sign in sign up
<<
Home , Aviculture, Battleground (2017), Bulweria, Coryphaena, Dodo, Echinometra mathaei

U.S. & Wildlife Service Midway Protection Project Final Environmental Assessment Sand Island, , Papahanaumokuakea- - Marine National Monument Front cover:

A male bonds with his chick Kukini,- a Hawaiian word for messenger. The mother - an albatross named Wisdom who is the world’s oldest known wild - is off foraging at sea. In her lifetime, Wisdom has raised over 35 chicks.

Credit: Kiah Walker/USFWS Volunteer Midway Seabird Protection Project Final Environmental Assessment

Sand Island, Midway Atoll, Papahanaumokuakea- - Marine National Monument

Prepared for: U.S. Fish and Wildlife Service

Cooperating Agency USDA APHIS Wildlife Services National Wildlife Research Center

Prepared by: Hamer Environmental L.P. and Planning Solutions, Inc.

January 2019

FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT EXECUTIVE SUMMARY

EXECUTIVE SUMMARY

Midway Atoll National Wildlife Refuge (MANWR) lies in the North approximately equidistant between North America and Asia. The refuge is also designated the Battle of Midway National Memorial and is within the Papahānaumokuākea Marine National Monument. The fringing , shallow lagoons, and 3 low-lying islands (Sand, Eastern, and Spit Islands), are the breeding grounds for millions of , the wintering grounds for thousands of shorebirds, and a refuge for critically endangered like the Hawaiian monk seal (Monachus schauinslandi) and Laysan (Anas laysanensis). Over 70% of the total global population of (Phoebastria immutabilis) breeds at the refuge; with the majority of the Midway population nesting on the 1,128-acre Sand Island. This Environmental Assessment (EA) documents an action proposed by the United States Fish and Wildlife Service (USFWS), Papahānaumokuākea Marine National Monument (PMNM) and presents the anticipated environmental effects of both the Proposed Action and a No Action alternative. The Proposed Action is to eradicate house mice (Mus musculus) from Sand Island by delivering a lethal dose of rodenticide to every in a manner that minimizes harm to island residents and the while still maintaining a high probability of success, and to maintain the island in rodent-free status in perpetuity. The toxicant to be employed as part of the Proposed Action would be Brodifacoum-25D Conservation, a pelleted rodenticide bait intended for conservation purposes for the control or eradication of invasive on islands or vessels. Implementation of the proposed action is currently being considered for Summer, 2019. This action was identified in the Papahānaumokuākea Monument Management Plan (PMMP), completed in December 2008, as Strategy AS-4 with a goal of developing an eradication plan within 5 years (USFWS 2008a). The need for the action was reinforced when, in 2015, mice were confirmed to be feeding on the backs and necks of adult albatross nesting on Sand Island, leading to nest abandonment and mortality of adults, and chicks. The Proposed Action was identified in the PMMP for many reasons, among them the fact that worldwide are a leading cause of island species including , bird, , and invertebrates. Forty to sixty % of all recorded bird and reptile extinctions are attributed to invasive rodents. The USFWS is proposing the Action to protect seabirds and their on MANWR’s Sand Island. The other islands of MANWR, Eastern and Spit, are not included because mice are not currently present on them. The USFWS is planning and would conduct the Action with technical support from Island Conservation (IC) and the Midway Restoration Partnership Group, which is a multidisciplinary stakeholder body including representatives from USFWS, Island Conservation, American Bird Conservancy, U.S. Department of Agriculture (USDA) and Plant Health Inspection Service (APHIS) National Wildlife Research Center (NWRC), the National Oceanographic and Atmospheric Agency (NOAA), U.S. Geological Survey (USGS), and the State of Hawaiʻi Office of Hawaiian Affairs. Removing mice would improve the MANWR’s ability to restore the natural island ecosystem, benefitting native coastal plants and . The proposed action would improve seabird nesting and could aid in the recovery of rare seabirds such as the short-tailed albatross (Phoebastria

ES-1 MIDWAY SEABIRD PROTECTION PROJECT FINAL ENVIRONMENTAL ASSESSMENT EXECUTIVE SUMMARY albatrus), Bulwer’s (Bulweria bulwerii), and Tristram’s storm-petrel (Oceanodroma tristrami). The proposed action involves the aerial broadcast of bait pellets containing rodenticide into all potential mouse territories on Sand Island along with supplemental hand broadcasting of bait in sensitive areas (e.g., around ponds that cannot be covered) and placing bait stations in commensal areas. The USFWS also considered other alternatives and methods to eradicate mice on Sand Island but these were dismissed from analysis since they failed to meet the purpose and need of the project. Mouse eradication would occur in the summer dry season to maximize the probability of success by targeting mice when their food resources are lowest and their abundance is declining. The proposed action also includes mitigation measures on Eastern Island where Laysan would be cared for and protected (see Appendix C). Conducting the operation during this period would also minimize the risk of rainfall washing rodenticide pellets into the ocean and is also a period of low winds, which makes bait application easier to control. It is also the time of year when relatively few wintering shorebirds, which may be susceptible to primary and secondary poisoning, are present; relatively few seabirds are also present at this time because the majority have completed breeding and have left the island. The lower numbers of seabirds present on MANWR during this time of year also reduces, but does not eliminate, the potential for collisions between operational aircraft and seabirds. The endangered Laysan duck, a year-round resident species, would be present at this time. Also, several species of water and land may be present during implementation. Protective measures would be implemented to avoid and minimize impacts to them. This EA evaluates the likelihood and magnitude of these incidental bird mortalities or injuries and describes the avoidance and minimization measures that will be implemented to reduce bird mortalities. This EA will also be used for consultation under Section 7 of the Act and for consulting with NOAA related to Essential Fish Habitat. The USFWS is responsible for the final decision on the Action along with the plans for implementation and monitoring. The USFWS anticipates that the Proposed Action would have a less than significant adverse impact on the environment when it is conducted as outlined in this EA, including the implementation of the avoidance, minimization, and mitigation policies and actions.

ES-2 FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT EXECUTIVE SUMMARY

Project: Midway Seabird Protection Project Agency: U.S. Fish and Wildlife Service Papahānaumokuākea Marine National Monument 300 Ala Moana Boulevard, Room 5-231 Honolulu, Hawai‘i 96850 Submit Comments to: [email protected] Cooperating Agencies: USDA APHIS NWRC Affected Location: Sand Island, MANWR, Northwestern Hawaiian Islands Proposed Action: Eradication of invasive mice on Sand Island, MANWR via rodenticide. Island & Project Area: 1,128 acres (456.5 hectares) Required Permits & Approvals: • National Environmental Policy Act (NEPA) Environmental Assessment • Section 7, Endangered Species Act (ESA) • Section 106, National Historic Preservation Act (NHPA) • Essential Fish Habitat • Federal Insecticide, Fungicide, and Rodenticide Act (FIFRA) Supplemental Label • National Pollutant Discharge Elimination System (NPDES) Permit • Marine National Monument Permit Anticipated Determination: No Significant Impacts Parties Consulted: See Chapter 7.0 Consultants: Hamer Environmental Planning Solutions, Inc.

ES-3 MIDWAY SEABIRD PROTECTION PROJECT FINAL ENVIRONMENTAL ASSESSMENT EXECUTIVE SUMMARY

THIS PAGE INTENTIONALLY LEFT BLANK

ES-4 FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT TABLE OF CONTENTS

TABLE OF CONTENTS

PURPOSE AND NEED ...... 1-1 1.1 INTRODUCTION ...... 1-1 1.2 PURPOSE OF THE ACTION ...... 1-2 1.3 NEED FOR THE ACTION ...... 1-3 1.3.1 Effects of Introduced Invasive Species on Islands ...... 1-3 1.3.2 Effects of Introduced mice on MANWR ...... 1-4 1.3.3 MANWR Ecosystem Management and conservation plans ...... 1-6 1.3.3.1 Papahānaumokuākea Monument Management Plan (PMMP) ...... 1-6 1.3.3.2 Conservation Action Plan for Black-footed albatross and Laysan albatross...... 1-7 1.3.3.3 U.S. Fish and Wildlife Service Regional Seabird Conservation Plan ...... 1-7 1.4 REGULATORY REQUIREMENTS ...... 1-7 1.5 PROJECT PARTICIPANTS ...... 1-9 1.5.1 U.S. Fish and Wildlife Service (USFWS) - Lead Agency ...... 1-9 1.5.2 U.S. Department of Agriculture, Animal and Plant Health Inspection Service, Wildlife services (WS) and the National Wildlife Research Center (NWRC) - Cooperating Agency ...... 1-10 1.5.3 Memorandum of Understanding (MOU) Between Island Conservation, USDA (APHIS), American Bird Conservancy and the USFWS ...... 1-10 1.5.4 Island Conservation ...... 1-10 1.5.5 Midway Restoration Partnership ...... 1-11 1.6 ORGANIZATION OF THE EA ...... 1-11 DESCRIPTION OF THE PROJECT AND ALTERNATIVES ...... 2-1 2.1 PRINCIPLES OF RODENT ERADICATION IN ISLAND ...... 2-1 2.2 MANWR TECHNICAL CONSIDERATIONS ...... 2-3 2.2.1 Constraints on Successful Eradication ...... 2-4 2.2.1.1 Manmade Structures ...... 2-4 2.2.1.2 Native Species...... 2-5 2.2.1.3 Resident Community on Sand Island ...... 2-6 2.2.2 Bait Distribution Approaches ...... 2-6 2.2.2.1 Bait Stations ...... 2-7 2.2.2.2 Broadcast ...... 2-7 2.2.2.2.1 Hand Broadcast ...... 2-7 2.2.2.2.2 Aerial Broadcast ...... 2-8 2.2.2.3 Summary ...... 2-8 2.2.3 Rodenticide and Bait Considerations ...... 2-9 2.3 PREFERRED ALTERNATIVE: MOUSE ERADICATION WITH RODENTICIDE VIA MULTIPLE DISTRIBUTION METHODS 2-11 2.3.1 Bait and Its Distribution Methods ...... 2-11 2.3.1.1 Primary Distribution Method – Aerial Broadcast ...... 2-11 2.3.1.2 Secondary Distribution Method – Supplemental Hand Broadcast ...... 2-13 2.3.2 Number and Rate of Bait Applications ...... 2-14 2.3.3 Mechanical Trapping ...... 2-15 2.3.4 Pre-Application Preparations ...... 2-15 2.3.5 Application Schedule ...... 2-16 2.3.6 Mitigation Actions and Effectiveness Monitoring ...... 2-16 2.3.7 Post-Application Actions ...... 2-17 2.3.8 Contingency Planning ...... 2-17 2.4 ALTERNATIVE 2: NO ACTION ...... 2-17

i

2.5 ALTERNATIVES CONSIDERED BUT REJECTED ...... 2-19 2.5.1 Emerging Technologies ...... 2-19 2.5.2 Mouse Eradication Using Cholecalciferal (Agrid3) Rodenticide ...... 2-20 2.5.3 Use of First-Generation Rodenticide ...... 2-20 2.5.4 Mouse-Proofing Structures ...... 2-21 2.5.5 Aerial Broadcast Rodenticide Only ...... 2-22 2.5.6 Hand Broadcast Rodenticide Only ...... 2-22 2.5.7 Bait Station Rodenticide Only ...... 2-22 2.5.7.1 Non-Target Species ...... 2-22 2.5.7.2 Operational Challenges ...... 2-23 2.5.8 Mouse Control ...... 2-23 2.6 PROJECT COSTS ...... 2-24 AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, & MINIMIZATION MEASURES ...... 3-1 3.1 GENERAL DESCRIPTION OF MANWR ...... 3-1 3.2 RESOURCE TOPICS EXCLUDED FROM DETAILED ANALYSIS ...... 3-2 3.2.1 Land Use ...... 3-2 3.2.1.1 Wildlife Refuge ...... 3-4 3.2.1.2 National Memorial ...... 3-5 3.2.1.3 Airport ...... 3-5 3.2.2 Access ...... 3-5 3.2.2.1 Air Transportation ...... 3-5 3.2.2.2 Marine Transportation ...... 3-5 3.2.3 Climate ...... 3-6 3.2.4 Visual Resources ...... 3-7 3.2.5 Fuel Facilities ...... 3-7 3.2.6 Energy ...... 3-7 3.2.7 Communications ...... 3-7 3.2.8 Air Quality ...... 3-8 3.2.9 Geologic Resources ...... 3-8 3.2.10 Socioeconomic and Environmental Justice ...... 3-8 3.2.11 Solid Waste Disposal ...... 3-9 3.2.12 Historical and Cultural Resources and Values...... 3-9 3.2.12.1 Programmatic Agreement and Treatment of Midway’s Historic Properties ...... 3-11 3.2.12.2 Historic Preservation Plan ...... 3-11 3.3 RESOURCES TOPICS ANALYZED IN DETAIL IN THIS EA ...... 3-11 3.3.1 Airfield Operations ...... 3-11 3.3.1.1 Existing Conditions ...... 3-11 3.3.1.2 Probable Impacts ...... 3-12 3.3.2 Water Resources ...... 3-12 3.3.2.1 Existing Conditions: Groundwater ...... 3-12 3.3.2.2 Existing Conditions: Surface Water ...... 3-13 3.3.2.3 Existing Conditions: Potable and Non-Potable Water Systems ...... 3-13 3.3.2.4 Probable Impacts ...... 3-14 3.3.3 Noise ...... 3-16 3.3.3.1 Existing Conditions ...... 3-16 3.3.3.2 Probable Impacts ...... 3-17 3.3.4 Wastewater and Storm water ...... 3-18 3.3.4.1 Existing Conditions ...... 3-18 3.3.4.2 Probable Impacts ...... 3-19 3.3.5 Hazardous Materials and Wastes ...... 3-19

ii FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT TABLE OF CONTENTS

3.3.5.1 Existing Conditions ...... 3-19 3.3.5.2 Probable Impacts ...... 3-22 3.3.6 Biological Resources ...... 3-23 3.3.6.1 Evaluation Criteria for Effects on Biological Resources ...... 3-23 3.3.6.2 Non-Target Species ...... 3-25 3.3.6.3 Environmental Fate of Brodifacoum in Soil and Water ...... 3-28 3.3.6.4 Terrestrial Environment ...... 3-32 3.3.6.4.1 Terrestrial Vegetation: Existing Conditions ...... 3-32 3.3.6.4.2 Dominant Vegetation Types ...... 3-33 3.3.6.4.3 Endangered Plants ...... 3-36 3.3.6.4.4 Probable Impacts: No Action Alternative ...... 3-37 3.3.6.4.5 Probable Impacts: Preferred Alternative ...... 3-38 3.3.6.5 Seabirds ...... 3-39 3.3.6.5.1 Existing Conditions: ...... 3-40 3.3.6.5.2 Existing Conditions: Nocturnal seabirds ...... 3-41 3.3.6.5.3 Existing Conditions: Noddies ...... 3-42 3.3.6.5.4 Existing Conditions: ...... 3-43 3.3.6.5.5 Existing Conditions: Great ...... 3-43 3.3.6.5.6 Existing Conditions: ...... 3-43 3.3.6.5.7 Existing Conditions: Boobies ...... 3-44 3.3.6.5.8 Probable Impacts: No Action Alternative ...... 3-44 3.3.6.5.9 Probable Impacts: Preferred Alternative ...... 3-45 3.3.6.6 Shorebirds ...... 3-54 3.3.6.6.1 Existing Conditions ...... 3-54 3.3.6.6.2 Pacific golden plover ...... 3-54 3.3.6.6.3 Ruddy turnstone ...... 3-55 3.3.6.6.4 Wandering tattler ...... 3-55 3.3.6.6.5 Sanderling ...... 3-55 3.3.6.6.6 Probable Impacts: No Action Alternative ...... 3-58 3.3.6.6.7 Probable Impacts: Preferred Alternative ...... 3-58 3.3.6.7 Bristle-thighed curlew ...... 3-62 3.3.6.7.1 Existing Conditions ...... 3-62 3.3.6.7.2 Probable Impacts: No Action Alternative ...... 3-64 3.3.6.7.3 Probable Impacts: Preferred Alternative ...... 3-64 3.3.6.8 Laysan duck ...... 3-68 3.3.6.8.1 Existing Conditions ...... 3-68 3.3.6.8.2 Probable Impacts: No Action Alternative ...... 3-70 3.3.6.8.3 Probable Impacts: Preferred Alternative ...... 3-71 3.3.6.9 Land Birds ...... 3-81 3.3.6.9.1 Existing Conditions ...... 3-81 3.3.6.9.2 Probable Impacts: No Action Alternative ...... 3-84 3.3.6.9.3 Probable Impacts: Preferred Alternative ...... 3-84 3.3.6.10 Cattle Egrets ...... 3-86 3.3.6.10.1 Existing Conditions ...... 3-86 3.3.6.10.2 Probable Impacts: No Action Alternative ...... 3-87 3.3.6.10.3 Probable Impacts: Preferred Alternative ...... 3-87 3.3.6.11 Terrestrial ...... 3-88 3.3.6.11.1 Existing Conditions ...... 3-88 3.3.6.11.2 Probable Impacts: No Action Alternative ...... 3-88 3.3.6.11.3 Probable Impacts: Preferred Alternative ...... 3-88 3.3.6.12 Herpetofauna ...... 3-89 3.3.6.12.1 Existing Conditions ...... 3-89 3.3.6.12.2 Probable Impacts: No Action Alternative ...... 3-89 3.3.6.12.3 Probable Impacts: Preferred Alternative ...... 3-89 3.3.6.13 Terrestrial Invertebrates ...... 3-90 3.3.6.13.1 Existing Conditions ...... 3-90

iii

3.3.6.13.2 Probable Impacts: No Action Alternative ...... 3-90 3.3.6.13.3 Probable Impacts: Preferred Alternative ...... 3-90 3.3.6.14 Marine Fish ...... 3-91 3.3.6.14.1 Existing Conditions ...... 3-91 3.3.6.14.2 Probable Impacts: No Action Alternative ...... 3-92 3.3.6.14.3 Probable Impacts: Preferred Alternative ...... 3-92 3.3.6.15 Coral ...... 3-95 3.3.6.15.1 Existing Conditions ...... 3-95 3.3.6.15.2 Probable Impacts: No Action Alternative ...... 3-97 3.3.6.15.3 Probable Impacts: Preferred Alternative ...... 3-97 3.3.6.16 Essential Fish Habitat ...... 3-98 3.3.6.16.1 Existing Conditions ...... 3-98 3.3.6.16.2 Probable Impacts: No Action Alternative ...... 3-103 3.3.6.16.3 Probable Impacts: Preferred Alternative ...... 3-103 3.3.6.17 Marine Invertebrates ...... 3-104 3.3.6.17.1 Existing Conditions ...... 3-104 3.3.6.17.2 Probable Impacts: No Action Alternative ...... 3-105 3.3.6.17.3 Probable Impacts: Preferred Alternative ...... 3-105 3.3.6.18 ...... 3-106 3.3.6.18.1 Existing Conditions ...... 3-106 3.3.6.18.2 Probable Impacts: No Action Alternative ...... 3-107 3.3.6.18.3 Probable Impacts: Preferred Alternative ...... 3-107 3.3.6.19 Sea Turtles ...... 3-108 3.3.6.19.1 Existing Conditions ...... 3-108 3.3.6.19.2 Probable Impacts: No Action Alternative ...... 3-110 3.3.6.19.3 Probable Impacts: Preferred Alternative ...... 3-111 3.3.6.20 Marine Mammals, Rays, and Sharks ...... 3-113 3.3.6.20.1 Existing Conditions: Dolphins and Hawaiian Monk Seals:...... 3-113 3.3.6.20.2 Probable Impacts: No Action Alternative ...... 3-117 3.3.6.20.3 Probable Impacts: Preferred Alternative ...... 3-117 3.3.6.20.4 Existing Conditions: False Killer Whale ...... 3-120 3.3.6.20.5 Existing Conditions: Blue Whale ...... 3-120 3.3.6.20.6 Existing Conditions: Fin Whale ...... 3-121 3.3.6.20.7 Existing Conditions: Sei Whale...... 3-121 3.3.6.20.8 Existing Conditions: Sperm Whale ...... 3-121 3.3.6.20.9 Existing Conditions: North Pacific Right Whales ...... 3-121 3.3.6.20.10 Existing Conditions: Giant Manta Ray ...... 3-122 3.3.6.20.11 Existing Conditions: Oceanic Whitetip Shark ...... 3-123 3.3.6.20.12 Probable Impacts: No Action Alternative ...... 3-124 3.3.6.20.13 Probable Impacts: Preferred Alternative ...... 3-124 3.3.6.21 Human Health and Safety ...... 3-124 3.3.6.21.1 Existing Environment ...... 3-124 3.3.6.21.2 Aerial Operations Safety ...... 3-125 3.3.6.21.3 Ground Operations Safety ...... 3-126 3.3.6.21.4 Probable Impacts ...... 3-126 CUMULATIVE IMPACTS ...... 4-1 4.1 CUMULATIVE IMPACTS ...... 4-1 4.2 PAST ACTIONS ...... 4-1 4.3 CURRENT ACTIONS ...... 4-2 4.3.1 Plant Restoration and Invasive Weed Control ...... 4-2 4.3.2 Lead Abatement ...... 4-2 4.3.3 Marine Debris Removal ...... 4-2 4.3.4 Mouse Control/Management efforts ...... 4-2

iv FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT TABLE OF CONTENTS

4.4 FUTURE ACTIONS ...... 4-3 4.5 PROBABLE CUMULATIVE IMPACTS ...... 4-3 4.5.1 No Action Alternative ...... 4-3 4.5.2 Preferred Alternative ...... 4-4 LIST OF PREPARERS AND CONTRIBUTORS ...... 5-1 5.1 PREPARERS ...... 5-1 5.2 CONTRIBUTORS ...... 5-1 REFERENCES ...... 6-1 CONSULTATION & DISTRIBUTION ...... 7-1 7.1 PRE-DRAFT SCOPING PROCESS ...... 7-1 7.2 INTERVIEW PROCESS ...... 7-2 STATEMENT OF COMPLIANCE FOR IMPLEMENTATION OF THE PROPOSED ACTION AT MANWR ...... 8-1 8.1 MIDWAY SEABIRD PROTECTION PROJECT NEPA AND COMPLIANCE PROCESS ...... 8-1 8.2 NATIONAL ENVIRONMENTAL POLICY ACT (1969) ...... 8-1 8.3 ENDANGERED SPECIES ACT OF 1973 ...... 8-1 8.4 MAGNUSON-STEVENS ACT ESSENTIAL FISH HABITAT ...... 8-1 8.5 PAPAHĀNAUMOKUĀKEA MARINE NATIONAL MONUMENT (PMNM) PERMIT ...... 8-1 8.6 NATIONAL HISTORIC PRESERVATION ACT, SECTION 106 (1966) ...... 8-2 8.7 EPA- APPROVED RODENTICIDE LABEL ...... 8-2 8.8 EO 12372 INTERGOVERNMENTAL REVIEW ...... 8-2 8.9 MIGRATORY BIRD TREATY ACT AND EO 13186 FEDERAL PROTECTION OF MIGRATORY BIRDS ...... 8-2 8.10 EO 13112 RESPONSIBILITIES OF FEDERAL AGENCIES PERTAINING TO INVASIVE SPECIES ...... 8-2 8.11 INTEGRATED PEST MANAGEMENT, 517 DM 1 AND 569 FW 1 ...... 8-3 8.12 EPA – NATIONAL POLLUTANT DISCHARGE ELIMINATION SYSTEM (NPDES) GENERAL PERMIT ...... 8-3 GLOSSARY OF TERMS ...... 9-1 APPENDIX A. MIDWAY ATOLL NATIONAL WILDLIFE REFUGE BIO-SECURITY PLAN ...... A-1 APPENDIX B. SUMMARY OF PROPOSED MONITORING ...... B-1 APPENDIX C. SUMMARY OF RISKS AND PROPOSED PROTECTION PLAN FOR LAYSAN DUCKS AND SHOREBIRDS ...... C-1 APPENDIX D. SEEP CHECK PROTOCOL ...... D-1 APPENDIX E. CAPTIVE CARE PROTOCOL ...... E-1 APPENDIX F. PUBLIC COMMENTS AND AGENCY RESPONSES ...... F-1

v

LIST OF FIGURES

FIGURE 1.1 LOCATION MAP ...... 1-1 FIGURE 1.2 MIDWAY ATOLL NATIONAL WILDLIFE REFUGE (MANWR) ...... 1-2 FIGURE 1.3 SAND ISLAND, MANWR ...... 1-3 FIGURE 1.4 ADULT LAYSAN ALBATROSS ON SAND ISLAND SHOWING EFFECTS OF PREDATORY MICE ...... 1-5 FIGURE 1.5 MOUSE BITING THE HEAD OF ADULT LAYSAN ALBATROSS ON SAND ISLAND ...... 1-5 FIGURE 2.1 GLOBAL MOUSE ERADICATION ATTEMPTS ON ISLANDS OVER TIME AND BY ISLAND SIZE ...... 2-2 FIGURE 3.1 GENERALIZED LAND USES ON SAND ISLAND ...... 3-4 FIGURE 3.2 WATER RESOURCES ON SAND ISLAND, MANWR ...... 3-14 FIGURE 3.3 MONTHLY SPECIES ABUNDANCE OR BREEDING ACTIVITY ON SAND ISLAND, MANWR ...... 3-27 FIGURE 3.4 MAP OF VEGETATIVE COVER ON SAND ISLAND, MANWR ...... 3-33 FIGURE 3.5 MEAN COUNTS FOR YEARS 1993-2001 AND 2017 OF BRISTLE-THIGHED CURLEW, PACIFIC GOLDEN PLOVER, WANDERING TATTLER, AND RUDDY TURNSTONE COMBINED BY MONTH ON MANWR ...... 3-56 FIGURE 3.6 MONTHLY COUNTS ON MANWR CONDUCTED IN 2017 FOR BRISTLE-THIGHED CURLEWS (BTCU), PACIFIC GOLDEN PLOVERS (PGPL), WANDERING TATTLERS (WATA), AND RUDDY TURNSTONES (RUTU) ...... 3-58 FIGURE 3.7 BRISTLE-THIGHED CURLEW ...... 3-63 FIGURE 3.8 LAYSAN DUCK ...... 3-69 FIGURE 3.9 EXPECTED BRODIFACOUM DEGRADATION CURVE FOLLOWING AN ERADICATION OPERATION...... 3-79 FIGURE 3.10 RESULTS OF ISLAND-WIDE MONTHLY AVIAN SURVEYS CONDUCTED FROM JUNE THROUGH SEPTEMBER 2017 FOR LAND BIRDS ...... 3-83 FIGURE 3.11 SPATIAL DISTRIBUTION OF BENTHIC COVER AND CORAL COMPOSITION AT NOAA CREP RAPID ECOLOGICAL ASSESSMENT SITES AT MIDWAY IN 2008 ...... 3-97 FIGURE 3.12 NOAA MIDWAY ISLAND DETAILED HABITAT COVER IMAGE AT 13M RESOLUTION...... 3-99 FIGURE 3.13 AUGMENTATION OF SAND ISLAND DURING THE COLD WAR ...... 3-101 FIGURE 3.14 MARINE SURVEY AREA FOR SEAWALL REPAIR PROJECT ...... 3-102 FIGURE 3.15 POSSIBLE CORAL TRANSLOCATION SITE FOR THE SEAWALL REPAIR PROJECT ...... 3-102 FIGURE 3.16 HAWAIIAN MONK SEAL ...... 3-113 FIGURE 3.17 SAND ISLAND HAWAIIAN MONK SEAL SIGHTINGS: 2012-2016...... 3-115

vi FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT TABLE OF CONTENTS

LIST OF TABLES

TABLE 1.1 STRATEGY AS-4 OF THE PMMP ...... 1-7 TABLE 2.1 PRINCIPLES OF SUCCESSFUL RAT AND MOUSE ERADICATIONS ...... 2-3 TABLE 2.2 MANMADE STRUCTURES ON SAND ISLAND, MANWR ...... 2-5 TABLE 2.3 SUMMARY OF TECHNICAL CONSIDERATIONS ...... 2-9 TABLE 2.4 RODENTICIDES APPROVED FOR CONSERVATION USE IN THE UNITED STATES ...... 2-10 TABLE 3.1 MANWR ADMINISTRATIVE CO-TRUSTEES ...... 3-1 TABLE 3.2 SUMMARY OF THE BUILT ENVIRONMENT ON SAND ISLAND, MANWR ...... 3-3 TABLE 3.3 AVERAGE RAINFALL ON SAND ISLAND, MANWR ...... 3-7 TABLE 3.4 TYPICAL PERSONNEL PRESENT ON SAND ISLAND, MANWR ...... 3-9 TABLE 3.5 EFFECTIVE PERCEIVED NOISE LEVEL OF SELECTED AIRCRAFT ...... 3-18 TABLE 3.6 LUC SITES ON SAND ISLAND, MANWR ...... 3-21 TABLE 3.7 PROJECT RELATED HAZARDOUS MATERIALS, WASTE, AND TOXINS ...... 3-22 TABLE 3.8 GENERALIZED PROPORTION OF DAILY FOOD INTAKE THAT MUST BE BAIT FOR BIRDS TO REACH AN LD50 THRESHOLD WHERE THE DOSAGE (D) OF A TOXIN IS LETHAL (L) TO 50% OF OF THE SPECIES IN LABORATORY TESTS ...... 3-24 TABLE 3.9 SUMMARY OF RISKS TO NONTARGET SPECIES ...... 3-25 TABLE 3.10 PRELIMINARY RISK ASSESSMENT BY SPECIES, CONSEQUENCE, AND POTENTIAL MITIGATION ...... 3-29 TABLE 3.11 VEGETATION COMMUNITIES AND PLANT ASSOCIATIONS ON SAND ISLAND, MANWR ...... 3-34 TABLE 3.12 BREEDING SEABIRDS OF MANWR (N = 21 SPECIES) ...... 3-40 TABLE 3.13 PRIMARY AND SECONDARY TOXICITY OF BRODIFACOUM ...... 3-47 TABLE 3.14 NUMBER AND SPECIES OF BIRDS REPORTED STRUCK BY AIRPLANES AT HENDERSON FIELD, MANWR: 2004-2010...... 3-51 TABLE 3.15 DUCKS, SHOREBIRDS, AND WADING BIRDS OF MANWR (N = 7 SPECIES) ...... 3-54 TABLE 3.16 PRIMARY AND SECONDARY TOXICITY OF BRODIFACOUM TO SHOREBIRDS ON MANWR ALONG WITH THE LAYSAN DUCK AND ...... 3-60 TABLE 3.17 ESTIMATED NUMBER OF INJURIES AND MORTALITIES TO EACH LAYSAN DUCK AGE FROM THE PROPOSED PROJECT ...... 3-73 TABLE 3.18 PRIMARY AND SECONDARY TOXICITY OF BRODIFACOUM TO ON MANWR ...... 3-85 TABLE 3.19 ADDITIONAL SPECIES RECORDED AT EACH SITE AT MANWR DURING ROVING DIVER SURVEY ...... 3-107 TABLE 3.20 HAWAIIAN MONK SEAL BEACH COUNT DATA SUMMARIZED BY SECTOR ...... 3-116 TABLE 3.21 HAWAIIAN MONK SEAL BEACH COUNT DATA SUMMARIZED BY MONTH ...... 3-117 TABLE 5.1 PRINCIPAL PREPARES OF THIS ENVIRONMENTAL ASSESSMENT ...... 5-1 TABLE 5.2 CONTRIBUTORS TO THIS ENVIRONMENTAL ASSESSMENT...... 5-1 TABLE 7.1 SCOPING LETTER RECIPIENTS ...... 7-1 TABLE 7.2 SUMMARY OF SCOPING INTERVIEW...... 7-2

vii MIDWAY SEABIRD PROTECTION PROJECT FINAL ENVIRONMENTAL ASSESSMENT LIST OF ACRONYMS

ACRONYMS AND ABBREVIATIONS

Acronym Phrase a.i. Active Ingredient ARFF Aircraft Rescue and Fire Fighting ARMCO 1942 Battle of Midway structure (hut) ASAP Alien Species Action Plan AST Above-ground Storage Tank BASH Bird Aircraft Strike Hazard BMP Best Management Practices BOQ Bachelor Officer Quarters CAMU Corrective Action Management Unit CERCLA Comprehensive Environmental Response, Compensation, and Liability Act CFR Code of Federal Regulations COPC Chemicals of Potential Concern CR Critically Endangered D Dose dB Decibels DNL Day-Night Average - Noise Level EA Environmental Assessment EC50 Half Maximal Effective Concentration (of a toxicant; measured in ppm) EIS Environmental Impact Statement EPNdB Effective Perceived Noise Level (in decibels) EO Executive Order ESA Endangered Species Act of 1973 ETOPS Extended Twin-Engine Operations FAA Federal Aviation Association FIRFA Federal Insecticide, Fungicide and Rodenticide Act FOD Foreign Object Debris FONSI Finding of No Significant Impact GIS Geographic Information System GPS Global Positioning Satellite HDLNR Hawai‘i Department of Land and Natural Resources IUCN International Union for Conservation of Nature L Lethal LD50 Lethal Dosage of a toxin to 50 % of laboratory tested animals; measured in mg (D)/kg of animal body weight LOAEL Lowest Observable Adverse Effect Level LUC Land Use Controls MANWR Midway Atoll National Wildlife Refuge/Battle of Midway National Memorial MBTA Migratory Bird Treaty Act MG Million Gallon MMPA Marine Mammal Protection Act MSL Mean Sea Level

viii FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT LIST OF ACRONYMS

Acronym Phrase NAF Naval Air Facilities NEPA National Environmental Policy Act NGO Non-governmental Organization NHPA National Historic Preservation Act NISA National Invasive Species Act NISC National Invasive Species Council NOAEL No Observable Adverse Effect Level NT Near Threatened NWHI Northwestern Hawaiian Islands NWR National Wildlife Refuge NWRC National Wildlife Research Center/Complex NWRS National Wildlife Refuge System OBWLF Old Bulky Waste Landfill OSHA Occupational Safety and Health Administration PCB Polychlorinated Biphenyl pH Potential of Hydrogen PMDY Henderson Field Airport International Air Transport Assoc. Airport Code PMMP Papahānaumokuākea Monument Management Plan PMNM Papahānaumokuākea Marine National Monument PPE Personal Protection Equipment ppm Parts Per Million RCRA Resource Conservation and Recovery Act SARA Superfund Amendments and Reauthorization Act SMP Structure Management Plan TACA Toxic Substances Control Act USC United States Code of Laws USDA U.S. Department of Agriculture USEPA United States Environmental Protection Agency USFWS U.S. Fish and Wildlife Service UST Underground Storage Tank

ix MIDWAY SEABIRD PROTECTION PROJECT FINAL ENVIRONMENTAL ASSESSMENT LIST OF ACRONYMS

THIS PAGE INTENTIONALLY LEFT BLANK

x FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT PURPOSE AND NEED

PURPOSE AND NEED

1.1 INTRODUCTION This Environmental Assessment (EA) evaluates the potential environmental effects of one action alternative and the No Action Alternative for mouse eradication on Sand Island, Midway Atoll National Wildlife Refuge (MANWR). The MANWR is located within the Papahānaumokuākea Marine National Monument (PMNM), located in the Northwestern Hawaiian Islands (NWHI) (see Figure 1.1, Figure 1.2, and Figure 1.3). Over 3 million birds, encompassing 25 different species, can be found on MANWR’s 3 islands and all of them are susceptible to by mice. MANWR is home to the largest albatross colony in the world and is the most important and successful breeding ground for black-footed albatross (Phoebastria nigripes) and Laysan albatross (Phoebastria immutabilis). The NWHI are, respectively, home to approximately 97.5 and 99.7% of the total worldwide population of these 2 species, and both of these species are considered “near threatened” (NT) by the International Union for Conservation of Nature (IUCN). MANWR alone is globally significant, supporting 36% of all Black-footed albatross and 73% of all Laysan albatross. Sand Island alone has approximately 360,000 pairs of Laysan albatross and 15,084 pairs of black-footed albatross. As of 2004, MANWR also now hosts a translocated population of Laysan ducks (Anas laysanensis), which are considered “critically endangered” (CR) by the IUCN and is listed as endangered under the Endangered Species Act of 1973 (ESA), as amended. This final EA was used by U.S. Fish and Wildlife Service (USFWS) to solicit public input and determine whether implementation of the alternatives will have a significant impact on the environment. This EA is part of the USFWS decision-making process in accordance with the National Environmental Policy Act (NEPA), as amended and its implementing regulations.

Figure 1.1 Location Map

Source: https://www.papahanaumokuakea.gov/visit/ (2017)

1-1 MIDWAY SEABIRD PROTECTION PROJECT FINAL ENVIRONMENTAL ASSESSMENT PURPOSE AND NEED 1.2 PURPOSE OF THE ACTION The purpose of the proposed action is to implement Strategy AS-4 from the PMMP and completely eradicate the invasive house mouse from Sand Island within the MANWR and to maintain its rodent-free status in perpetuity. To eradicate invasive mice, a lethal dose of rodenticide would be delivered to every rodent on the island in a manner that minimizes harm to island residents and the ecosystem while still maintaining a high probability of successful eradication.

Figure 1.2 Midway Atoll National Wildlife Refuge (MANWR)

Source: Planning Solutions, Inc. (2017)

Within 1 year of project implementation, non-native mice will be eradicated (population = 0) from Sand Island on Midway Atoll National Wildlife Refuge for the benefit and protection of nesting albatross species (e.g., Laysan, short-tailed, and black-footed), other nesting seabirds (e.g., Bonin Petrel), and their habitats.

1-2 FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT PURPOSE AND NEED

Figure 1.3 Sand Island, MANWR

Source: Planning Solutions, Inc. (2017)

1.3 NEED FOR THE ACTION

1.3.1 EFFECTS OF INTRODUCED INVASIVE SPECIES ON ISLANDS Worldwide, invasive species are causing negative ecological and economic impacts. The impacts from invasive predatory mammals, including mice and rats, are one of the leading causes of species on islands (Blackburn et al. 2004; Duncan and Blackburn 2007). The extinction of many island species of mammal, bird, reptile, and invertebrate have been attributed to the impact of invasive rodents (Andrews 1909, Atkinson 1985, Daniel and Williams 1984, Hutton et al. 1984, Tomich 1986), with estimates of between 40-60% of all recorded bird and reptile extinctions globally being directly attributable to invasive rodents (Atkinson 1985).

1-3 MIDWAY SEABIRD PROTECTION PROJECT FINAL ENVIRONMENTAL ASSESSMENT PURPOSE AND NEED

1.3.2 EFFECTS OF INTRODUCED MICE ON MANWR Invasive house mice and black rats (Rattus rattus) became established on MANWR’s Sand Island more than 75 years ago during military occupancy. Black rats were eradicated in 1996 and house mice are now the sole rodent and invasive mammal present in the NWHI.1 Mouse predation of albatross on Sand Island was first identified in 2015 and the area where mouse predation occurs and the number of birds bitten has been observed to increase with each successive year. Mice were confirmed as the ultimate cause of for bitten Laysan albatross during the 2015/2016 breeding season on MANWR with 42 dead adults and 70 nests abandoned in a 4 ac. (1.6 ha) area (DuhrSchultz et al. in press). Time lapse photography recorded mice repeatedly crawling onto and biting the head, neck, and backs of adult birds (see Figure 1.4 and Figure 1.5). Necropsy and histopathology results from recovered carcasses indicate bacterial infection from wounds as the cause of death (USGS, National Wildlife Health Center, Honolulu). This same phenomenon was again present in the 2016/2017 breeding season with a 6-fold increase in the number of affected albatross to 242 dead adults, 1,218 bitten birds, and 994 abandoned nests and site locations from 3 to 50 distinct areas, as well as an expansion of total affected area from 4 acres to 27 acres (11 ha) (USFWS, Unpublished b). Mouse attacks declined in 2017/2018, however, the reason for the decline is not known. A combination of environmental conditions, the cyclical nature of mouse populations, and active management to reduce attacks are thought to have played a role. Beyond the predation of seabirds, rodents have deleterious effects on entire island ecosystems. Briefly summarized, they:

• Are a disease vector.

• Feed opportunistically on plants and alter the floral communities of island ecosystems (Campbell and Atkinson 2002). Their feeding competes with native species and in some cases degrades the quality of nesting habitat for birds that depend on the vegetation (Wegmann 2009, Young et al. 2010).

• Prey upon terrestrial invertebrates and , affecting the abundance and age structure of intertidal invertebrates directly (Navarrete and Castilla 1993).

• Have been demonstrated to lower total soil carbon, nitrogen, phosphorous, mineral nitrogen, marine-derived nitrogen, and pH relative to rodent-free islands (Fukami et al. 2006).

• Affect invertebrate and marine algal abundance, changing intertidal community structure from an algae-dominated system in rodent free areas to an invertebrate dominated system in rodent-invaded areas (Kurle et al. 2008).

1 The 1996 eradication effort was geared solely towards black rats and was not intended to eliminate house mice.

1-4 FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT PURPOSE AND NEED

Figure 1.4 Adult Laysan albatross on Sand Island showing effects of predatory mice

Source: USFWS (2016)

Figure 1.5 Mouse biting the head of Adult Laysan albatross on Sand Island

Source: USFWS (2016)

A comparison of rodent-infested and rodent-free islands, and pre- and post-rodent eradication projects have shown that rodents depressed the population size and recruitment of birds (Campbell 1991, Jouventin et al. 2003, Thibault 1995), reptiles (Bullock 1986, Cree et al. 1995, Towns 1991, Whitaker 1973), plants (Pye et al. 1999), and terrestrial invertebrates (Bremner et al. 1984, Campbell et al. 1984). In an analysis by Brooke et al. (2017), drawing on data of eradication of non-native mammals on islands from around the world, they examined the population growth rates (lambda) of 181 seabird populations of 69 seabird species following successful mammal

1-5 MIDWAY SEABIRD PROTECTION PROJECT FINAL ENVIRONMENTAL ASSESSMENT PURPOSE AND NEED

eradication projects. Rate of growth (lambda) is the ratio of population size at the end of one interval to population size at the end of the previous interval. When lambda = 1.0 the population density is stable while values >1.0 indicate increasing populations. After successful eradication, the median population growth rate (lambda) was 1.119 and seabird populations with positive growth-rates greatly outnumbered those in decline. Their study confirmed that invasive mammal eradication is usually followed by growth of seabird populations. Predation of vulnerable populations of native seabirds is a real and ongoing threat on Sand Island and demands an immediate and effective response. Eradication of the house mouse from MANWR would also facilitate the protection and restoration of multiple native species and habitats present in the refuge.

1.3.3 MANWR ECOSYSTEM MANAGEMENT AND CONSERVATION PLANS

1.3.3.1 Papahānaumokuākea Monument Management Plan (PMMP) The PMMP, completed in December 2008, was the result of an extensive public review process and identified 6 priority management needs, with supporting action plans and corresponding desired outcomes, for the PMNM. The 6 priorities were (PMMP, ES-3 to ES-5): 1. Understanding and Interpreting the Northwestern Hawaiian Islands; 2. Conserving Wildlife and Habitats; 3. Reducing Threats to Monument Resources; 4. Managing Human Uses; 5. Coordinating Conservation and Management Activities; and 6. Achieving Effective Monument Operations. As evidenced by these priorities, protecting the land and waters of the NWHI from the impacts of alien species is critical to achieving the Monument’s primary goal of resource protection. A specific component of Priority 3 was the development of an Alien Species Action Plan (ASAP). The ASAP advanced a series of strategies to address and mitigate the ongoing effects of invasive alien species on the PMNM. Specifically, the ASAP identified the eradication of the house mouse on Sand Island, MANWR as Strategy AS-4 (see Table 1.1).

1-6 FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT PURPOSE AND NEED

Table 1.1 Strategy AS-4 of the PMMP Strategy AS-4: Eradicate the house mouse population on Sand Island, Midway Atoll, within 15 years. After the eradication of the black rat (Rattus rattus) at Midway Atoll and the Polynesian rat (Rattus exulans) at Kure Atoll, the house mouse (Mus musculus) on Sand Island, Midway, remains the only invasive mammal left in the NWHI. Mice can cause high mortality in seabirds as large as albatrosses (Wanless et al. 2007). In addition, Midway now hosts a translocated population of endangered Laysan ducks that are likely to be negatively affected by the high mouse populations. Mice are also a threat to native plants and terrestrial invertebrates. Activity AS-4.1: Produce a house mouse eradication plan within 5 years and procure appropriate permits for chosen eradication techniques. The eradication of introduced rodents from islands is routine, and the successful removal of black rats at Midway Atoll in recent years has provided a model for mouse eradication. Mice present additional challenges, however, as they have much smaller home range sizes and different foraging and reproductive . A careful planning effort that emphasizes the minimization of effects to nontarget organisms at the site and the other biological differences that may affect the operation is necessary. Activity AS-4.2: Implement and complete house mouse eradication. All of Sand Island 1,128 acres (456 ha) will be treated with rodenticide, with active management to prevent nontarget impacts to native wildlife. Surveys of the affected ecosystem components before and after the operation will provide a valuable demonstration of the effects of introduced mice on biological communities. Source: Alien Species Action Plan, Papahānaumokuākea Monument Management Plan (USFWS, December 2008a)

One reason the PMMP-ASAP identified Strategy AS-4 was Wanless et al.’s (2007) observations that mice can and will predate seabirds once their competitors are removed. During the preparation of the PMMP, Wanless et al. (2007) observed mice depredating on Tristan albatross (Diomedea dabbenena) and Atlantic (Pterodroma incerta) on Gough Island in the South Atlantic Ocean. Based on their experience on Gough Island, they suggested that once mice are released from the ecological effects of other predators and competitors (e.g., on Sand Island, MANWR where rats have been removed and mice are the sole alien mammal), the mice may begin to predate seabird chicks. Since completion of the PMMP in 2008, Wanless et al.’s predictions have been shown to be accurate on Sand Island, MANWR where mice are now the sole alien mammal.

1.3.3.2 Conservation Action Plan for Black-footed albatross and Laysan albatross In a multi-agency report on conservation action plans for the black-footed albatross and Laysan albatross (Naughton et al. 2007), the authors state that one of the primary threats to these 2 species is predation by introduced mammals. The authors recommended a conservation action to eradicate house mice from MANWR to protect these 2 species of seabirds.

1.3.3.3 U.S. Fish and Wildlife Service Regional Seabird Conservation Plan In addition, the USFWS Regional Seabird Conservation Plan (USFWS 2005) identified the eradication or control of introduced predators and other invasive species that have a negative impact on seabirds as one of the top priorities for seabird conservation. They specifically recommend eradicating introduced predators in the Pacific Islands where the Laysan and black- footed albatross breed or have historically bred.

1.4 REGULATORY REQUIREMENTS The proposed action of eradicating invasive mice on Sand Island, MANWR by the USFWS, Papahānaumokuākea Marine National Monument, constitutes a Federal action, which makes it

1-7 MIDWAY SEABIRD PROTECTION PROJECT FINAL ENVIRONMENTAL ASSESSMENT PURPOSE AND NEED

subject to review under the National Environmental Protection Act (NEPA) and other applicable statutes, regulations, and . USFWS is required to integrate and consider the potential environmental effects that its actions may have on the human and natural environment prior to taking action. It accomplishes this by evaluating the environmental consequences of proposals to ensure that environmental values are given appropriate consideration in agency decision-making along with economic and technical factors within the agency’s mission. This EA has been prepared in consultation with other agencies, private organizations, and the public. If, after circulating the report for public and agency comment, USFWS finds that the proposed project would not have a significant adverse effect on the quality of the environment, it would prepare a Finding of No Significant Impact (FONSI). Notification of the EA and FONSI will be available at: https://www.fws.gov/refuge/midway_atoll/ If at any point in the preparation of an EA, USFWS determines that the proposal would have a significant adverse effect on the quality of the environment, it would initiate preparation of an Environmental Impact Statement (EIS). In addition to NEPA, other applicable laws and regulations include Section 7 of the Endangered Species Act (ESA) of 1973, Essential Fish Habitat under Magnuson-Stevens Conservation and Management Act and the Migratory Bird Treaty Act (MBTA) of 1918. Specific Federal laws that apply in addressing alien species and invasive species in the NWHI, in addition to the ESA, include the Lacey Act of 1900, as amended (18 USC 42, 16 USC 3371), and the National Invasive Species Act (NISA) of 1996. This EA is intended to address and satisfy many, but not all, of the above requirements. A full list of compliance needs for this project can be found in Chapter 8. The Magnuson-Stevens Act mandates that Federal Agencies conduct an Essential Fish Habitat (EFH) consultation with NOAA Fisheries regarding any actions authorized, funded, or undertaken that may adversely affect EFH. This EA serves as a basis for EFH impact review and consultation for the proposed action to eradicate mice at MANWR. The USFWS’s resource managers are authorized, and mandated by law, to conserve and restore wildlife and habitats that are under the agency’s jurisdiction. With a mission to work collaboratively with others to “conserve, protect, and enhance fish, wildlife, and plants and their habitats for the continuing benefit of the American people,” it is the responsibility of USFWS to address any threats that introduced species pose to the habitat and native wildlife of MANWR. Eliminating invasive mice from MANWR’s Sand Island is pertinent to the management strategy of USFWS for the refuge. The proposed action would be carried out in compliance with various Federal laws, EOs, and legislative acts including: • National Wildife Refuge System Improvement Act of 1997 (Public law 105-57); • The Fish and Wildlife Act of 1956 (16 U.S.C. 742a-742j, not including 742 d-l, 70 Stat. 1119), as amended; • National Environmental Policy Act of 1969 (NEPA); • Endangered Species Act (ESA) of 1973 (16 U.S.C. 1531-1544, 87 Stat. 884);

1-8 FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT PURPOSE AND NEED

• Migratory Bird Treaty Act of 1918 (MBTA)2; • EO 13186 Responsibilities of Federal Agencies to Protect Migratory Birds (66 FR 3853, Jan. 17, 2001); • Marine Mammal Protection Act of 1972 (MMPA), as amended; • Magnuson-Stevens Fishery Conservation and Management Act and Essential Fish Habitat; • EO 13089 Coral Reef Protection (June 11, 1998); • Water Pollution Control Act of 1948, as amended; • EO 13112 Invasive Species as amended 12/08/2016 by E0 13751; • National Historic Preservation Act of 1966 (NHPA); • Federal Insecticide, Fungicide, and Rodenticide Act of 1947 (FIFRA); and • Papahānaumokuākea Marine National Monument 50 CFR Part 404 – Permit Requirement.

1.5 PROJECT PARTICIPANTS

1.5.1 U.S. FISH AND WILDLIFE SERVICE (USFWS) - LEAD AGENCY The U.S. Fish and Wildlife Service is a bureau of the U.S. Department of Interior that works with others to conserve, protect, and enhance fish, wildlife, and plants and their habitats for the continuing benefit of the American people. The National Wildlife Refuge System within the USFWS has primary jurisdiction and statutory authority for management of MANWR resources, thus it is the Lead agency for this Seabird Protection Project. MANWR is part of the Papahānaumokuākea Marine National Monument, the largest contiguous fully protected conservation area under the U.S. flag, and one of the largest marine conservation areas in the world. It encompasses 582,578 square miles (mi2) (1,508,870 square kilometers [km2]) of the Pacific Ocean, an area larger than all the country's national parks combined. The Northwestern Hawaiian Islands Marine National Monument was established by Presidential Proclamation 8031 on June 15, 2006 under the authority of the Antiquities Act (16 U.S.C. §431- 433). A year later, it was given its Hawaiian name, Papahānaumokuākea under Presidential Proclamation 8112. The Monument was expanded by Presidential Proclamation 9478 on August 26, 2016. It was expressly created to protect an exceptional array of natural and cultural resources. The stated mission of PMNM is to carry out seamless integrated management of the monument to ensure ecological integrity and achieve strong, long-term protection and perpetuation of NWHI ecosystems, Native Hawaiian culture, and heritage resources for current and future generations.

2 The Department of the Interior Solicitor's Office issued a binding opinion (M-Opinion) on December 22, 2017 (Memorandum M-37050) which states that a permit is not required for incidental take of migratory birds. Therefore, an MBTA permit will not be sought. However, this document does analyze the effects to migratory birds and avoidance and minimization measures that will be implemented as part of the USFWS' responsibility to conserve the trust resource of migratory birds (See 1.4.5)

1-9 MIDWAY SEABIRD PROTECTION PROJECT FINAL ENVIRONMENTAL ASSESSMENT PURPOSE AND NEED

1.5.2 U.S. DEPARTMENT OF AGRICULTURE, ANIMAL AND PLANT HEALTH INSPECTION SERVICE, WILDLIFE SERVICES (WS) AND THE NATIONAL WILDLIFE RESEARCH CENTER (NWRC) - COOPERATING AGENCY The mission of USDA APHIS Wildlife Services (WS) is to provide Federal leadership and expertise to resolve wildlife conflicts to allow people and wildlife to coexist. WS conducts program delivery, research, and other activities through its Regional and State Offices, the NWRC and NWRC Field Stations, as well as through its National Programs. The NWRC is the federal institution devoted to resolving problems caused by the interaction of wild animals and society. The NWRC applies scientific expertise to the development of practical methods to resolve these problems and to maintain the quality of the environments shared with wildlife. NWRC has special expertise with respect to this proposed project and has agreed to be a Cooperating Agency. NWRC actions in this project include: • NWRC might be considered for the chemical analysis of pre- and post-application brodifacoum residues in rodents or non-target animals. NWRC’s involvement would be limited to receiving samples from Midway project managers, conducting laboratory- based chemical analysis, and disposing of biological and chemical waste resulting from the analytical process in the NWRC waste management program. • The NWRC Registration Unit will, with direction from the Midway project planners, conduct all activities associated with preparation of a specific label for the APHIS rodenticide product Brodifacoum 25D Conservation (EPA Reg. No. 56228-37) used for the mouse eradication attempt, submission of the supplemental registration application to the Environmental Protection Agency and all communication required to obtain a registration approval.

1.5.3 MEMORANDUM OF UNDERSTANDING (MOU) BETWEEN ISLAND CONSERVATION, USDA (APHIS), AMERICAN BIRD CONSERVANCY AND THE USFWS In April 2015, the USFWS entered into agreement with Island Conservation for the “purpose of furthering wildlife conservation and ecosystem management interests and responsibilities for the islands, atolls, and reefs under the jurisdiction of the United States.” This partnership promotes an integrated and coordinated approach to protecting, restoring and managing native populations of plants and animals and island ecosystems impacted by invasive alien species including but not limited to rodents, , cats and plants. In 2016, 2 additional parties were added to this national level agreement, including APHIS and the American Bird Conservancy.

1.5.4 ISLAND CONSERVATION Island Conservation (IC) is an active partner in this Seabird Protection Project under the MOU. IC is a non-profit organization with the mission “to prevent extinctions by removing invasive species from islands.” Working in partnership with local communities, government management agencies, and conservation organizations, Island Conservation seeks to identify islands that have the greatest potential for preventing extinction of globally threatened species classified as Critically Endangered or Endangered (i.e., CR or EN) on the IUCN’s Red List of Threatened Species. Island Conservation develops plans for, and implements the removal of, invasive alien

1-10 FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT PURPOSE AND NEED

species and conducts field research to document the environmental effects of this work to inform the planning and execution of future conservation projects. Island Conservation seeks to prevent extinctions by focusing its efforts on islands where the concentration of biodiversity, and consequently the threat of species extinction, is greatest. Of the 245 recorded animal extinctions since 1500 AD, 80% were on islands. Of those, in cases where causes could be identified, non-native and invasive species were responsible for 54% of these extinctions. Eradicating the primary threat—introduced invasive vertebrates—is one of the most critical and effective means for preserving threatened populations of plants and animals and restoring island ecosystems. To date, Island Conservation has deployed teams to protect 994 populations of 389 species on 52 islands.

1.5.5 MIDWAY RESTORATION PARTNERSHIP For the Midway project, and consistent with the National MOU, Island Conservation formed a partnership team with a variety of regulatory agencies and conservation organizations to provide individual input and project support in the planning phase of the mouse eradication. Participants in the Midway Restoration Partnership include the USFWS; USDA, APHIS WS and NWRC; NOAA; U.S. Geological Survey; Office of Hawaiian Affairs; and the American Bird Conservancy.

1.6 ORGANIZATION OF THE EA The remainder of this document is organized as follows: • Chapter 2 describes the proposed action and its alternatives in detail, including those alternatives that were initially considered but ultimately rejected from further evaluation. • Chapter 3 provides information regarding the existing environment and identifies the kinds of environmental effects which each alternative is likely to have. • Chapter 4 discusses the potential for cumulative effects. • Chapter 5 lists the agencies, organizations, and individuals who prepared and contributed to this report. • Chapter 6 provides the references for the information sources that were relied upon during preparation of this report. • Chapter 7 describes the scoping process which was implemented during the planning process. • Chapter 8 includes a statement of the project’s consistency and compliance with relevant laws, EOs, agency policies and permits. • Chapter 9 provides a glossary of terms.

1-11 MIDWAY SEABIRD PROTECTION PROJECT FINAL ENVIRONMENTAL ASSESSMENT PURPOSE AND NEED

THIS PAGE INTENTIONALLY LEFT BLANK

1-12 FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT DESCRIPTION OF THE PROJECT AND ALTERNATIVES

DESCRIPTION OF THE PROJECT AND ALTERNATIVES This chapter provides detailed information about the Proposed Action and alternatives to the Proposed Action, including those alternatives that were initially considered but ultimately rejected from further evaluation because they did not meet the project objectives outlined in Section 1.2. It also describes the procedures that would be used in the implementation of the project, methods, estimated costs, and the anticipated schedule for project implementation. The information contained in this chapter draws from the Feasibility Study: House Mouse Eradication on MANWR (Island Conservation 2017), including the discussion of methods and potential alternatives. Sections 2.1 and 2.2 provide principles and considerations that all alternatives must address to fulfill the project’s objectives. Section 2.3 presents the details of the Proposed Action; Section 2.4 outlines the No Action Alternative; and Section 2.5 discusses the alternatives that have been considered but rejected.

2.1 PRINCIPLES OF RODENT ERADICATION IN ISLAND ECOSYSTEMS Rodent eradication has become a common tool for agencies charged with the maintenance and management of island ecosystems, and its use to protect and conserve threatened species and ecosystems has increased significantly over the past decade. By 2015, there were 944 documented attempts at rodent eradication—across 10 different rodent species—on 692 islands globally (Samaniego-Herrera 2016). In total, 87 (9.2%) of those attempts were specifically related to mouse eradication; these attempts were reported from 76 islands in 17 countries (Samaniego 2016). Most of these mouse eradication attempts were in the temperate region, with only 32 of the 87 (36%) of them being mouse eradication attempts on tropical islands. Island size does not appear to be a limiting factor for successful eradication. Eradication attempts and successful eradication operations have been reported from small offshore islets to very large islands, including the largest known mouse eradication on Macquarie Island at approximately 31,810 acres (12,873 hectares). Mouse eradications are currently being planned or considered for very large islands such as Floreana Island in the Galapagos archipelago (~44,478 ac or 18,000 ha) and Marion Island in South Africa (~82,780 ac or ~33,500 ha) (Parkes 2014). Figure 2.1 below summarizes the outcomes of attempts globally at mouse eradication on islands over time and by island size.

2-1 MIDWAY SEABIRD PROTECTION PROJECT FINAL ENVIRONMENTAL ASSESSMENT DESCRIPTION OF THE PROJECT AND ALTERNATIVES

Figure 2.1 Global Mouse Eradication Attempts on Islands Over Time and by Island Size

Source: DIISE 2016; Samaniego 2016 cited in Island Conservation (2017)

While no guarantee of success exists, as evidenced by Figure 2.1, mouse eradication from islands is achievable provided that basic principles of rodent eradication are applied. Attempted mouse eradication from all islands between 1975 and 2015 had an overall success rate of 71%. In comparison, success of rat eradication programs is approximately 87% (Russell and Holmes 2015). The reason(s) behind the lower success rate of mouse eradications is undetermined and has been the subject of several investigations (Samaniego 2016, Mackay 2011). Factors contributing to failure could include some, or a combination of, the following factors: (i) natural history differences between rats and mice could make it harder to target the complete population of mice, including their foraging behavior, travel distances, and home range sizes; (ii) the fact that mice have been a secondary target of multi-species eradication efforts; or (iii) ineffective eradication strategies or management approaches that failed to reach all mice, so that remnant mice remain and recolonize the island. More recently, 6 of 10 mouse eradication projects on islands greater than 988 ac (400 ha) have been successful. All of these 10 projects were conducted as multi-species eradications, typically known to be more challenging than single-species eradications; and were in temperate regions. Within the , the largest mouse eradication attempt to-date has been on Mer Island (~1,134 ac or 459 ha) and is still being evaluated for success. Of the 3 largest mouse eradication attempts where a positive or negative outcome has been determined (i.e., success or failure), 2 were successful (Samaniego 2016). Between 2005 and 2015, the success of mouse eradications increased to 93.3% of 31 attempted eradications. This increase in the success rate has been attributed to better international cooperation

2-2 FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT DESCRIPTION OF THE PROJECT AND ALTERNATIVES

and knowledge sharing between eradication practitioners regarding lessons learned (Veitch et al. 2002, 2011), and the establishment of best practices principles for the eradication of rodents in temperate (Broome et al. 2014) and tropical ecosystems (Keitt et al. 2015). Today, successful rodent eradications on all but the smallest of islands (i.e., less than ~12 ac or 5 ha) rely on the use of rodenticides, and specifically, anticoagulant rodenticides (Database of Island Invasive Species Eradications [DIISE] 2016). Mice were successfully eradicated from Antipodes Island in the Subantarctic in 2018. The 5,189 ac (2,100 ha) island had a large mouse population of about 200,000 individuals. House mice were known to eat invertebrates, prey on bird chicks and eggs, and eat plant material including seeds. Mice had already wiped out two taxa of insects from Antipodes Island and Black- bellied storm petrels (fregetta tropica) and Subantarctic little ( elegans) only bred on the mouse free offshore islands. All successful eradications, regardless of species or location, are based on 3 fundamental principles (Cromarty et al. 2002): 1. Every individual of the target species must be put at risk with the proposed eradication technique(s). 2. The technique(s) employed must remove individuals from the target population at a rate greater than they can breed (i.e., their replacement rate). 3. Immigration of new specimens of the target species must be zero, or effectively managed to zero, by identifying and eliminating immigrant specimens. These principles have been further developed with specific regard to rat and mouse eradications as summarized in Table 2.1 below (see Howald et al. 2007).

Table 2.1 Principles of Successful Rat and Mouse Eradications No. Principle 1 Deliver a highly palatable bait containing a toxic rodenticide into every potential rodent territory. 2 Ensure bait is available in enough quantity and for long enough that every mouse has access to a lethal dose. 3 Time the baiting operation to when the rodent population is most likely to consume the bait. 4 The short-term risks and impacts to nontarget wildlife, people, and the environment from disturbance and the rodenticide is minimized wherever possible. The benefits of eradication must outweigh the costs. 5 Biosecurity procedures must be able to sustain the eradication, with effective prevention, detection, and an effective response to any incursion. Source: Howald et al. 2007 cited in Island Conservation (2017).

The principles outlined in Table 2.1 have been further refined into best practice guidelines for maximizing the probability of successful rodent eradications from temperate islands in New Zealand (Broome et al. 2014) and from tropical islands (Keitt et al. 2015).

2.2 MANWR TECHNICAL CONSIDERATIONS Island Conservation, in its 2017 Feasibility Report, acknowledges that, “the feasibility of an eradication is a multi-dimensional analysis that considers technical, environmental, social,

2-3 MIDWAY SEABIRD PROTECTION PROJECT FINAL ENVIRONMENTAL ASSESSMENT DESCRIPTION OF THE PROJECT AND ALTERNATIVES political, and legal factors.” These factors must all be considered, both individually and cumulatively, to determine if, and how, a successful mouse eradication can be achieved at Sand Island, MANWR.

2.2.1 CONSTRAINTS ON SUCCESSFUL ERADICATION While all rodent eradications from large islands apply the same fundamental set of principles (see Table 2.1 above), the technical approach to each project must be tailored to the unique environmental context, including its wildlife communities, ecosystem function, and landscape features. Because of this, the first step in evaluating the best technical approach consists of identifying the local constraints acting on the operation and how they impact the overall probability of success, as well as the potential for concomitant adverse impacts to the local environment. These constraints have the potential to limit the likelihood of a successful eradication if they are not understood, minimized, mitigated, or eliminated. The major constraints present on Sand Island, MANWR which must be considered in project design include the presence of manmade structures, underground utilities, native species that inhabit the island, and the resident community. These are discussed in the following sections.

2.2.1.1 Manmade Structures The built environment on Sand Island is extensive and is the result of a long history of human habitation and infrastructure development over more than a century. The structures on Sand Island vary in age, construction , condition, and whether and how they are used. The entire built environment on Sand Island can be divided into one of 3 categories, summarized in Table 2.2 below.

2-4 FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT DESCRIPTION OF THE PROJECT AND ALTERNATIVES

Table 2.2 Manmade Structures on Sand Island, MANWR Construction Category Use Description Vertical Inhabited Living Spaces Buildings that provide housing and dining facilities, food storage, and food preparation areas. Work Spaces Buildings that provide offices, utility buildings, covered storage, recreation facilities, covered garden, and nurseries. Abandoned Condemned Structures and spaces where human access is restricted or no Spaces longer permitted due to physical or other safety reasons. Horizontal In-use Airfield An active 7,897-foot (2,407 m) long and 151-foot (46 m) wide airfield, certified by the Federal Aviation Administration (FAA).3 The runway and taxiway is maintained as an emergency landing site for extended twin-engine operations (ETOPS) across the Pacific Ocean. Subterranean (both in use Utilities, High-voltage service boxes. and abandon) Access Points Electrical conduit junction sites. Military Sewer system access points. Structures Waterline junction sites and valves. Abandoned junction points for water, oil, and electrical services. Source: Island Conservation (2017).

All categories comprising the built environment on Sand Island present unique challenges to a successful eradication effort. These 3 categories of infrastructure each require specific strategies to ensure that rodenticide is applied to every potential mouse territory.

2.2.1.2 Native Species The presence of wildlife on and around islands targeted for eradication efforts present an inherent challenge. In essence, the operation must be able to deliver bait to every mouse on the island while minimizing the availability of or the exposure to the rodenticide to other non-target species susceptible to it. In addition to the toxicological risks, the operation may also impose disturbances and habitat alterations that could have negative impacts on the ecosystem as a whole. Although impacts to native species are only ecologically significant if they pose a population-level threat, in principle all risks to wildlife should be avoided, minimized, or mitigated whenever reasonably practicable. As noted in the discussion of principles of successful rodent eradication, the benefits of the effort must outweigh the environmental costs and mitigation strategies must be considered in the tradeoff framework (Broome et al. 2014). MANWR is of great significance (see Section 3.3.6) to several terrestrial and marine species, including: • The critically endangered, non-migratory resident Laysan duck.

3 In 2004, FAA issued a final rule that revised the Federal airport certification regulation [Title 14, Code of Federal Regulations (CFR), Part 139 (14 CFR Part 139)] and established certification requirements for airports serving scheduled air carrier operations in aircraft designed for more than 9 passenger seats but less than 31 passenger seats. In addition, this final rule amended a section of an air carrier operation regulation (14 CFR Part 121) so it would conform with changes to airport certification requirements. The revised Federal airport certification requirements went into effect on June 9, 2004.

2-5 MIDWAY SEABIRD PROTECTION PROJECT FINAL ENVIRONMENTAL ASSESSMENT DESCRIPTION OF THE PROJECT AND ALTERNATIVES

• The endangered short-tailed albatross (Phoebastria albatrus). • High-density breeding populations of Laysan albatross and black-footed albatross. • Threatened green sea turtles (Chelonia mydas) use the beach for basking and nesting; endangered hawksbill sea turtles (Eretmochelys imbricate) may also be present in MANWR but have not been confirmed to bask or nest there. • Endangered Hawaiian monk seals (Neomonachus schauinslandi) that use the beaches and fore-shore vegetation for resting and pupping.

2.2.1.3 Resident Community on Sand Island Successful island-based rodent eradications have been, for the most part, conducted on remote, uninhabited islands free from permanent human settlements or development. Human settlements, inadvertently, provide food and shelter to rodents. This situation is referred to as “commensal” in that it is used by both humans and rodents; it is a relationship in which the rodents benefit from the presence of the humans while the humans can be largely unaffected by the presence of rodents. This commensal behavior can represent a significant risk to an eradication attempt, decreasing the probability of success. There is precedent to rodent eradications from human inhabited islands—a total of 94 documented cases—but these operations require supplemental treatment strategies such as targeted hand application of bait, mechanical devices (e.g., traps), and modification of bait strategies (DIISE 2016). The infrastructure of Sand Island would be no exception. Local communities are frequently key-stakeholders in island eradication projects and must be considered from the outset where an eradication effort takes place on an inhabited island. The year- round community on MANWR numbers roughly 57 people and includes: (i) USFWS staff and volunteers; (ii) base operations staff from Chugach Alaska Corp.; and (iii) transient contractors or researchers. The community would bear the direct or indirect negative impacts of invasive rodent infestations and stand to benefit from the outcomes of eradication and subsequent restoration efforts. Moreover, these local communities play a significant role in the eradication process and maintaining the rodent-free status of Sand Island, which frequently requires changes in behavior. In brief, their involvement is essential to the success of all effective eradication projects.

2.2.2 BAIT DISTRIBUTION APPROACHES The only viable approach to completely remove the house mouse from Sand Island, MANWR is the use of rodenticide(s). Other tools and strategies used to control mice were considered but rejected because there is no evidence that they would have a reasonable probability of eradicating mice on MANWR. See Section 2.5 for a more detailed discussion of approaches which were considered but ultimately rejected from further evaluation. Use of rodenticide requires the delivery of rodenticide-impregnated bait into every potential mouse territory. In order to accomplish this, there are 3 basic methods, and a fourth hybrid method: (i) bait stations; (ii) hand broadcast; (iii) aerial broadcast using a helicopter and bait-sowing bucket; and (iv) a combination of the above methods. Each of these methods is described in greater detail in the following subsections.

2-6 FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT DESCRIPTION OF THE PROJECT AND ALTERNATIVES

2.2.2.1 Bait Stations Bait stations have been used in rodent eradication projects on islands ranging from very small to relatively large (DIISE 2016). Bait stations typically used in island eradications have been either commercially manufactured plastic bait boxes, or plastic pipes, or other similar material. Either design allows rodents to enter the stations freely to access bait placed within it, but prevent larger, non-target species from easily entering and feeding on the bait and greatly extend the longevity of the bait. The major advantage of bait stations is that the bait can be delivered to all mice while preventing most non-target species from gaining access to the station or the poisoned bait within. For eradication purposes, bait stations are systematically placed on a grid pattern in all habitats across the entire island. Once placed, bait crews would arm and check stations regularly and re- arm each station over a period of months until bait take by rodents declines to zero. Bait stations were previously and successfully used at MANWR in the 1990s to eradicate R. rattus. In that effort, bait stations were spaced at ~164 ft. (50 m) intervals with live traps in between, ensuring that at least 2 stations were found in every potential rat home range. Due to smaller territory size when targeting mice, bait stations would need to be at smaller intervals.

2.2.2.2 Broadcast In the 1990s, island managers began to explore more efficient techniques to undertake eradications more quickly and on larger islands and adopted methods of broadcasting rodenticide impregnated bait using both hand broadcast and aerial broadcast (see additional discussion below). Bait, in the form of cylindrical or spherical pellets, is broadcast evenly across the landscape at a prescribed application rate calculated as kilograms per hectare or pounds per acre (i.e., kg/ha or lbs./ac.) so that all rodents have access to it for long enough to find the bait, overcome any neophobia, and consume a lethal dose. A strategy which broadcasts bait across the island, whether by hand, helicopter, or both can achieve success in a significantly shorter timeframe than with bait stations as it exposes all target individuals at one time and avoids the risk of neophobia related to animals entering bait stations.

2.2.2.2.1 Hand Broadcast Broadcast of bait by hand is the second most-documented approach to island rodent eradication efforts (DIISE 2016). Bait is typically distributed by a team of baiters who systematically walk on parallel transects, stopping at predetermined intervals to distribute pellets as evenly as possible in a quadrant or circle.4 As the end of one transect is approached, the teams move over to the next transects and start anew, treating the entire area of the island in a “rolling front”. For successful hand broadcast of bait, significant advance preparation is required to ensure efficient application and to minimize delays and errors in bait application. Complexity increases with the size of the land area and the topography of the island being treated. Transects through heavy vegetation would be opened using mechanical or hand tools to create a path to all baiting points in advance, and each broadcast point would be marked using flagging or pins and recorded

4 See: http://rce.pacificinvasivesinitiative.org

2-7 MIDWAY SEABIRD PROTECTION PROJECT FINAL ENVIRONMENTAL ASSESSMENT DESCRIPTION OF THE PROJECT AND ALTERNATIVES by global positioning system (GPS) technology. Bait must also be staged at various accessible locations across the island to minimize the time and effort needed for baiters to refresh their supplies as they progress across the treatment transects.

2.2.2.2.2 Aerial Broadcast Aerial broadcast for rodent eradication projects involves using a commercial-grade bait bucket slung under a helicopter, guided by GPS to evenly distribute bait across the entire area of the island. The set rate at which the bait exits the bucket, the width of the treatment swath, and speed are calibrated to achieve a desired application rate. The pilot is guided by a computer connected to a GPS and guidance system to keep the helicopter on pre-programmed bait application flight lines. The bait flow from the bucket is controlled by the pilot at all times, opening and closing the bait bucket on demand to apply bait in desired areas and minimize bait application in other areas, such as the marine environment. The hopper can be fitted with a deflector to broadcast bait out to one side, allowing the helicopter to fly parallel to and along the shoreline with minimal unintentional bait applications. Bait application along the shoreline can be accomplished with minimal bait drift into the marine environment with the use of a deflector. The infrastructure on MANWR, provides a highly suitable base from which to implement an aerial broadcast operation. This includes support equipment for loading bait and large operational areas for loading and refueling. The airfield services available there include: (i) equipment staging; (ii) aircraft storage; (iii) fuel supply; and (iv) necessary fire, medical, and support infrastructure.

2.2.2.3 Summary Table 2.3 below summarizes the relative advantages and disadvantages of the 3 methods of bait application discussed in the preceding sections.

2-8 FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT DESCRIPTION OF THE PROJECT AND ALTERNATIVES

Table 2.3 Summary of Technical Considerations Method Advantages Disadvantages Bait • Requires least total bait. • Rodents are often neophobic and take time to enter bait stations. Stations • Controlled bait delivery. • Difficult to ensure all rodents have access to bait. • Lower mobilization of • Labor intensive and significant set up time. bait into ecosystem. • Presents logistical challenges and safety risks related to managing a • Easier removal of large team conducting repetitive and demanding physical labor residual bait when and accessing all parts of the island. effort concluded. • Relatively slow, with total eradication requiring a period of • Lower risk of primary between several months to several years. exposure to non-target • Non-target species are exposed to secondary risks for a prolonged species. period. • Limits requirement for • Requires a very large number of bait stations and access trails to specialized skills. bait stations with prolonged impact to vegetation. • Prolongs potential for direct impacts to ground-nesting seabirds and seabird burrows and repeated disturbance to monk seals and turtles. Broadcast • High likelihood that all • Requires more bait. in general rodents have access to • Bait is readily available to non-target species. bait. • Minimizes any neophobic behavior towards bait delivery devices • Shorter period of bait availabillity. Hand • Allows greater control • Labor intensive. Broadcast over bait application. • Presents logistical challenges and safety risks related to managing a specific • Limits requirement for large team conducting repetitive and demanding physical labor. specialized skills. • Requires transect paths along with access to all areas with dense vegetation and fragile areas. It is not possible to bait entire island in a single day. Precision needed is not achievable. • Potential for direct impact to ground-nesting seabirds and seabird burrows and repeated disturbance to monk seals and turtles. Aerial • Fastest method for • Requires specialized skills and equipment. Broadcast applying bait to a large • Depending on the label, may or may not be restricted over specific area. inhabited areas. • Relies on a smaller team • Distribution is limited to exposed areas and does not allow of workers. distribution of bait within buildings, under structures or • No need to clear trails. underground. • May result in some unintentional bait entering the ocean. Source: Island Conservation (2017)

2.2.3 RODENTICIDE AND BAIT CONSIDERATIONS Several factors must be considered in the selection of a bait product and associated toxicant for rodent eradication. Under ideal conditions, the bait product would have demonstrated 100% efficacy under both laboratory and field trials on the target island. In addition, it would have been

2-9 MIDWAY SEABIRD PROTECTION PROJECT FINAL ENVIRONMENTAL ASSESSMENT DESCRIPTION OF THE PROJECT AND ALTERNATIVES successfully used to eradicate the target species (i.e., house mice) on other islands. It would also have low risk to non-target species. From an eradication perspective, the rodenticide used on Sand Island, MANWR must: • Contain an active ingredient that is known to be highly toxic to mice; • Be palatable and demonstrated to cause little or no aversion in mice; • Be deliverable to every potential mouse territory on the island; • Be legally registered for use in compliance with the Federal Insecticide, Fungicide and Rodenticide Act (FIRFA); • Be of low attractiveness to non-target species. Presently there are 10 different rodenticides approved for use and available on the market in the United States. However, only 3 anticoagulant rodenticide products are currently registered for conservation use and therefore available for the eradication of rodents from island ecosystems; they are characterized in Table 2.4 below.

Table 2.4 Rodenticides Approved for Conservation Use in the United States Name Description Diphacinone D50 A 0.07 to 0.17 ounces (oz) (2-5 g) pelleted bait product containing 50 parts per million (ppm) diphacinone, adopted for use primarily in Hawaiʻi for main island conservation, landscape control, and eradication from offshore islets. It is made by Hacco, Inc. and Ditrac 50D by Bell Labs Brodifacoum-25D A 0.03 to 0.07 oz (1-2 g) pelleted bait product containing 25 ppm brodifacoum, designed for use on islands with Mediterranean climates. It is produced by Bell Labratories, Inc. Brodifacoum-25W A 0.07 oz (2 g) pelleted bait containing 25 ppm brodifacoum, designed for use on islands with wet to very wet climates. It is produced by Bell Laboratories, Inc. Source: Island Conservation (2017)

The major advantage of anticoagulant rodenticides is the delayed onset of symptoms, usually requiring between 2 and 3 days to take effect. They are not known to cause aversion, referred to as “bait shyness,” as per acute toxicants which rodents may learn to associate with their sublethal effects, and rodents would continue to feed on the bait even after symptoms develop (Parkes et al. 2011). Some rodents are even known to continue to feed on bait during the latent period, the time between ingestion of a lethal dose and mortality (Howald et al. 2005, Buckelew et al. 2005). First generation anticoagulant rodenticides, such as diphacinone (see Table 2.4), are multi-feed rodenticides and rodents need to consume bait over a sustained period of days to achieve mortality, which may vary from 3 to 12 days of ingestion for a lethal dose. Diphacinone has been successfully used to eradicate rats from islands, typically delivered by bait stations with relatively few instances of broadcast being the delivery method (DIISE 2016). It has been used in efforts to eradicate house mice on islands twice, with one instance being a reported success and one a failure (Samaniego 2016). Second generation anticoagulant rodenticides, such as brodifacoum, can be toxic after a single exposure or feeding event to sensitive rodents (Kaukeinen 1993). In some laboratory trials of brodifacoum, 100% mortality of rats and mice have been reported after a 3-day choice test (Pitt et

2-10 FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT DESCRIPTION OF THE PROJECT AND ALTERNATIVES

al. 2011). Globally, a total of 87 house mouse eradication attempts have been carried out, and all but the 2 mentioned above using diphacinone, utilized brodifacoum.

2.3 PREFERRED ALTERNATIVE: MOUSE ERADICATION WITH RODENTICIDE VIA MULTIPLE DISTRIBUTION METHODS The Preferred Alternative consists broadly of: • The use of Brodifacoum-25D rodenticide with supplemental trapping. • Multiple bait distribution methods – primarily aerial broadcast but also supplemental hand broadcast and bait stations (Section 2.3.1). • Three bait applications spaced roughly 7-12 days apart (Section 2.3.2), projected to start in early July (Section 2.3.5). • Implementing mitigation measures identified for selected species in Chapter 3. • Implementing effectiveness monitoring to assure mitigation is carried out and determine if mitigation is producing the desired results. • Employing long-term biosecurity protocols to prevent the reinvasion of the island by mice and rats. See Appendix A for a summary of long-term biosecurity protocols developed for MANWR. Implementation of the proposed action is currently being considered for July/August 2019 with mitigation operations potentially extending into 2020.

2.3.1 BAIT AND ITS DISTRIBUTION METHODS The toxicant to be employed as part of the Preferred Alternative would be Brodifacoum-25D Conservation, a pelleted rodenticide bait intended for conservation purposes for the control or eradication of invasive rodents on islands or vessels. Island Conservation’s Feasibility Report (2017) specifically recommends the use of a bait containing brodifacoum and notes that use of a less robust alternative rodenticide could compromise the success of the eradication effort. The rodenticide bait used for Sand Island would be dyed a green color to reduce the likelihood of birds picking up the bait. The Preferred Alternative consists of the use of brodifacoum to eradicate all mice on Sand Island, MANWR via 1 primary method and 2 supplemental distribution methods. The 3 methods (Island Conservation 2017) are discussed in the following 3 subsections.

2.3.1.1 Primary Distribution Method – Aerial Broadcast Aerial broadcast (Section 2.2.2.2.2) of a pelleted bait containing rodenticide across the entire emergent land area, except: • The portion of the airfield within the Foreign Object Debris (FOD) management area, which is delineated by large, conspicuous painted aircraft control lines on the tarmac.

2-11 MIDWAY SEABIRD PROTECTION PROJECT FINAL ENVIRONMENTAL ASSESSMENT DESCRIPTION OF THE PROJECT AND ALTERNATIVES

• Portions of the coastal fringe where metal retaining structures combine with eroding sand to create undercut areas, and piers. • Within a buffer zone around the fresh water ponds that cannot be covered by tarps or cannot be protected by other means. • Indoor residential/commensal areas; aerially baiting over all buildings will occur. Bait would be applied using bait stations or by hand broadcast and along the narrow sea wall on the harbor entrance. Aerial broadcast would also likely occur over rock piles on the island. These piles are rock staged for shore stabilization. There are also rocks that are used to stabilize the shoreline in some areas and these would also be aerial broadcast with a helicopter equipped with a deflector to maximize the chances of eradication success while minimizing the chances of any bait entering the marine environment. Aerial broadcast is the most widely-used method of delivery of pelletized bait on islands and is effective in delivering bait into every potential open above ground mouse territory. The helicopter broadcasting the rodenticide would fly at speeds ranging from 29-58 mph (46 - 93 km/hr.) and at an average expected altitude of approximately 164 ft. (50 m) above the ground. The bait hopper would be on a long-line 20-30 ft. (6 - 9 m) slung below the helicopter. Therefore, the bait hopper would be approximately 134 to 144 ft. (41 – 44 m) above the ground during the rodenticide drop. Except for landings and take-offs at the runway, the helicopter would fly at an average altitude of 164 ft. (50 m) and the bucket would be >134 ft. (>41 m) above the ground most of the time. The height of the helicopter would likely be higher when over forested areas or buildings. These numbers are an estimate and may change as needed to complete the baiting operation safely and effectively. The helicopter pilot would be certified for aerial bait application and in compliance with FAA safety requirements. Three bait drops are planned on Sand Island using 2 helicopters to maximize the chance of completing each drop in a single day, but a reduction to 1 helicopter is possible and the same effort would occur over multiple days. It is expected each drop would take a total of 20 flight hours or 60 total flight hours for the entire operation; total flight time could be longer. Check flights and boundary flights will also add to the total flight time. Bait would be delivered in shipping containers from the point of manufacture in the Mainland U.S. Any ships delivering equipment or bait to MANWR would have final inspections in Honolulu or their departure port before departing for MANWR. The shipping of bait and equipment needed for the eradication operation to MANWR are part of the Proposed Action (project) covered by this EA. The rodenticide bait would be stored in locked shipping containers. The shipping containers would protect the bait from exposure to the elements and would allow a controllable area to access bait during aerial broadcast operations. Bait would be packaged in either 50 lbs. (23 kg) bags, or in large (up to 700 lbs. or 318 kg) bags and loaded into large cardboard boxes on skids. The shipping containers would remain locked and staged in a large storage building adjacent to the airport runway on Sand Island and opened periodically for inspection prior to the eradication, and during the baiting operation. Bait would be loaded either manually by hand from bags into the hopper, or directly from bulk bags.

2-12 FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT DESCRIPTION OF THE PROJECT AND ALTERNATIVES

2.3.1.2 Secondary Distribution Method – Supplemental Hand Broadcast Supplemental hand broadcast of pelleted bait and the use of bait stations (Section 2.2.2.1) with rodenticide would be used in the following areas: • The locations listed above where aerial broadcast would not be conducted. • Key shoreline habitats (e.g., across causeways too narrow for aerial broadcast including a potential narrow strip of land between the coast and the airport runway (at the southeast end of the runway). • Piers will not be broadcast over. Bait stations or bait trays will be placed on these structures by hand. • Within a buffer zone around the fresh water ponds that cannot be covered by tarps or cannot be protected by other means. In commensal environments such as within aboveground and underground structures. Any work related to hand broadcasting will occur proximately to each of the aerial bait drops. For successful hand-broadcast of bait, preparation is required to ensure efficient application and to minimize delays and errors in bait application. Preparations for bait application include preparing the baiting grid and staging bait. In general, about one person-hour of preparation for every 2 person-hours of baiting would be needed for hand-broadcasting. The areas that would be designated for hand-broadcasting operations would not require any vegetation clearing or preparation of transects. The areas to be hand-broadcast would include: 1) an area buffer around the water catchment pond located between the runways; 2) a potential narrow strip of land between the coast and the airport runway (at the southeast end of the runway); 3) portions of a narrow sea wall that extends from the north end of the harbor southward and; 4) unoccupied commensal environments including baiting in aboveground and underground structures. For hand-broadcasting of bait around the water catchment pond, an estimated 16 person-hours would be needed to prepare the baiting grid and stage bait, and an additional 32 person-hours per application to apply bait per application. If hand-broadcasting occurs on the narrow strip of land at the southeast end of the runway, another 16 hours of prep and 32 hours of baiting per application would be needed. Some additional time may be spent hand broadcasting around the margins of covered seeps and ponds. Overall, for the entire operation, a maximum of 240 total person-hours is estimated to be needed for hand-broadcasting outside of commensal areas given 3 planned scheduled drops, 8-person teams, and 10-hour work days. Within the commensal environment, where people are typically not present, hand broadcast may be employed, along with the following supplemental bait station methods for distributing bait inside occupied structures: • Hand placement of loose bait trays, with an open top to expose bait. • Bait bolas, where the bait is held within rodent accessible material that can be hung in place to keep out of water. • Enclosed bait stations (either tubes or bait boxes with lids).

2-13 MIDWAY SEABIRD PROTECTION PROJECT FINAL ENVIRONMENTAL ASSESSMENT DESCRIPTION OF THE PROJECT AND ALTERNATIVES

Each structure, in use, abandoned, or condemned and above ground or below ground, would be individually identified and assigned a unique mouse removal strategy, known as a Structure Management Plan (SMP); all SMPs would outline a structure-specific mixture of secondary distribution methods (hand broadcast and/or bait stations) plus the possible use of traps. The following factors would be considered and recorded when developing each individual SMP: • Structure category and use; • Numerical identification; • GPS coordinates and GIS map; and • Details regarding how it is to be treated throughout the implementation period including detailed maps of bait application and placement. When available, photographs of the structure and a floor plan showing treatment options may be included.

2.3.2 NUMBER AND RATE OF BAIT APPLICATIONS Bait must be delivered at an application rate that ensures bait is made available to all mice on the island for long enough that each individual mouse would be exposed to and consume a lethal dose of the toxicant. The application rate needs to account for other species that compete with mice to gain access to the bait. Some competitor species, like invertebrates such as and insects, can consume a significant amount of the bait but are not affected by the rodenticide (Island Conservation 2017). Other avian competitor species, such as the myna, cattle egret, canary, shorebirds, and Laysan duck, could also consume bait and would be negatively affected by the exposure. Bait application rates are set to ensure that adequate amounts of bait are available for long enough, regardless of the number of species or individuals that consume it. Based on other eradication efforts, a minimum of 4 nights bait availability may be necessary (Keitt et al. 2015). The total application rate on Sand Island is expected to be up to 214 lbs/acre (239 kg/ha). Up to this amount of bait will be applied over the course of 2-3 applications, with approximately 18 to 28 days between the first and final application of bait. The specific baiting strategy will be designed to ensure bait is available to all mice for a sufficient period of time, including intercepting any new generation of mice that may have been missed or emerge an earlier bait application. All applications will be made in compliance with the EPA supplemental bait label. The bait would be applied according to a helicopter flight plan that would take into account the need to: (i) apply bait relatively evenly and to prevent any gaps in coverage; (ii) accommodate island topography; (iii) minimize bait spread into the marine environment; (iv) minimize disturbance to native wildlife and; (v) ensure human safety. The onboard computer linked to the GPS would serve as the primary method of monitoring where bait has been applied to the island. Data from the onboard computer would be downloaded from the computer and evaluated on a laptop computer to assess where bait has been applied and total area treated, in order to calculate the approximate actual bait application rate. A monitoring team would be staged on Sand Island to collect near-real time data during the bait application to ensure that the application rate stays within the legal and optimal application rates.

2-14 FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT DESCRIPTION OF THE PROJECT AND ALTERNATIVES

In addition, a monitoring plan would be implemented to determine if: (i) eradication of mice is progressing as expected (mice mortality), and (ii) potential risks and expected impacts of the rodenticide in the environment and to non- native species are documented and in compliance with applicable permits and guidelines (e.g., NEPA permits and labels). At a minimum, sampling of marine water, fish, birds, and rodents would be made. To implement the monitoring plan, a field team would be present on Sand Island for approximately 6 weeks.

2.3.3 MECHANICAL TRAPPING Mechanical traps including, but not restricted to, snap-traps and glue boards would be employed in conjunction with hand broadcast and/or bait stations in the commensal environment and other areas where bait is not aerially broadcast (such as piers), as deemed appropriate per the terms of each structure’s SMP (see Section 2.3.1). These traps would be used as an additional detection and removal method within structures to appropriately target mice. Traps would be activated and checked on the same schedule as bait stations within buildings, except live traps, which would be checked daily. Traps and other detection devices may be used to inform timelines for managing and deactivating bait stations in buildings.

2.3.4 PRE-APPLICATION PREPARATIONS The collection of water in the storage tanks would be maximized prior to the eradication effort. The water storage tanks, when filled to capacity, have the ability to provide water to the island for a 2-year period. Once the tanks are full, they would be disconnected from the water collection system and the intake-grates covered prior to bait application. MANWR must remain an active residential community before, during, and after the Midway Seabird Protection Project. Certain activities, such as maintaining the airfield’s operability, potential construction projects, biological monitoring, and maintenance of facilities and utilities would continue and the eradication effort must accommodate these activities. However, modifications to the day-to-day systems, processes, and infrastructure can, and in certain cases must, be made to minimize the risk of eradication failure. These may include: • Procedural changes to food delivery, handling, storage, and disposal. • Procedural changes to grow operations at the hydroponic gardens. • Procedural changes to food preparation and consumption. • Modifications to waste disposal and the on-island landfill. • Development of a biosecurity plan which would prevent reintroduction of mice and rats. See Appendix A for a summary of the biosecurity plan developed for MANWR. Food would be stored at all times in rodent proof containers. Some food comes in rodent proof containers (e.g. canned food), but food that may be more accessible (loose vegetables, bags of rice) would be put in containers that can be closed tight. Additional bait stations/traps may be deployed in areas where food is stored as well. At a minimum, each grow operation (e.g. tables, pots on ground) would be assessed and determined if it should be removed or modified to prevent mice from accessing food. Baits stations, bait trays, and/or traps would be placed in the shade house and

2-15 MIDWAY SEABIRD PROTECTION PROJECT FINAL ENVIRONMENTAL ASSESSMENT DESCRIPTION OF THE PROJECT AND ALTERNATIVES

modifications may be made to further prevent mice from accessing food. Any food growing operation would be closely monitored for signs of mouse activity and be responded to accordingly. For food preparation and consumption, personnel would be trained in proper procedures, which would likely include emphasis on not leaving food out and immediate clean up. Garbage would be stored in sealed containers at all phases (in buildings, outside, when awaiting incineration), and the garbage collection schedule may likely be modified to include more frequent collections and complete incineration. Additional discussion of human health and safety factors are provided in Section 3.3.6.21.3.

2.3.5 APPLICATION SCHEDULE In MANWR, the average period of lowest rainfall is from April through July, with increasing rainfall from late summer through winter. Reports from island biologists suggest that the condition of low vegetation such as grasses are deteriorated due to dry conditions and trampling by albatross chicks. Thus, the lowest availability of vegetation as a food source to mice, and likely low point in their annual population cycle, is probably between May and July. The low precipitation and low level of vegetation food sources make the summer a very attractive timeframe for an eradication effort employing bait broadcast. The role that nesting albatross play in providing alternative food sources to mice (i.e., dead chicks and adults, and regurgitated food), and the influence it may have on the annual mouse population cycle is unclear, but it may be relevant. To date, data suggests that the mice breed year-round, even though it is likely that the rate at which they breed fluctuates on an annual and intra-annual basis, depending on climate (Brown and Singleton 1999, Drost and Fellers 1991). The presence of potential nontarget species on Midway is another consideration in selecting the rodenticide application period. There is no point in the year when non-target species are not present on Sand Island; however, the late summer timeframe is considered better than other times because relatively few potential non-target species are typically present (see Section 3.3.6). Based on these considerations, the first rodenticide application would occur in early July. A second application would occur roughly 7-12 days later, and a third application conducted roughly 7-12 days after the second application. If a significant storm event (i.e., tropical storm) is forecast to occur within 7 days after the application of the rodenticide, then the application would be postponed until the threat of the storm has past. Other dynamic elements, such as the presence of bird airstrike hazards (BASH), high winds, and other factors may also require adjustments in the project implementation schedule. This measure, like many other considerations detailed in this Chapter, is included to avoid and minimize potential adverse environmental effects and increase the potential for success.

2.3.6 MITIGATION ACTIONS AND EFFECTIVENESS MONITORING Mitigation measures and actions identified for multiple species in Chapter 3 would be carried out and monitored for their effectiveness. Effectiveness monitoring tracks the success in achieving desired outcomes and evaluating environmental effects (see Appendix B). Mitigation includes specific measures or practices that would reduce, avoid or eliminate the effect of the proposed

2-16 FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT DESCRIPTION OF THE PROJECT AND ALTERNATIVES action on non-target species. In this EA, the identified mitigation measures are part of the Proposed Action (project) and necessary to support a FONSI. Examples of mitigation measures in Chapter 3 include: training ground-based staff to identify endangered plants and how to avoid stepping on burrows; capturing and moving vulnerable species to avoid rodenticide exposure; measures to minimize bait entering the marine environment; and measures to reduce impacts to sea turtles and monk seals. For detailed descriptions see the protection and mitigation section for each species in Chapter 3. The Proposed Action (project) also includes mitigation measures on Eastern Island where Laysan ducks would be cared for and protected. A detailed protection and mitigation plan for Laysan ducks and shorebirds is provided in Appendix C.

2.3.7 POST-APPLICATION ACTIONS Post-application actions include 3 critical tasks: • Post-treatment efficacy monitoring and post treatment environmental effects and residue monitoring for rodenticide in the terrestrial, freshwater aquatic, and marine environments and for documenting mortality of non-target organisms (see Appendix B). • The remaining potable water would undergo additional filtration as needed and no additional fresh water would be collected until a monitored degradation of residual bait pellets and rodenticide residue indicated that the potable water system was free of rodenticide. See Section 3.3.2 for a detailed discussion of water resources and management systems present on Sand Island, MANWR. • Implementation of the long-term biosecurity plan to prevent the reintroduction of mice on Sand Island, MANWR (see Appendix A).

2.3.8 CONTINGENCY PLANNING There is a chance that the project could fail to eradicate all mice after the main implementation phase. The follow-up actions should one or more mice be detected after the project is implemented will vary based on the time and circumstances of the detection. Additional baiting may occur in areas where mice are found if the circumstances dictate there is a reasonable chance of eliminating any remaining mice without additional adverse effects. Should the project fail entirely, and a mouse population persists, the partners will undergo reviews to determine the main causes of failure, and to identify possible solutions to those failures to inform a new strategy for future eradication attempts. Mouse population control using cholecalciferol would continue proactively and reactively to reduce mouse attacks on albatross while a new eradication strategy is being developed.

2.4 ALTERNATIVE 2: NO ACTION Analysis of the “No Action” Alternative, the alternative in which current conditions and trends are projected into the future in the absence of the proposed action, is required pursuant to NEPA [40 CFR 1502.14(d)]. In this case, under the No Action Alternative, Sand Island’s mouse population would not be the subject of a targeted eradication project, but mouse control efforts started in 2016 would continue to protect nesting seabirds and human health and safety.

2-17 MIDWAY SEABIRD PROTECTION PROJECT FINAL ENVIRONMENTAL ASSESSMENT DESCRIPTION OF THE PROJECT AND ALTERNATIVES

Mouse control on Sand Island currently consists of: (i) using traps and rodenticide bait stations in commensal areas such as food storage and preparation areas to maintain human health standards, as well as in areas near buildings, manmade structures, at the airport, and on shipping docks receiving conveyances; (ii) multi-catch live trapping; and (iii) rodenticide bait stations and hand broadcasting of AGRID3 Pelleted Bait for seabird and listed candidate plant protection in areas where mouse predation of seabirds and mouse damage to listed or candidate plants is detected. The bait product AGRID3, (active ingredient 0.075% cholecalciferol; Bell Laboratories, Madison, WI; approved for restricted use at MANWR under a supplemental label for mouse control) is a non-anticoagulant. It disrupts calcium homeostasis by increasing calcium absorption from the small intestine and mobilization from the bones into the blood stream, as well as decreasing calcium excretion by the kidneys. The effectiveness of this product on rodents has been proven in limited hand-broadcast situations, and it is relatively safe to nontarget species if used according to label directions (Marshall 1984, Eason et al. 2000, DurShultz et al. in press, USFWS, Unpublished b). Trapping includes: (i) multi-catch live traps; and (ii) mechanical traps (snap-traps and glue boards) used in the commensal environment as deemed appropriate per the terms of each structure’s SMP (see Section 2.3.1).

Hand broadcasting previously involved 2 separate hand-broadcast applications of AGRID3 pellets. Following label instructions, a 17.8 lbs./ac. (20 kg/ha) application rate was used over 27 ac (11 ha) where evidence of mouse attacks occurred in 2016. Future applications starting in November 2017 prior to the albatross’s main laying season, would occur at the same rate and area and then would occur as needed if and where mouse attacks occur throughout the incubation period, up until the following February. It should be noted that there were no observations of any non-target organism such as shorebirds or Laysan ducks interacting with AGRID3 bait pellets in the field or being found sick or dead in the colony as a result of the baiting process in 2016/2017 (DurShultz et al, in press; USFWS, Unpublished b). Toxicity tests by Eason et al. (2000) found that the lethal dose for a mallard duck is 4.4 lbs. (2000 g) of 0.08% cholecalciferol. There is the potential for Laysan ducks to consume some bait, but it is unlikely they would consume lethal amounts. Further, to reach a lethal dose, a Laysan duck would need to ingest 3 times its body weight in pellets, which is unlikely to occur (DurShultz et al. in press). Both control measures (i.e., trapping and rodenticide treatment) are ongoing. There are currently no other activities taking place at Sand Island, MANWR in terms of mouse management or control; however, other unrelated invasive species management and conservation operations at MANWR would continue per their respective agency plans. In addition, any other non-USFWS programs or projects would continue to be implemented under their respective authorities. The No Action Alternative would be contrary to: (i) general principles of wildlife management; (ii) the PMMP’s strategy AS-4; which calls for conservation actions to continue predator control and eradication of predators at Laysan and black-footed albatross nesting colonies; (iii) multi- agency conservation action plans for the black-footed and Laysan albatross (Naughton et al. 2007) that recommend eradicating house mice from MANWR to protect these 2 species of seabirds; and (iv) the USFWS Regional Seabird Conservation Plan (USFWS 2005) that identified the eradication

2-18 FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT DESCRIPTION OF THE PROJECT AND ALTERNATIVES

or control of introduced predators and other invasive species that have a negative impact on seabirds as one of the top priorities for seabird conservation. The USFWS Regional Seabird Conservation Plan specifically recommended eradicating introduced predators in the Pacific Islands where the Laysan and black-footed albatross breed. The purpose of the NWRS is to administer a national network of lands and waters for the conservation, management, and where appropriate, restoration of fish, wildlife, and plant resources and their habitats within the US and outlying territories for the benefit of present and future generations of Americans. Further, the No Action Alternative would be contrary to the USFWS’ policy on maintaining the biological integrity, diversity, and environmental health of the NWRS. Finally, it would not achieve the USFWS’s project objectives summarized in Section 1.2. It is included here because it is appropriate under the environmental review process mandated by NEPA, and to provide a baseline for comparison with the potential impacts of the proposed action.

2.5 ALTERNATIVES CONSIDERED BUT REJECTED The USFWS evaluated several different approaches to house mouse eradication. Ultimately the alternative approaches summarized below were rejected from further evaluation because of the lack of evidence that these alternative tools and strategies would have a reasonable likelihood of eradicating mice on MANWR or thus achieving the purpose and need for the Project. These approaches are briefly described in the following subsections.

2.5.1 EMERGING TECHNOLOGIES Ongoing research continues into the development of technologies for the control of introduced rodents. If these tools, currently in development, ultimately become available they may have either incremental or transformative impacts on the way in which eradication projects are approached. These technologies include: (i) reproductive inhibitors; (ii) species or specific toxins; and (iii) genetic bio-control. In a review of rodent control methods, Buckle and Smith (2015) state that reproductive control is not yet safe or effective for field use. Fertility control has been used with limited success as a method of pest management in a few species. Experimental sterilization methods have included chemicals and proteins delivered by vaccine, and genetically-modified viral pathogens. However, the effectiveness of these experimental techniques in the wild, and their impacts to nontarget animals are unknown. Aerial application of rodenticide is a more practical, effective, and a safer method to eradicate rats than repeated baiting of uncertain oral contraceptives on a remote island across seasons or capturing, vaccinating, and releasing every member of a single gender of the Palmyra rat population. This lack of data and tools disqualifies the use of fertility control from detailed consideration (Tobin and Fall 2005). Currently, however, the several methods mentioned above are in development and none are likely to be available prior to 2023, the deadline for eradication of house mice on Sand Island, Midway Atoll established by the PMMP-ASAP Strategy AS-4. Therefore, this approach was rejected from further evaluation because it would not achieve the project objectives summarized in Section 1.2.

2-19 MIDWAY SEABIRD PROTECTION PROJECT FINAL ENVIRONMENTAL ASSESSMENT DESCRIPTION OF THE PROJECT AND ALTERNATIVES

2.5.2 MOUSE ERADICATION USING CHOLECALCIFERAL (AGRID3) RODENTICIDE Cholecalciferol (Agrid 3 by Bell Labs), a non-anticoagulant rodenticide, can kill rodents more quickly than the anticoagulant rodenticides by disrupting calcium absorption and decreasing calcium excretion in the kidneys. It is an attractive alternative because of its lower toxicity to birds. This product is currently being used at MANWR to control mice (not eradicate them) by hand broadcast in limited areas. However, these non-anticoagulant rodenticides are untested on islands larger than 22 ha (~54 ac) (Howald et al. 2007). Furthermore, there is no cholecalciferal product registered by the EPA for aerial broadcast and the purpose of island-wide eradications for mice. Using MANWR as a test island, without a high probability of success, would be inappropriate due to the high financial cost of the operation. These factors, lack of field-testing on islands comparable to Midway, potential bait avoidance, and greater human safety risk disqualifies them from detailed consideration for use on a MANWR eradication.

2.5.3 USE OF FIRST-GENERATION RODENTICIDE Island Conservation’s Feasibility Report (2017) specifically recommends the use of a bait containing brodifacoum and notes that use of a less robust alternative rodenticide could compromise the success of the eradication effort. As mentioned in Section 2.2.3, first generation anticoagulant rodenticides, such as diphacinone, are multi-feed rodenticides and rodents need to consume bait over a sustained period of days to achieve mortality, which may vary from 3-12 days of ingestion of a lethal dose. Diphacinone has been successfully used to eradicate rats from islands, typically delivered by bait stations with relatively few instances of broadcast being the delivery method (DIISE 2016). It has been used in efforts to eradicate house mice on islands only twice, with one instance being a reported success and one a failure (Samaniego 2016). Second generation anticoagulant rodenticides, such as brodifacoum, can be toxic after a single exposure or feeding event to sensitive rodents (Kaukeinen 1993). In some laboratory trials of brodifacoum, 100% mortality of rats and mice have been reported after a 3-day choice test (Pitt et al. 2011). Globally, a total of 87 house mouse eradication attempts have been carried out, and all but the 2 mentioned above using diphacinone, utilized brodifacoum. Of the 31 house mouse eradication attempts carried out between 2005 and 2015, 28 of them were successful; this represents a 93.3% success rate (DIISE 2016). The anticoagulants brodifacoum and diphacinone both bind to the same molecular site in the liver. Ingestion of these compounds would only lead to death in the exposed animal if a toxic threshold of one or more of these compounds is maintained in the liver consistently for a sufficient period of time. Brodifacoum molecules bind comparatively tightly in the liver, so the same brodifacoum molecules that are ingested on Day 0 are highly likely to remain in the liver for an extended period (Fisher et al. 2003). Thus, a single feeding of a threshold quantity of brodifacoum is sufficient to ultimately induce lethal toxic effects in the exposed animal. Diphacinone, on the other hand, binds comparatively loosely in the liver, with a half-life measured in days (Fisher et al. 2003), so diphacinone molecules ingested on Day 0 are much less likely to remain in the liver for multiple days from a single feeding. Thus, to maintain a threshold lethal level of diphacinone in the liver consistently for sufficient time, a rodent needs to consume diphacinone in multiple feedings over multiple days so that it accumulates in the system faster than the liver processes it out. In application, this means

2-20 FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT DESCRIPTION OF THE PROJECT AND ALTERNATIVES

bait must be available to each rodent to feed on at multiple times over a few days, and thus a longer period of time compared to brodifacoum. Comparisons of Diphacinone to Brodifacoum Diphacinone presents less risk to non-target species than brodifacoum (Howald et al. 2007, Erickson and Urban 2004). As described above, species that might ingest bait pellets opportunistically or accidentally are much less likely to ingest a lethal amount of diphacinone in one exposure (Hoare and Hare 2006). Predatory and scavenger species are also less likely to consume a lethal amount of toxin when preying on individuals that are dead or dying from primary exposure to the rodenticide, since diphacinone is retained only at a low level in body tissues. However, the successful eradication of rodents on islands using baits is partly dependent upon overcoming bait avoidance by rodents and inadequate bait consumption. While diphacinone generally presents less risk to non-target species than brodifacoum, its low retention in body tissues and its multiple feeding requirement results in a higher risk of failure in rodent eradication efforts (Donlan et al. 2003). All other operational considerations being equal, the task of delivering enough bait to all rodents on the island to ensure 100 % mortality is less certain to succeed when using diphacinone than when using brodifacoum. Rodents that survive diphacinone bait application may be resistant to first-generation anticoagulants (e.g., diphacinone), or could be unable to find or consume adequate bait to maintain lethal effects of diphacinone dose, avoid bait after a single feeding, or a combination of these. These rodents, and possibly their offspring, could consequently be more difficult to eradicate from the island in any subsequent attempts. The use of the much more field-tested brodifacoum provides a greater degree of confidence that the eradication program would succeed. While both sides of this balancing equation are important, the need to ensure eradication success is particularly acute. An incomplete eradication attempt would provide few conservation returns in the long term, since mice at low densities have the ability to reproduce at higher rates and would quickly reoccupy vacant territories throughout the island. The only way to attain cost-effective, and ecologically secure, long-term conservation returns on this investment is through a successful and complete eradication. In summary, the safety risks associated with an aerial or hand broadcast operation can be somewhat mitigated with the use of a more potent toxicant such as brodifacoum. Therefore, the USFWS has rejected this alternative from further consideration.

2.5.4 MOUSE-PROOFING STRUCTURES Modifying existing manmade structures to be rodent-free may be a viable strategy to eliminate house mice within buildings but would come with considerable cost and have little utility once mice were eradicated from the remainder of the island. Further, mouse-proofing would not achieve the project objectives outlined in Section 1.2, and therefore has been rejected as an eradication strategy. It is more efficient and cost-effective to specifically target mice in and around manmade structures as part of a comprehensive eradication campaign. Therefore, the USFWS has rejected this alternative from further consideration.

2-21 MIDWAY SEABIRD PROTECTION PROJECT FINAL ENVIRONMENTAL ASSESSMENT DESCRIPTION OF THE PROJECT AND ALTERNATIVES

2.5.5 AERIAL BROADCAST RODENTICIDE ONLY An aerial broadcast only approach to mouse eradication on Sand Island, MANWR would not effectively administer bait to the buildings and other infrastructure located there, whether occupied or abandoned. This factor alone means that a supplemental treatment by some other method, such as hand broadcast, would be required to address this mouse habitat. Conducting a mouse eradication project on Sand Island, MANWR solely relying on aerial broadcast of rodenticide would be incomplete in its application (i.e., within structures and in below-ground infrastructure) and unlikely to attain the project objectives outlined in Section 1.2. Aerial broadcast only would also potentially lead to some bait drifting into the marine environment in areas where the shoreline is too narrow for aerial applications such as causeways or where seawalls and armored shorelines exist. Therefore, the USFWS has rejected this alternative from further consideration.

2.5.6 HAND BROADCAST RODENTICIDE ONLY If hand broadcast alone were employed, then distribution transects would need to be spaced a maximum of approximately 66 ft. (~20 m) apart to ensure enough bait was available in every mouse territory. To achieve that end, Island Conservation (2017) estimates that more than 200 worker-days would be necessary to achieve a single complete hand broadcast operation across all of Sand Island, MANWR. This approach would: (i) be very labor-intensive; (ii) would need to be very carefully implemented and managed; (iii) would require a period of 2 or more months to complete, assuming a work crew of 40 individuals and between 2 and 3 bait applications across the entire area of Sand Island; and (iv) would increase the risk of damage to seabird burrows and vegetation as a result of worker movements. It is unlikely that MANWR’s facilities and resources could effectively accommodate the large workforce needed to complete each application in a single day. Reducing the number of individuals working on the eradication would increase the number of days needed per application and would increase the complexity of the project. Finally, this time- estimate does not include applying bait within condemned and abandoned structures, where hand broadcast is not considered a viable method due to human health and safety concerns. Conducting a mouse eradication project on Sand Island, MANWR solely relying on hand broadcast of rodenticide would be time consuming, costly, unable to be adequately applied in some condemned structures, and unlikely to attain the project objectives outlined in Section 1.2. Therefore, the USFWS has rejected this alternative from further consideration.

2.5.7 BAIT STATION RODENTICIDE ONLY Specific considerations related to using bait stations as the sole means of mouse eradication on Sand Island, MANWR work against this alternative include the following:

2.5.7.1 Non-Target Species One of the benefits of bait stations is the reduced level of primary exposure to non-target species. On MANWR, bait stations would offer some, but not complete protection to terrestrial birds. Moreover, bait stations would not eliminate the secondary exposure risk through mobilization of rodenticide into the environment by contaminated insects (e.g., cockroaches, ants, etc.) and other invertebrates (e.g., crabs). As such, a bait station approach would minimize the non-target

2-22 FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT DESCRIPTION OF THE PROJECT AND ALTERNATIVES

poisoning risk but would not eliminate the need for additional mitigation measures to protect the Laysan duck and shorebirds.

2.5.7.2 Operational Challenges The logistics of implementing mouse eradication on Sand Island, MANWR using bait stations alone would be significantly more complex than rat eradication, primarily due to their smaller home range, which would require a much higher density of bait stations. Bait stations would need to be deployed at a maximum of approximately 66 ft. (~20 m) intervals, with approximately 33 ft. (~10 m) offering greater confidence that at least one bait station would be accessible to each mouse on the island. Assuming a 33 ft. x 33 ft. (10 m x 10 m) grid across the island, Island Conservation estimates that a minimum of 45,200 bait stations would be needed to cover the total area of Sand Island. This figure does not account for key commensal habitat which would require additional treatments. The Feasibility Report (Island Conservation 2017) estimates that more than approximately 280 miles (mi.) (450 kilometers [km]) of trails would need to be opened, flagged and maintained to support crews walking to install, service, monitor, and decommission these stations. These trails would need to be opened in key habitat such as the coastal fringe in high density Scaevola sp. and through habitat with Bonin petrel burrows, which are found wherever the substrate allows for excavation by the birds and thus are common on Sand Island. It is likely some burrows would be stepped on and collapse suffocating adults or young. Island Conservation estimates that, assuming a manageable crew size of 40 workers, this would require 200+ days, and an individual station would need to be visited at a minimum of 5-day intervals. Because conducting a mouse eradication project on Sand Island, MANWR relying solely on bait stations would be time consuming, impactful to sensitive seabird habitat, incomplete in its application (i.e., within condemned structures), and unlikely to attain the project objectives outlined in Section 1.2, the USFWS has rejected this alternative from further consideration.

2.5.8 MOUSE CONTROL Under this potential alternative, project planners considered mouse removal with the goal of reducing but not eliminating the mouse population on Sand Island, MANWR. A Mouse Control alternative would create and implement a strategic plan for long-term mouse control, with a regimented management plan written and approved, funds allocated, staff assigned, progress reports, and other elements that would distinguish it in scale, intensity, and duration from the mouse control efforts currently underway at Sand Island, MANWR. The net conservation benefit achieved by successful mouse control, as compared to total eradication, could be similar. However, the risk posed to non-target wildlife and island personnel as a result of continued control operations are greater than the risks related to an eradication operation due to the indefinite timeline for which a control operation would need to persist. The long-term presence of rodenticide and the repeated disturbances related to ongoing control operations would place non-target species at a continuous risk. In addition, should the scheduled operations be interrupted due to inclement weather or some other circumstance, the mice could quickly reproduce to, or beyond, current levels and repopulate Sand Island, MANWR. Should this occur, it would require a further intensification of control operations once more. Control of the island-wide mouse population to levels low enough to eliminate them

2-23 MIDWAY SEABIRD PROTECTION PROJECT FINAL ENVIRONMENTAL ASSESSMENT DESCRIPTION OF THE PROJECT AND ALTERNATIVES

as an ecosystem-wide threat would require constant maintenance of an ecologically beneficial mouse control program. Doing so would be far less cost-effective, increase personnel safety risks, and would not result in the permanent conservation benefit derived from island-wide eradication of house mice. Because this alternative would not achieve the project objectives in Section 1.2, the USFWS has rejected this alternative from further consideration.

2.6 PROJECT COSTS The USFWS’ total estimated implementation budget for the Midway Seabird Protection Project is $4.5 to 5 million dollars.

2-24 FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, & MINIMIZATION MEASURES

3.1 GENERAL DESCRIPTION OF MANWR MANWR is located at approximately 28˚15ʹ N and 177˚20ʹ W in the NWHI and consists of 3 sandy islets. Sand Island, where the mouse eradication project would take place, is the largest islet at 1,117 acres (452 ha); the other 2 islets are Eastern Island 366 acres (148 ha) and Spit Island 15 acres (6.1 ha) (Figure 1.2). Together, they encompass a total land area of 1,549 acres (627 ha) with a mean elevation of approximately 10 ft. (3 m) above mean sea level (+MSL). Together these 3 islets lie in the southern portion of a large, elliptical barrier reef measuring nearly 5 mi (8 km) in diameter. MANWR is one of the northernmost land masses in the NWHI, located approximately 1,313 mi (2,113 km) northwest of Honolulu, Hawaiʻi (see Figure 1.1). MANWR became an overlay refuge in 1988, while remaining under the jurisdiction of the Department of the Navy. On October 31, 1996 President William Clinton officially established MANWR as a standalone refuge by Presidential EO No. 13022. On September 13, 2000, the lands and waters of MANWR were designated as the Battle of Midway National Memorial. In addition, on June 15, 2006, the lands and waters of MANWR were incorporated into the PMNM by President George Bush’s Presidential Proclamation No. 8031. MANWR’s Co-Trustees in administrative matters, along with their respective responsibilities are summarized in Table 3.1 below.

Table 3.1 MANWR Administrative Co-Trustees Co-Trustee Responsibilities Secretary of Commerce, National Primary responsibility regarding the management of the marine areas of Ocean Atmospheric the Monument, in consultation with the Secretary of the Interior. Administration (NOAA) Secretary of the Interior, USFWS Sole responsibility for the areas of the Monument that overlay MANWR, the Battle of Midway National Memorial, and the Hawaiian Islands National Wildlife Refuge, in consultation with the Secretary of Commerce. State of Hawaiʻi, Department of Primary responsibility for the Northwestern Hawaiian Islands Marine Land and Natural Resources Refuge and State Seabird Sanctuary on Kure Atoll. Nothing in the Proclamation diminishes or enlarges the jurisdiction of the State of Hawaiʻi. State of Hawaiʻi, Office of Consultation on matters pertaining to Native Hawaiian culture. Hawaiian Affairs Source: USFWS (2017)

Together, USFWS and its Co-Trustees coordinate with the U.S. Coast Guard as they exercise their law enforcement, search and rescue, and medical evacuation responsibilities in the Central Pacific. The Coast Guard works with USFWS to store aircraft fuel on MANWR for mission-related use, and occasionally crews will stay on Midway during extended operations. Although geographically part of the Hawaiian Islands archipelago, MANWR is not part of the State of Hawaiʻi and is an unincorporated territory of the United States. Therefore, the State Hawaiʻi has no jurisdiction on MANWR. Current funding to operate MANWR comes from the

3-1 MIDWAY SEABIRD PROTECTION PROJECT FINAL ENVIRONMENTAL ASSESSMENT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

USFWS, supplemented by the FAA, which fully funds airport operations costs and a share of infrastructure operations costs. A small amount of funding is generated by the other users of MANWR, such as other Federal agencies conducting activities on Midway.

3.2 RESOURCE TOPICS EXCLUDED FROM DETAILED ANALYSIS The resource topics in this section are those that do not have the potential to be affected by the proposed action or action alternatives. They are briefly summarized here in the interest of completeness and to provide context for readers, but there is no attendant discussion of impacts because no such potential exists.

3.2.1 LAND USE The built environment on Sand Island is extensive and a result of a long history of human habitation and infrastructure, representing over a century of occupation. The structures on Sand Island vary in age, construction type, condition, and on how they are used (if at all). The oldest structure dates back to 1903 and the Commercial Pacific Cable Company (one structure remains from that time); this and other historic buildings are discussed in Section 3.2.12. Nearly all the other structures, except the memorials, date to the Navy’s use of Midway from the 1930s to 1993. The types of structures present on Sand Island and their use are summarized in Table 3.2; more generalized land uses are shown in Figure 3.2. Eastern Island lies nearly a mile to the east of Sand Island. It is uninhabited today but was occupied during World War II. Eastern Island has been uninhabited since 1970. Like Sand Island, Eastern Island is naturally composed of coral sand and enhanced with man-made fill. A large part of Eastern Island is an abandoned three-runway airfield, and the rest is mainly open fields, roads, and sand beaches. Rats were eradicated from Midway Atoll in 1995 and house mice are not present on Eastern Island. There are no structures on Eastern island other than the dock and former unused runways. The only human activities on Eastern Island currently are: (a) bird surveys; (b) vegetation restoration and; (c) guzzler maintenance for the Laysan duck.

3-2 FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

Table 3.2 Summary of the Built Environment on Sand Island, MANWR Type Use Inhabited Space Living spaces, such as housing and dining and recreation facilities, food storage, and food preparation sites. These include: • Clipper House dining hall • Captain Brooks pub • Charlie Barracks • Houses near Charlie Barracks (i.e., Pranee, Dolphin, and Yoodee houses) • Midway Mall, which includes a store among other facilities Work spaces, including offices (NWR Visitor Center and Chugach offices), Aircraft Rescue and Fire Fighting (ARFF) facilities, utility buildings, covered storage, covered gardens, and plant nurseries. Abandoned Condemned structures where routine access is not permitted due to safety or other Structures concerns, but all buildings can be entered with proper personal protection equipment (PPE). Memorials Monuments to the Battle of Midway and other features of historic importance. Aboveground Including water tanks, fuel storage tanks, electrical equipment, and other facilities. Utility Infrastructure Subterranean The majority of this infrastructure is defunct and has been abandoned in place, but select Utility facilities are still in use in inhabited areas and include utility service boxes, electrical and Infrastructure communications conduits, water and fuel pipelines and valves, and wastewater pipelines and valves. Source: USFWS (2017)

3-3 MIDWAY SEABIRD PROTECTION PROJECT FINAL ENVIRONMENTAL ASSESSMENT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

Figure 3.1 Generalized Land Uses on Sand Island

Source: USFWS (2017)

In addition to the physical infrastructure summarized above, there are 3 types of land use designations present on Sand Island, MANWR. Each is summarized in one of the following subsections.

3.2.1.1 Wildlife Refuge As noted above, on October 31, 1996, President William Clinton officially established MANWR as a standalone refuge via EO No. 13022. The refuge encompasses all islands and waters within the 12-nautical mi. territorial sea of the atoll, totaling 581,864 ac. (235,472 ha). The refuge is administered by the USFWS in order to: • Maintain and restore natural biological diversity within the refuge. • Provide for the conservation and management of fish and wildlife, and their habitats, within the refuge. • Fulfill the international treaty obligations of the United States with respect to fish and wildlife. • Provide opportunities for scientific research, environmental education, and compatible wildlife-dependent recreation activities.

3-4 FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

• Recognize and maintain the historic significance of MANWR in a manner compatible with refuge purposes. NOAA National Marine Fisheries Service (NMFS) also monitors Hawaiian monk seals in MANWR and has taken periodic actions with seals to enhance their survivability as part of a cooperative conservation effort between NOAA, USFWS, USCG, and several non-governmental organizations (NGOs).

3.2.1.2 National Memorial On September 13, 2000, Secretary of the Interior Bruce Babbitt designated the lands and water of MANWR as being within the Battle of Midway Memorial, “so that the heroic courage and sacrifice of those who fought against overwhelming odds to win an incredible victory will never be forgotten.” MANWR is the first National Memorial co-located with a National Wildlife Refuge. Several monuments are present on Sand Island, honoring those who sacrificed their lives with remnants of the actual historic Battle of Midway during the Second World War.

3.2.1.3 Airport Henderson Field Airport, designated by the International Air Transport Association as airport code PMDY, has a 7,900 ft. (2,407 m) runway capable of handling almost any type of aircraft. PMDY is a fully-certified airport maintained according to the standards specified in the FAA’s Title 14 CFR, Part 139. Midway is used as a required emergency landing site for extended twin-engine operations (ETOPS) flights across the Pacific Ocean. Under current regulations, twin-engine aircraft must be within a maximum of 180 minutes from a Title 14 CFR Part 139-certified airfield in case of an emergency.

3.2.2 ACCESS MANWR and PMNM are not typically open to visitors, and only those with a role in the operation, maintenance, or purpose of the refuge or the monument can access Sand Island. The only means of accessing MANWR are via air transport or surface vessel. These means of access currently adhear to a biosecurity plan designed to address a myriad of threats and concerns across the entire PMNM, adhearance to the biosecurity plan to prevent the reintroduction of mice on Sand Island (see Appendix A) will not adversely affect continued access to MANWR and PMNM.

3.2.2.1 Air Transportation Currently, flights to MANWR are determined on an as-needed basis for the purposes of transporting support staff, scientists, volunteers, and restocking perishable foods. While the frequency of these flights varies, typically 2 flights occur per month, originating from Honolulu International Airport. A Gulfstream III jet aircraft is typically used, with a crew of 2 and a total capacity of 15 passengers.

3.2.2.2 Marine Transportation The only vessels permitted within MANWR are those that service activities and personnel within the monument, and those transiting through without stopping; as a consequence, very few vessels

3-5 MIDWAY SEABIRD PROTECTION PROJECT FINAL ENVIRONMENTAL ASSESSMENT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

call at MANWR. Small craft typically enter the inner harbor and moor dockside, or in the harbor. Larger vessels used to resupply Sand Island and research vessels generally make their landing at the cargo pier, inside Midway’s lagoon but outside the inner harbor. Larger passenger vessels, when present, are required to remain outside the reef and shuttle passengers onto the atoll via launches or tending craft due to port security requirements. Annually, visits to MANWR by marine vessels generally include the following: (i) 1 barge, associated with construction projects, bringing construction materials as well as general operations material; (ii) 2 or more support barges; and (iii) between 3 and 5 NOAA research ships. The only access to Eastern Island is to travel by small boat from Sand Island across the channel to the small wooden pier located on the northwest part of Eastern Island.

3.2.3 CLIMATE The climate of MANWR is influenced by the marine tropical and marine Pacific air masses, depending on the season. During the summer months, the Pacific High-Pressure System becomes dominant, with the ridgeline extending across the Pacific north of Kure and MANWR s. This places the region under the influence of easterly winds, with marine tropical and trade winds prevailing. During the winter, particularly between November and January, the Aleutian Low, a semi-permanent low-pressure system, moves south over the North Pacific, displacing the Pacific High before it. The Kure-Midway region is then affected by either the marine Pacific or marine tropical air masses, depending upon the intensity of the Aleutian Low or the Pacific High-Pressure systems. The weather on Sand Island, MANWR is monitored at Henderson Field Airport (see Section 3.2.2.1). Two seasons dominate the annual climate-cycle. During the warm season, which typically extends from late June through early October, the average high temperature is 85 ˚F (29 ˚C) and the average low temperature is 78 ˚F (25˚C). During the cool season, which typically spans from late December to mid-April, the average high temperature is 71 ˚F (21˚C) and the average low temperature is 63 ˚F (17˚C). The average annual rainfall is 41.3 in (104 cm), with January being the wettest month with an average of 5 in (13 cm) of rain. June is typically the driest month with 2.2 in. (5 cm) of rain (weatherbase.com). Dramatic variation in precipitation timing and volume is possible during El Niño and La Niña conditions. Sand Island is located at 28 12 N, and daylight hours range from 10:22 and 13:55 hours per day, with the longest day occurring in June during the boreal summer. The typical seasonality of rainfall at MANWR is depicted in Table 3.3 below, showing higher monthly totals and number of days with precipitation in the winter months. During some exceptional years, such as 2015-2016, the pattern may be reversed. Precipitation normally occurs as rain, ranging from mist to moderately intense rainfall, with total monthly averages ranging up to 5 in (13 cm). Despite significant seasonal fluctuation, rainfall occurs throughout the year, generally in the form of light, intermittent showers. However, rainfall is generally lowest in April, May, June and July (Table 3.3).

3-6 FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

Table 3.3 Average Rainfall on Sand Island, MANWR Average Precipitation 1974-2016 (in.) Jan. Feb. Mar. Apr. May Jun. Jul. Aug. Sep. Oct. Nov. Dec. Annual 5 3.8 3 2.5 2.3 2.2 3.3 4.3 3.5 3.5 3.8 4.1 41.3 Average No. of Days with Precipitation 1977-2016 Jan. Feb. Mar. Apr. May Jun. Jul. Aug. Sep. Oct. Nov. Dec. Annual 16 14 12 11 9 9 15 15 15 14 14 16 160 Source: Weatherbase.com (2017)

3.2.4 VISUAL RESOURCES Sand Island, MANWR is comprised of dunes, grasslands, forested areas, beaches, the protective barrier reef and lagoon, each with its own distinct character and visual appeal. It also encompasses several monuments to the Battle of Midway during the Second World War, as well as other points of human interest. However, none of these features has been formally designated as a protected vista or scenic viewpoint. In addition, because visitation to Sand Island is purposefully limited, there are very few individuals present to observe the natural and manmade environment present. Eastern Island is naturally composed of coral sand and enhanced with man-made fill. A large part of Eastern Island is an abandoned three-runway airfield, and the rest is mainly open fields, roads, and sand beaches. Eastern Island has even fewer visitors (mostly USFWS staff) than Sand Island.

3.2.5 FUEL FACILITIES A new jet fuel tank farm was constructed in 2007 with a capacity of 450,000 gallons (1,703,435 liters); the facility consists of 9 50,000-gallon (189,271-liter), aboveground storage tanks located on the southwest side of the Inner Harbor (see Figure 1.3). The tank farm stores a sufficient amount of fuel to operate electrical generators, vehicles, and aircraft for a year. The USCG and USFWS have an interagency agreement that covers this cooperative effort and outlines shared costs associated with fuel.

3.2.6 ENERGY Electrical power at MANWR is supplied by generators utilizing jet fuel. Two generators operate in automatic duplex mode such that, in most cases, only 1 generator is needed to meet Sand Island’s demand. If 1 generator exceeds capacity, the second generator automatically comes online and automatically shuts off when electrical demand reduces. The current system for generating electricity is sufficient for the existing population. Midway has 2 electrical distribution grids. A relatively new electrical distribution grid serves most of Sand Island, but portions of the older grid still provides power to the old airport hangar and the finger pier area. Finally, photovoltaic panels are used to power select energy demands, such as food storage refrigerated containers.

3.2.7 COMMUNICATIONS Telecommunication is provided by satellite service and includes T-1 data and VOIP service. Communications are provided to the inhabited portions of Sand Island via a fiber optic distribution system. Radio communication is also widely used in place of cellular phone service; the service area extends throughout both Sand and Eastern Islands.

3-7 MIDWAY SEABIRD PROTECTION PROJECT FINAL ENVIRONMENTAL ASSESSMENT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

3.2.8 AIR QUALITY Air quality at MANWR is generally excellent. Sources of emissions on the atoll are minimal and include electrical generators, motor vessels, vehicles, tools, and occasional plane flights (U.S. Army Space and Missile Defense Command, 2003).

3.2.9 GEOLOGIC RESOURCES Midway is one of the northernmost coral atolls in the world and only one other atoll 90 mi. (145 km) to the north of Midway, Kure Atoll, also has emergent land. MANWR is 28.7 million years old. It is the result of a classic geomorphological sequence along the Hawaiian ridge, ranging from brand new volcanic basalt islands at the Southeastern end of the chain to true ring-shaped atolls at the Northwestern end of the chain and submerged seamounts even farther to the northwest. As its name implies, Sand Island is comprised of coral sand, as is Eastern Island and the relatively recently formed Spit Island. Despite that the foundation of the atoll is volcanic, no in-situ volcanic rocks are found on the islands. The former volcanic island slowly eroded and sank to an elevation well below current sea level. As sea level rose and the island continued to sink, coral grew on top of the volcanic rock to form the atoll that is present today. Portions of both Sand and Eastern Islands are man-made fill. The fill is primarily believed to be coral sand from the dredging of the channel and harbor at MANWR. Other sources of fill are believed to be relatively minor and include: (i) large rocks from Hawai‘i and the US mainland, used to build shoreline revetments; and (ii) some organic material, such as soil from the US mainland, reportedly imported to help establish trees on the islands. Landfills, which are discussed in Section 3.3.5, also created landforms, most notably the “bulky dump” finger extending out from the south side of Sand Island (see Figure 1.3).

3.2.10 SOCIOECONOMIC AND ENVIRONMENTAL JUSTICE Sand Island, MANWR does not have an established demographic or socioeconomic structure. The population of the Atoll fluctuates based on the needs of MANWR and PMNM; all inhabitants are on the Atoll strictly to operate, maintain, and address these two purposes. As all inhabitants are professionals employed in these undertakings, there are no environmental justice populations present there. The year-round community of people on MANWR includes USFWS staff and volunteers, base operations staff (mostly comprised of Thai nationals employed by Chugach Management) and temporary contractors or research scientists. Sand Island has extensive infrastructure, processes, and utilities in place (see Table 2.2) to support the approximately 60 people that live and work there, including housing, common eating spaces (known as the Clipper House), small scale agriculture, recreation facilities, transportation infrastructure, recycling, and liquid and solid waste disposal systems. In effect, Sand Island is a functioning, albeit small, municipality. Atoll residents live in renovated Navy housing, including single family homes, duplexes, and Bachelor Officer Quarters (BOQ). One BOQ, known as “Charlie Barracks,” which contains 36 rooms, has been set aside for transient and visitor use. Almost all of the residents and transients eat at the Clipper House, where 3 meals a day are served buffet style. Most supplies, particularly

3-8 FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES foodstuffs, are flown to the island on chartered aircraft. Approximately twice a year, a ship brings in equipment, additional food, and supplies too large or heavy for air transport. Eastern Island has no infrastructure or utilities and no residents. The typical complement of personnel present on Sand Island, MANWR is summarized in Table 3.4 below.

Table 3.4 Typical Personnel Present on Sand Island, MANWR Type Current Purpose Number Refuge Staff 4 Full-time USFWS employees tasked with MANWR operations. Operations 37 Chugach Management Services employees responsible for the Contractors operation and maintenance of the active infrastructure on Sand Island. Construction <30 During construction season, between August and October, additional workers responsible for construction and demolition operations on Sand Island. FAA-Airport 4 Airport Maintenance and Management Volunteers 1-4 Volunteers periodically assist the regular MANWR staff in biological and habitat management activities. Researchers Variable Sand Island, MANWR also hosts transient researchers, USFWS employees, and USCG personnel on an irregular and periodic basis. Transients Variable Refuge employees, Co-Trustee staff, USCG and other law enforcement entities, contractors, researchers, and other Federal and state employees. Source: USFWS (2017)

3.2.11 SOLID WASTE DISPOSAL Solid waste generated in MANWR is disposed of via several methods. Most solid waste, such as household and food waste, is temporarily stored in open plastic trash bins with periodic trash collection via flatbed truck. Once collected, the solid waste is then burned in an oil-fired incinerator, dependent on the availability of waste fuel, or burned in an unlined open-air pit. Once burned, the ash is collected and disposed of at the existing landfill present on Sand Island. Aluminum cans are collected, compacted, stored, and then periodically sent to a recycling facility in the main Hawaiian Islands. Glass is collected, crushed, and buried in the landfill. Because the capacity of the existing landfill is limited both in capacity and in the types of waste which it can accommodate, it is only used for items which cannot be incinerated (e.g., some manmade marine debris).

3.2.12 HISTORICAL AND CULTURAL RESOURCES AND VALUES Study of Midway’s heritage resources was initiated in 1986 by the National Park Service when it conducted a survey of World War II-era properties eligible for designation as a National Historic Landmark. Nine structures, all defensive positions on the west side of Sand Island, were identified on Midway to convey a close association with the pivotal Battle of Midway, including ammunition magazines (ARMCO huts), a pillbox, and gun emplacements. Later that year, the 9 defensive

3-9 MIDWAY SEABIRD PROTECTION PROJECT FINAL ENVIRONMENTAL ASSESSMENT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

positions on Sand Island identified as eligible by the National Park Service and surrounding buffer areas were designated as a landmark. Between 1992 and 1994, the Navy sponsored studies of the Naval Air Facility on Midway, including archival research, interviews, and field surveys. The initial field effort consisted of an architectural history survey of the structures, buildings, and objects located on Sand and Eastern Islands. The study of Cold War Resources was conducted in 1993-94 by contractors hired by the Department of the Navy in order to identify the most important Cold War-era resources, even though they were less than 50 years old, as part of the Base Closure process. The historian hired to conduct the inventory, research, and make recommendations regarding the significance of the buildings on Midway was a specialist in the Cold War period. The Cold War-era buildings were constructed on Midway between 1957 and 1969. The recommendation accepted by the Navy was that the Cold War-era buildings and structures on Midway lacked architectural merit, were not directly associated with President Nixon’s visit, and do not convey a direct link to the events that occurred during the Cold War. The Navy subsequently demolished many of the Cold War-era buildings and structures prior to the transfer to the FWS. In addition to the landmark structures, 69 buildings, structures, and objects associated with the 1903-1945 historic period on Sand and Eastern Islands were determined to be eligible according to criteria established for the National Register of Historic Places. The properties evaluated as significant are associated with 3 major themes: colonization, initial years of base construction and the Battle of Midway, and 1942-1945 base construction. In 1859, Captain N.C. Brooks was the first Westerner to “discover” Midway aboard the Gambia from Honolulu. He claimed Midway for the U.S., based on the Act of 1856, which authorized Americans to temporarily occupy uninhabited islands to obtain guano. Captain Brooks named the atoll “Middlebrooks,” reflecting its position between the U.S. west coast, , and himself. The United States took formal possession of the unoccupied islands in 1867. Later, the name was changed to Midway. The Commercial Pacific Cable Company Site, which includes 5 buildings (Nos. 619, 623, 626, 628, and 643) on Sand Island, has been determined by the Secretary of the Interior to be eligible for inclusion in the National Register of Historic Places. Originally constructed by the Commercial Pacific Cable Company, these buildings are the only remnants of the initial permanent colonization of the Midway Islands. Archaeological surveys of Sand and Eastern Islands were conducted in 1992 and 1994. Surface inspections, 68 subsurface core samples, and 5 shovel-test units were conducted in disturbed sediments with as much as 2 meters of fill in some areas. Heavy construction and military use since 1940 changed the island considerably. Prior to the military era, these islands were also periodically scoured by storms and high winds that may have removed or buried evidence of pre- 1900 use. A review of Hawaiian chants, genealogy, mythology, and oral histories found numerous references to distant low lying islands with abundant birds and marine life. Kuaihelani which translates to “back bone of heaven” is a name that is referenced often with these sources and is potentially associated with Midway. This homeland of the gods is described as a floating island

3-10 FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

in the sky, which can derive from a lagoon’s refection in the clouds, such as at Midway Atoll. Pihemanu is a recent name that was given to the Atoll and means “loud din of birds” perfectly describing the current experience at MANWR. While no evidence of Polynesian/Hawaiian or pre- 1900 historic period cultural remains have been found to date, there is still the potential for a buried discovery. In addition, the chants, genealogies, mythologies and oral histories confirm that Hawaiians traveled to the islands and atolls within the Northwestern Hawaiian Islands over the span of several centuries.

3.2.12.1 Programmatic Agreement and Treatment of Midway’s Historic Properties In 1996, the Navy’s Pacific Division, Naval Facilities Engineering Command, the Advisory Council on Historic Preservation, and USFWS signed a programmatic agreement directing how MANWR’s historic properties were to be treated during the closure of Naval Air Facility Midway. These properties were assigned to 1 of 6 categories of preservation treatment: reuse and maintain, secure and abandon in place, abandon in place and leave as is, fill or cover, relocate, or demolish. USFWS was required to prepare a long-term Historic Preservation Plan, which it completed in 1999.

3.2.12.2 Historic Preservation Plan The December 2010 Midway Atoll National Wildlife Refuge Historic Preservation Plan defines a program to integrate historic preservation planning with the wildlife conservation mission of USFWS at MANWR. The plan focuses on the long-term management conditions and goals for preserving and stabilizing historic properties. It also recommends procedures for treating new discoveries, caring for collections, and implementing a visitor program that includes historic preservation work. In the future, the Co-Trustees will incorporate submerged cultural resource protection into such plans.

3.3 RESOURCES TOPICS ANALYZED IN DETAIL IN THIS EA The resource categories in this section are those that have the potential to be affected by the proposed action. In the following subsections the existing condition of each resource is characterized, followed by a discussion of the potential for impacts resulting from the Preferred or No Action Alternatives, including any mitigation activities which could avoid, minimize, or mitigate these potential impacts. Because no bait will be dropped on Eastern Island, and only a small area of the island will be used for mitigation activities associated with the care and maintenance of Laysan ducks (see Appendix C), Eastern Island is not described or assessed in detail in Chapter 3.

3.3.1 AIRFIELD OPERATIONS

3.3.1.1 Existing Conditions As described previously in Section 3.2.1.3, the active airport on Sand Island, MANWR is Henderson Field Airport. Henderson Field has a 7,900-foot (2,405 m) runway, is fully certified as an airport maintained according to all applicable FAA standards, and is capable of handling virtually any type of aircraft. Because of its capacity, Henderson Field Airport is used as a required

3-11 MIDWAY SEABIRD PROTECTION PROJECT FINAL ENVIRONMENTAL ASSESSMENT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES emergency landing site for ETOPS flights across the Pacific Ocean. It is staffed by air traffic control and ARFF personnel sufficient to operate as an ETOPS landing site. This airport is one of the dominant features present on Sand Island and, along with the harbor, is the principal means of accessing MANWR. The old aircraft runways present on Eastern Island are no longer used.

3.3.1.2 Probable Impacts No evidence suggests that the physical airfield, which consists primarily of hardtop asphalt and cement, support house mice. However, the proposed action does call for the treatment of potential mouse habitat in areas immediately surrounding the airport runway, including any subterranean utility access points. This treatment is particularly necessary in areas where vegetation has created significant ground cover and in the airfield utility access points, which create breaks in the otherwise uniformly sealed perimeter surfaces. A runway safety zone skirting, taxiway safety zone skirting, and 2 blast pads make up the area immediately adjacent to surfaces requiring Foreign Object Debris (FOD) management; these areas are delineated by large, conspicuous painted aircraft control lines on the tarmac. Baiting within these sites—up to but not within the aircraft control lines—and the airfield utility access points was assessed by airfield management and determined not to present a FOD hazard. Should bait pellets inadvertently drift onto the active airfield, pellets would be removed via an FOD-assessment following each bait distribution event. Additional flights and helicopter operations would be required to conduct the proposed action. Additional flights would include charter aircraft required to bring personnel, bait, and equipment to Midway. It is anticipated that these additional flights would not exceed an average of 6 operations a day over the 3-week bait distribution period. Given airport staffing levels, the low level of other operations, and apron area size, which once supported frequent and numerous military operations, the increase in operations anticipated to occur during the proposed action would not adversely affect the airport’s ability to continue to serve its ETOPS emergency landing site function or other functions.

3.3.2 WATER RESOURCES Fresh water is defined as water having a sufficiently low salinity to allow for its consumption without ill effect. Salinity is a measure of the total amount of salt in water. Salinity and total dissolved solids are essentially synonymous. Water is considered “fresh” when its salinity is less than 0.5 parts per thousand (ppt). The salinity of ocean water is generally 35 ppt. Under the Safe Drinking Water Act, the EPA has established a Secondary Drinking Water Regulation concentration limit of 0.5 ppt (which is the same as 500 milligrams per liter (mg/L) and 500 parts per million (ppm)) for total dissolved solids.

3.3.2.1 Existing Conditions: Groundwater Midway Atoll’s subsurface geology and hydrogeology is similar to other atolls in the Northern Hawaiian chain and elsewhere. The atoll’s geology is discussed briefly in Section 3.2.9. There are no hydrogeology-specific publications publicly available regarding Sand Island’s groundwater resources. However, a hydrogeological study of Wake Island (AFCEE 2006), another atoll where

3-12 FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

bait has been used to eradicate rodents and annual rainfall is similar, provides insight into atoll hydrogeology. Numerous wells have been drilled into the uppermost saturated zone of Sand Island to assess, remediate, and monitoring landfills and petroleum releases. The results of those investigations have yielded no evidence that Sand Island’s groundwater processes or conditions differ meaningfully from other atolls. Rainfall percolates through the coral sand and joins the underlying groundwater aquifer of Sand Island, MANWR. This unconfined (i.e., not confined under pressure beneath relatively impermeable rock) groundwater remains in contact with the atmosphere due to the porous nature of the coral sand soils. These porous soils at or below sea level—approximately 6 to 8 ft. (1.8-2.4 m) below grade—also are infiltrated and saturated by sea water. As the freshwater encounters the sea water, a zone of brackish water of variable salinity develops as they intermix. The depth and thickness of this brackish layer varies with the magnitude of tidal fluctuations, the amount of rainfall, and the permeability of the sediments through which the tidal signal occurs. The shallowest and most interior portions of Sand Island’s aquifer are the freshest, due to saturation with percolating rainwater; this groundwater flows down and outward towards the shore, where it discharges into the ocean. The gradients and flux are small (GeoEngineers 2011); at Wake Island it was assessed that it would take 30 to 50 years for groundwater to migrate from the center of the island to either shore (a distance of approximately 1,000 ft. [305 m]) (AFCEE 2006). Residence times at Sand Island could be as much as 3 times longer because it is roughly 6,000 ft. (1,828 m) across. There are no wells used for potable drinking water on MANWR.

3.3.2.2 Existing Conditions: Surface Water There are no streams or lakes on Sand Island, MANWR but there are several natural and man- made seeps present on the Atoll, which provide habitat for Laysan ducks, shorebirds, and migratory waterfowl. The seeps include small pools near the water tanks, in the area of former housing along Henderson Street, and west of the active dump (Figure 3.1). There are no 100-year floodplains on Sand Island, MANWR; however, the Atoll is surrounded on all sides by the Pacific Ocean and is periodically subjected to storm surges. The coral sand that makes up the island is sufficiently permeable that erosion and flowing surface water is not present except where storm water flows are concentrated by hardened surfaces. Storm water is discussed further in Section 3.3.4.

3.3.2.3 Existing Conditions: Potable and Non-Potable Water Systems Freshwater is a critical resource for people and wildlife present on Sand Island, MANWR. During substantial rain events, runoff from the airport runway is channeled into a pond adjacent to the runway and pumped into 3 large storage tanks for both potable and non-potable uses (Figure 3.2). Each of the 3 tanks has a 4.2 million-gallon (MG) (15,898,729 liters) capacity, for a total holding capacity of 12.6 MG (47,696,188 liters). Water is transferred from these 3 tanks to 2 approximately 80,000-gallon (302,833-liter) tanks in town, as needed, and from there directed into either the potable or non-potable systems. The non-potable system was originally built by the Navy for its use, and is now used for fire suppression, watering plantings, and other non-potable uses. The potable system consists of a treatment facility and separate piping system that distributes potable water to inhabited buildings.

3-13 MIDWAY SEABIRD PROTECTION PROJECT FINAL ENVIRONMENTAL ASSESSMENT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

The treatment system for potable water consists of sediment filtration to 5 microns and chlorination. In some of the older buildings, where the plumbing is aging, the water is also filtered at the tap to address secondary particulate. The potable water in use on Sand Island is sampled and tested monthly, at randomly selected points in the system, to ensure the safety of the supply.

Figure 3.2 Water Resources on Sand Island, MANWR

Source: USFWS (2017)

Typically, water is transferred from the pond to the 3 large runway tanks once a year, following a heavy rainfall or storm event. As storm events are more frequent during the winter months, water is usually collected and transferred at these times. Total annual water use is approximately 5 MG (18,927,058 liters) across all uses; thus, when the 3 large storage tanks are full, they provide over 2 years of water capacity.

3.3.2.4 Probable Impacts Evaluation criteria for effects on water resources are based on water availability, quality, and use, and the regulations associated with them. A proposed action can be characterized as having adverse impacts on water resources where it can be established that one or more of the following shall occur as a result of its implementation: • Reduce water availability to supply to existing users; • Overdraft groundwater basins;

3-14 FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

• Exceed safe annual yield of water supply sources; • Affect water quality adversely; • Endanger public health by creating or worsening a hazard to human health; • Threaten or damage unique hydrologic characteristics; or • Violate established laws or regulations adopted to protect water resources. The proposed action would not reduce water availability or exceed the safe yield of water supply sources. The 3 water tanks on Sand Island would be filled during the winter rains so that they hold sufficient water to supply all island residents and visitors for a period of roughly 2 years. Because all the water is drawn from the pond that collects storm water runoff from the runways, there is no danger of groundwater basin overdraft when the 3 tanks are filled in the winter season prior to the proposed action. The watershed for the seep is at least 200 acres (81 ha) and is primarily new or aging asphalt pavement. At that size, less than 2 in (5 cm) of rain is necessary to provide sufficient runoff to completely fill the 3 tanks. Winter storms in December, January, and February commonly drop 4 to 5 in. (10-13 cm) of rain each month (Table 3.3). Because all water for human consumption for a 2-year period would be collected in enclosed tanks prior to broadcasting the bait, no public health threat is anticipated associated with drinking water or other water use. The effect of flood hazards on a proposed action may also be important if such an action is in an area with a high probability of flood or, with greater relevance to MANWR, storm surge. With regard to surface water quality, no adverse effects are anticipated as a result of the proposed action. Brodifacoum, the toxin proposed for the mouse eradication effort, has very low water solubility (Primus et al. 2005). Bait that enters any given water column, whether marine or fresh, would dissolve into individual grain particles, and the brodifacoum molecules would remain bonded to these grain particles. In addition, the bait proposed for use during the eradication operation would contain a poison concentration of approximately 0.0025% brodifacoum. At this concentration level, brodifacoum which if inadvertently introduced to Sand Island’s surface water, groundwater, or the marine environment, would likely be below risk levels. No bait would be dropped on Eastern Island since this island is already mouse-free. Similarly, no groundwater quality adverse effects are anticipated as a result of the proposed action. Should brodifacoum dissolved at low concentrations (it is nearly insoluble) enter the groundwater aquifer, it would be degraded by bacteria in the subsurface, be diluted, and take years or decades for that water to discharge into the marine environment. Hydrogeological studies at Wake Island found that groundwater chemistry and temperature conducive to the biodegradation of organic compounds like brodifacoum, similar conditions are present at Sand Island. The Wake Island study also found that significant dilution, driven by vertical groundwater movement from tidal fluctuations, would occur (AFCEE 2006). Therefore, no short- or long-term discharge of brodifacoum-impacted groundwater into the marine environment is anticipated. If large pellet fragments containing brodifacoum do enter the marine environment, they could be ingested by organisms present there. However, the pellets would rapidly disintegrate into

3-15 MIDWAY SEABIRD PROTECTION PROJECT FINAL ENVIRONMENTAL ASSESSMENT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

fragments too small to interest most consumers; these potential impacts are discussed further in Section 3.3.6. In order to best address the potential for brodifacoum contamination of surface water, groundwater, or the marine environment, USFWS and Island Conservation have developed a series of Best Management Practices (BMPs) to minimize this potential. These measures would be used as needed and may include, but are not limited to, the following:

• Hand broadcast in buffer zones around surface water (if they cannot be covered or protected by other means), including seeps and non-beach shorelines where necessary.

• Directional deflectors mounted on the bait buckets during aerial broadcast in areas adjacent to the shoreline.

• Pre and post application sampling of potable water sources using professional collection and lab processes consistent with procedures used at Palmyra by USDA, APHIS NWRC in 2012. (Pitt, et al. 2012)

• Burying artificial seeps, temporarily avoiding or covering or avoiding natural seeps during aerial broadcast. Guzzlers will be inspected and cleaned before being refilled.

3.3.3 NOISE

3.3.3.1 Existing Conditions Sound is measured with instruments that record instantaneous sound levels in decibels (dB). A- weighted sound level measurements (dBA) are used to characterize sound levels that can be sensed by the human ear. “A-weighted” denoted the adjustment of the frequency content of a noise event to represent the way in which the average human ear responds to the noise event. All sound levels analyzed in this EA are A-weighted. Noise levels, which result from multiple, single-events, are used to characterize community noise effects from aircraft operations and are measured in the Day-Night Average (DNL). The noise metric incorporates a “penalty” for nighttime noise events to account for increased annoyance. The DNL metric is the energy-averaged sound level measured over a 24-hour period, with a 10- dB penalty assigned to noise events occurring between 10:00 p.m. and 7:00 a.m. DNL values are obtained by averaging sound exposure level values for a given 24-hour period. DNL is the preferred noise metric of the US Department of Housing and Urban Development, the FAA, the US Environmental Protection Agency, and the Department of Defense for modeling airport environs. Most people are exposed to sound levels of a DNL of 50 to 55 dBA or higher on a daily basis. Noise levels in residential areas vary depending on the housing density and location. As a frame of reference, a normal suburban area is exposed to approximately 55 dBA, increasing to 60 dBA for an urban residential area, and 80 dBA in the downtown section of a city. For Sand Island, MANWR, aircraft arrivals and departures are too infrequent to establish a reliable and accurate DNL metric. Noise impacts for this EA are calculated using Effective Perceived Noise Level (EPNdB). EPNdB is used by the FAA as the noise certification metric for large transport and turbojet aircraft and helicopters. Maximum sound level is important in contrasting

3-16 FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

the interference caused by an aircraft noise event with conversation, sleep, or other common activities. As evidenced by the Hawaiian language name for MANWR —Pihemanu, which means “loud din of birds”—the dominant source of noise on Sand Island is the resident bird population. The extreme concentration of birds present on Sand Island combined with wind noise results in a natural, ambient noise level that exceeds 65 dBA during the day. At night seabirds remain active; thus, while reduced, the sound level remains higher than would normally be found in typical undeveloped areas. The density of human use on Sand Island, MANWR is very low. During normal operations on the Atoll, primary anthropogenic sources of noise include aircraft, engines used to generate power, air compressors, and refrigeration equipment. On relatively still nights when wind noise is limited, the power generation equipment can be heard throughout the inhabited portion of Sand Island when outside. Other sources of transient noise include construction and demolition activity, utility vehicles, and golf carts.

3.3.3.2 Probable Impacts Noise impact analyses typically evaluate potential changes to the existing noise environment that would result from implementation of the proposed action. Potential changes in the acoustic environment can be: • Beneficial, if they reduce the number of sensitive receptors exposed to unacceptable noise levels or reduce the ambient sound level; • Negligible, if the total number of sensitive receptors to unacceptable noise levels is essentially unchanged; or • Adverse, if they result in increased sound exposure to unacceptable noise levels or ultimately increase the ambient sound level. Projected noise effects were evaluated quantitatively for the proposed action. Excessive noise can cause annoyance or irritation. Noise annoyance is defined by the EPA as any negative subjective reaction to noise by an individual or group. Aircraft noise effects can be described according to 2 categories: annoyance and human health concerns, such as hearing loss and sleep disturbance. Annoyance, which is based on perception, represents the primary effect associated with short-term aircraft noise. EPNdB is an acceptable unit for quantifying community annoyance to general environmental noise, including aircraft. Under the proposed action, up to 2 helicopters would be used to aerially disperse bait across Sand Island, MANWR. While no final decision has yet been made as to the type of helicopter which would be used, most helicopters of the size and range required for this project would present an EPNdB of between 85 and 91 dB in the vicinity of the aircraft. As a frame of reference, 2 models of helicopter that have been proposed for use in other rodent eradication projects on islands are the Bell 206B Jet Ranger and the Bell 206L4 Long Ranger. Table 3.5 below depicts effective perceived noise levels for these 2 aircraft.

3-17 MIDWAY SEABIRD PROTECTION PROJECT FINAL ENVIRONMENTAL ASSESSMENT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

Table 3.5 Effective Perceived Noise Level of Selected Aircraft Aircraft Model Flyover EPNdB Level Takeoff EPNdB Level Approach EPNdB Level Bell 206B Jet Ranger 85.4 88.7 90.6 Bell 206LR Long Ranger 85.2 88.4 90.7 Source: 15th Airlift Wing USAF (2009)

The use of helicopters to disperse bait would be conducted during daylight hours and would only disperse bait for approximately 2 days per drop for 20 hours each drop. Three applications are planned, resulting in a total of 60 hours of helicopter time. The noise generated by the helicopter would present a noticeable, albeit short term annoyance to the human and animal populations present on Sand Island, MANWR. The contractors operating the helicopter and bait bucket would wear standard protective hearing devices in accordance with Occupational Safety and Health Administration (OSHA) standards. Other operational personnel carrying out the proposed action would wear standard issue ear plugs, to be used at their discretion when in proximity to the aircraft. Other than the aircraft used to transport personnel to and from Sand Island, MANWR and the operation of the helicopter itself, the other aspects of the proposed action such as bait stations and bait dispersal by hand would not produce elevated levels of sound. It is anticipated that the proposed action would cause temporary adverse noise impacts during the bait distribution procedure(s). Once those treatments are complete, there would be no significant or lasting effect on the existing acoustic environment on Sand Island, MANWR. No helicopters would be flying over Eastern Island since this island is mouse-free. The No Action Alternative would not produce any noise and would have no impact on the acoustic environment of the Atoll.

3.3.4 WASTEWATER AND STORM WATER

3.3.4.1 Existing Conditions The potable water system (see Section 3.3.2.3) on Sand Island, MANWR incorporates collections points wherever potable water is used and conveyed to a septic system and leach field located in the central portion of the island between the runways and the housing (see Figure 1.3). The septic system and leach field was first installed in 1998 and is located in an elevated area with no threat of ponding during heavy rainfall. The previous Navy-built drainage system co-mingled wastewater and storm water, and the existing wastewater conveyance system utilizes portions of that older Navy system. Because of this, some storm water continues to be mixed with wastewater. Steps have been taken to reduce the storm water component by disconnecting building downspouts from the system and reducing the hardened surface areas that collect rainfall, allowing for more natural percolation into the ground. There is no storm water drainage system of the type common in urban areas and military basis in Hawai‘i and the mainland. The coral sand that makes up the island is very permeable and storm water quickly infiltrates and becomes groundwater, resulting in essentially no storm water flowing into the ocean on the ground surface. The few exceptions to this are: • At the few locations where hardscape extends to the shoreline, which is generally confined to the inner harbor and the eastern end of Runway 6-24.

3-18 FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

• There are few catch basins and ditches that direct storm water to the pond and sump where fresh water is collected for potable and non-potable uses. This system primarily serves to concentrate storm water for collection, not discharge to the ocean. The system does include “overflow ditches” on the south side of Runway 6-24 so that in the event a storm event overwhelmed the pond and the ability for the coral sand to absorb storm water, excess storm water can flow to the Pacific Ocean. Overflow does not occur during typical years, suggesting that a 10-year or greater storm event is necessary to generate an overflow condition. There are three freshwater seeps on Eastern Island. Because the island is uninhabited, there is no wastewater or storm water infrastructure present.

3.3.4.2 Probable Impacts The proposed action would not have any direct effect on the wastewater and storm water infrastructure. The wastewater system was designed and built when many more people were present on Sand Island and is more than sufficient to handle the increased wastewater flow associated with the increased number of people that would be present on Sand Island during the eradication effort. Similar numbers of people are present on Sand Island during annual bird counts, and no adverse effects to the system have been experienced at those times. The few locations where storm water flow occurs are those areas where rodenticide application is not necessary (i.e., hardscape of runways and the inner harbor) or where rodenticide would be hand broadcast (i.e., portions of non-beach shorelines and the water catchment system between the runways). Furthermore, the application would be conducted during the relatively dry portion of the year when storm events do not typically occur. The three existing freshwater seep ponds on Eastern will be temporarily filled with sand or covered with shade cloth to prevent Laysan ducks from accessing the water and to reduce conditions that foster botulism outbreaks. Fresh water for captive ducks and camping duck caretakers will be brought from Sand Island periodically. Therefore, no impacts are expected to freshwater resources on Eastern Island. Based on these considerations, the potential effects to wastewater and storm water infrastructure, flows, and quality on Sand Island would be less than significant.

3.3.5 HAZARDOUS MATERIALS AND WASTES

3.3.5.1 Existing Conditions Hazardous material is defined by the Comprehensive Environmental Response, Compensation, and Liability Act (CERCLA), as amended by the Superfund Amendments and Reauthorization Act (SARA) and the Toxic Substances Control Act (TACA), as any substance with physical properties of ignitability, corrosivity, reactivity, or toxicity that might cause an increase in mortality, serious irreversible illness, or incapacitating reversible illness, or pose a substantial threat to human health or the environment. Hazardous waste is defined by the Resource Conservation and Recovery Act (RCRA), which was further amended by the Hazardous and Solid Waste Amendments, as any solid, liquid, contained gaseous, or semisolid waste, or any combination of wastes that poses a substantial present or

3-19 MIDWAY SEABIRD PROTECTION PROJECT FINAL ENVIRONMENTAL ASSESSMENT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

potential hazard to human health or the environment. In general, hazardous materials and hazardous wastes include substances that, because of their quantity, concentration, physical, chemical, or infectious characteristics, might present substantial danger to public health or welfare or the environment when released or otherwise improperly managed. Evaluation of hazardous materials and wastes in this EA focuses on the storage, transport, and use of pesticides, fuels, petroleum, oil, other lubricants, and other chemicals. Evaluation may also extend to generation, storage, transportation, and disposal of hazardous wastes when such activity occurs at or near the site of the proposed action. In addition to presenting a threat to humans, improper release of hazardous materials and wastes can threaten the health and wellbeing of wildlife, botanical habitats, soil systems, and water resources. In the event of a release of hazardous materials or wastes, the extent of contamination varies based on the type of soil, topography, and water resources. Sand Island, MANWR, as former Naval Air Facility (NAF) Midway, has undergone hazardous material removal operations, but still contains chemicals of potential concern (COPC). Therefore, the Navy has implemented Land Use Controls (LUC) at the sites to limit exposure to contaminated material. These sites are listed in Table 3.6 below. Environmental investigations (for hazardous material/hazardous waste) were also conducted on Eastern Island. There are no LUCs present on Eastern Island. Several different types of LUCs were established at 4 landfills and 1 asbestos disposal area located on the Atoll. The landfills contain either municipal or bulky solid waste (i.e., construction debris) and the asbestos disposal area contains asbestos roofing materials. The landfills and the disposal area have been closed and covered with 1.5 to 4 ft. (0.5 to 1.2 m) of clean soil. The following activities are prohibited at the landfills and asbestos disposal area: • Excavation or soil disturbance resulting from human activities that could compromise the integrity of the landfill or soil cover; • Changing the intended land use of the landfill or disposal area; • Modifying or altering the landfill in any way that may adversely affect the landfill area or release or expose subsurface waste. The USFWS does conduct some operations in the LUC areas on an ongoing basis, including Verbesina encelioides eradication and annual albatross counts. Any project work conducted on Sand Island, MANWR, including any earthmoving activities with the potential to disturb contaminated soils, require communication with USFWS’ onsite refuge managers and engineers based in Portland, Oregon. All work plans are reviewed and must be checked by the Refuge Manager, and all visitors are given a briefing related to hazardous materials and provided with a map indicated restricted areas, including LUC sites, as part of their orientation upon arrival.

3-20 FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

Table 3.6 LUC Sites on Sand Island, MANWR Site No. Site Name Description Prior to Remediation 01 Old Bulky Waste Landfill (OBWLF) The OBWLF is an artificial peninsula extending from the south side of Sand Island, formed by the dumping of approximately 64,000 yd3 of bulky metal wastes, construction debris, vegetation waste, and scrap metal into the ocean. The ocean-exposed sides of the OBWLF are protected by concrete riprap. The marine and nearshore environment on the northeastern side of the OBWLF includes the location where a tug and a garbage barge were grounded next to the seawall, on the eastern side of the revetment (Ogden 1996). 02 Runway Landfill The Runway Landfill is located in a filled area beyond the east end of Runway 6-24. The landfill is surrounded by trees and is predominantly grass-covered (Ogden 1996). In operation from 1970 to 1997, the Runway Landfill received all non- hazardous, non-bulky municipal solid waste generated on Sand Island. To reduce the volume of debris, the waste placed in the landfill was burned and covered with soil. 04 New Asbestos Disposal Area The New Asbestos Disposal Area is located near the east end of Runway 6-24 on the east side of the Runway Landfill. The New Asbestos Disposal Area comprises approximately 18,000 ft.2 and contains corrugated asbestos roofing material in plastic bags. 09 Pesticide Storage Area, Former Bldg. The Pesticide Storage Area, Former Bldg. 629, is located near the corner of Morrell and Cannon Avenues. The building 629 was destroyed by fire several years before the EBS (Ogden 1994) was conducted. Records of the types and quantities of pesticides stored in the building were reportedly lost in the fire (Ogden 1997). Pesticide contamination likely occurred at the site due to past pesticide handling practices and possible pesticide spillage during the fire. 20 Old Power Plant, Bldg. 354 The Old Power Plant, Bldg. 354, located in the Superblock Area bound by Nimitz, Branon, Cannon, and Morrell Avenues, was once the main diesel power plant for Sand Island. The Superblock Area consisted of 8 centrally located maintenance and power generation facilities. Bldg. 354 housed diesel generators, oil-filled drums, sumps, switches, transformers, ASTs, and several USTs with evidence of leaks from multiple sources. Cable trenches, production distribution piping, and a storm drain leading to the Inner Harbor were also associated with the site. 34 Pesticide Shop, Bldg. 361 The Pesticide Shop, Bldg. 361, is located near the corner of Nimitz and Branon Avenues. It was formerly used as a transformer substation. Pesticides were mixed and stored at this site, and a sink in the building was reportedly used for pesticide mixing and washing. The facility contained numerous containers of pesticides and herbicides, which appeared to be properly labeled and stored (Ogden 1996). 53 Bldg. 348 Bldg. 348 is located at the southwest corner of the Superblock Area on Sand Island. Bldg. 348 was the former Public Works/Administration Building. Discarded electrical equipment parts observed behind the building were identified as potential PCB sources that may have impacted surface soil (Ogden 1996). n/a New Bulky Waste Landfill, Sand The New BWLF on Sand Island is located near the northern edge of Runway 15-33. Demolition- and removal-action- Island related debris, including concrete, metal debris, PCB-contaminated marine sediment, and asbestos, were placed in this landfill. This landfill also contains a CAMU where stabilized PCB- and pesticide-contaminated soil from various Midway removal actions was placed. Source: Third Five Year CERCLA Review of 12 Sites, Dept of Navy, NAVFAC (2013)

3-21 MIDWAY SEABIRD PROTECTION PROJECT FINAL ENVIRONMENTAL ASSESSMENT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

3.3.5.2 Probable Impacts Effects on hazardous materials or hazardous waste management may be considered adverse if the proposed action results in noncompliance with applicable Federal and state regulations, increases the amounts generated or present on Sand Island over current levels, or overwhelms existing hazardous waste management procedures and capacities. The effects of the proposed action may also be considered adverse if the proposed action disturbs or creates contaminated sites, resulting in adverse effects on human health or the environment. As noted in the previous section, as a former NAF (and ), the Atoll has several hazardous material, waste transfer, and storage areas. These existing conditions are the subject of other planning documents, including 5 Year CERCLA Reviews by the Naval Facilities Engineering Command. These project-related hazardous materials are summarized in Table 3.7 below.

Table 3.7 Project Related Hazardous Materials, Waste, and Toxins Substance Use Fuel(s) For use in all operational vehicles, including aircraft. Oils or Lubricants For use in all operational vehicles, including aircraft. Brodifacoum Rodenticide toxin. Source: Planning Solutions, Inc. (2017)

While vehicular fuel and lubricants are important to consider, the primary substance of concern is brodifacoum, a coumarin-based anticoagulant toxin. It is a vertebrate toxicant that acts by interfering with the blood’s ability to form clots, causing sites of even minor tissue damage to bleed continuously. Before brodifacoum can have a measurable physical effect, levels of the toxin in the liver must reach a toxic threshold, which varies widely by species. For a detailed discussion of the potential for adverse biological impacts related to the proposed broadcast of brodifacoum, see Section 3.3.6. Brodifacoum was the most commonly used rodenticide in the United States and was widely available for household use (Erickson and Urban 2004) until 2014 when the EPA prohibited second-generation anticoagulant rodenticides for use in products geared toward consumers. Now they are only registered for the commercial pest control and structural pest control markets. Second-generation anticoagulants registered in the United States include brodificoulm, bromadiolone, difenacoum, and difethialone. It is anticipated that the proposed action of broadcasting brodifacoum bait across Sand Island, MANWR through a combination of aerial, hand, and bait station dispersal methods (see Section 2.3) would have no long-term adverse impact on LUC sites or issues associated with the handling or management of hazardous waste. Operational staff carrying out the eradication project would adhere to all LUCs and other guidelines for vehicle refueling, disposing of hazardous wastes, and managing hazardous materials throughout its implementation. No aspect of the proposed action would require the alteration of the LUCs in force at the present time, or involve physical disturbance of contaminated sites, hazardous materials management, or hazardous waste disposal processes.

3-22 FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

3.3.6 BIOLOGICAL RESOURCES

3.3.6.1 Evaluation Criteria for Effects on Biological Resources The significance of effects on biological resources is based on: (i) the importance (i.e., legal, commercial, recreational, ecological, or scientific) of the resource, (ii) the proportion of the resource that would be affected relative to its occurrence in the region, (iii) the sensitivity of the resource to proposed activities, and (iv) the duration of ecological effects. It is anticipated that the primary result of the Proposed Action to introduce a toxicant to achieve the purpose and need would be a shift in the plant and animal communities and ecosystem processes on MANWR toward a state more representative of its ecosystem before mice were accidentally introduced. Using previous successful studies and similar applications of a toxin to remove mice from other islands, it is anticipated that this shift would include an increase in the breeding success of several species of seabirds on the island. However, there are too many variables that would contribute to the post- eradication ecosystem responses for effective analysis within the scope of this document. The short-term impacts from exposure to brodifacoum or any other rodenticide to individual animals is determined by 2 factors as follows: 1. The toxicity of the compound to that individual. 2. The probability of that individual’s exposure to the compound (Erickson and Urban 2004). The toxicity of a particular compound on an individual animal is often expressed as LD (Lethal Dose). A common value used to express toxicity is “LD50” which means that the dosage (D) of a toxin that is lethal (L) to 50 % of animals of the species in a laboratory test. The EPA has compiled laboratory data on the LD50 quantity of brodifacoum for a number of species. However, due to the difficulty and expense of obtaining extensive laboratory data, the LD50 values for most species remain unknown. Therefore, for the purpose of estimating impacts to wildlife, this document will use the following LD50 values to generalize potential toxicity for birds and mammals respectively using data from Erickson and Urban (2004) and USEPA (1991):

-6 For birds, an LD50 value of 4.1 x 10 ounces/pound (0.26 mg/kilogram) of body weight for bait containing brodifacoum at a concentration of 25 ppm will be used – this is the average LD50 value for the mallard (Anas platyrhynchos).

-6 For mammals, an LD50 value of 6.4 x 10 ounces/pound (0.4 mg/kilogram) of body weight for bait containing brodifacoum at a concentration of 25 ppm will be used – this is the average LD50 value for the laboratory rat (R. norvegicus). This toxicity model assumes that an animal’s body mass is the primary determinant of how much brodifacoum is required for that animal to reach an LD50 threshold, within each taxonomic category (in this case, birds and mammals). However, there are other variables that affect LD50 as well, but using LD50 values such as those above decreases the possibility that the model would underestimate the risk or potential impact to each species considered. The LD50 value for mallards used for the assessment of toxicity to birds for the proposed action is 4.1 x 10-6 ounces/pound (0.26 mg/kg) (Ross et al. 1980, USEPA 1991). Erickson and Urban (2004) use a similar model to determine the amount of bait needed to reach an LD50 threshold for birds with a mass of 0.88 ounces (25 grams), 3.53 ounces (100 grams), and

3-23 MIDWAY SEABIRD PROTECTION PROJECT FINAL ENVIRONMENTAL ASSESSMENT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

35.3 ounces (1,000 grams), compared to the average daily food intakes for each of these size classes (Table 3.8). However, it should also be noted that measured LD50 values are highly variable within a single species (Mineau 2018).

Table 3.8 Generalized proportion of daily food intake that must be bait for birds to reach an LD50 threshold where the dosage (D) of a toxin is lethal (L) to 50% of animals of the species in laboratory tests

Body Size Class Amount of Bait for LD50 % of Daily Food Intake 0.88 oz. (25 grams) 0.0092 oz. (0.26 grams) 4.2 3.53 oz. (100 grams) 0.037 oz. (1.04 grams) 10.8 35.3 oz. (1,000 grams) 0.37 oz. (10.4 grams) 19.2 Source: Adapted from Erickson and Urban (2004), using a brodifacoum concentration of 25 ppm.

Erickson and Urban (2004) use a similar model to determine the amount of bait needed to reach an LD50 threshold for mammals, using the same size classes as Table 3.8 above. Toxicants are also evaluated by their sub-lethal effects on animals. These are represented by metrics, such as NOAEL (no observable adverse effect level) and LOAEL (lowest observable adverse effect level). NOAEL is a dose or exposure level of a toxicant that produces no measurable toxic effects on the test group of animals and LOAEL is the lowest dose or exposure level of a toxicant that produces a measurable toxic effect on the test group of animals. Sub-lethal effects observed from anticoagulant exposure may include a variety of mild adverse effects, including prolonged clotting time, internal bleeding, piloerection, lethargy, diarrhea, bloody diarrhea, and/or anorexia (Anderson et al. 2011). The NOAEL value for mallards is <3.2 x 10-6 ounces/pound (<0.20 mg/kg) (EPA 1991).

Although toxicity is most frequently represented as the LD50, there is enough evidence to suggest that laboratory-derived LD50 values may not be appropriate for predicting field safety to non-target species in response to the use of second-generation anticoagulant rodenticides (Mineau 2018). This is because there is evidence in the case of a bio-accumulating substance such as brodifacoum that absorption processes are overwhelmed by large doses given acutely, and that smaller, regular intake such as through the food supply results in a higher proportion of the dose being retained and stored in the liver (Erickson 2004). Even sub-lethal doses can cause clotting, biochemical, and physiological abnormalities - including decreased body weight, increased liver size, increased heart size, and increased kidney size, which could or did cause mortality in laboratory settings. Therefore, there are many indications that LD50 values for rodenticide products have underestimated their true toxicity and may not reflect true risks in the field (Mineau 2018). As discussed above, absorption of the chemical from the gut is rate-limited and a single dose is not equivalent to dietary intake. Therefore, for rare species like the Laysan Duck, we used an approach proposed by Thomas et al. (2011) and improved on by Mineau (2014). This approach provides a relationship between liver residues and the probably of evident coagulopathy at necropsy (Thomas et al. 2011) and uses data on the proportion of a single dose that is retained in the liver over time (Mineau 2014). These two sets of information allow for a calculation of a “field-derived LD50” or more accurately ED50 (dose at which half of the test population will show an effect – not necessarily death) (Mineau 2018). However, to be even more protective at the individual level for a rare or threatened species like the Laysan duck, we used an ED5 (5% of the exposed population will be affected by coagulopathy – not necessarily death) for the final analysis.

3-24 FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

3.3.6.2 Non-Target Species Implementation of the Midway Seabird Protection Project would pose inherent risks to non-target species. These risks are summarized in Table 3.9 below.

Table 3.9 Summary of Risks to Nontarget Species Risk Description Primary Poisoning of nontarget species due to consumption of bait with rodenticide. Secondary Poisoning of nontarget species from consuming primary-contaminated prey (e.g., contaminated mice, crabs, insects, fish, etc.). Tertiary Poisoning of nontarget species from consuming secondary-contaminated prey. Disturbance Risks to non-target species from personnel and equipment conducting eradication (e.g., disturbance to resting monk seals from overflight noise). Crushed Burrows Risks of crushing the burrows of nesting seabirds as ground personnel walk into these areas during the baiting operation. Bird Strike Collisions between flying birds and helicopter during aerial operations. Source: USFWS (2017)

From a toxicological perspective, the risk to nontarget species during an eradication project is a function of the species present on the island and their behavior. Specific factors which require consideration include: (i) toxicological properties, composition, and delivery method of bait; (ii) the susceptibility of those species to the toxin; and (iii) the probability of exposure to the rodenticide by directly consuming it or indirectly by feeding on contaminated prey (Howald et al. 2007). Spatial and temporal risks should be eliminated, minimized, or mitigated to the maximum practicable extent, with due consideration to the species, their , and the population significance. On MANWR, although there is no single time of year during which all species with a potential nontarget concern are absent, there are times of the year when some species that have potential nontarget risks are present in very low numbers. Thus, examining the annual use of the island by each species can be used to identify the greatest risk periods so that they can be avoided (see Figure 3.3 below), along with using the latest population estimates of each species on the refuge. Rodenticide exposure, disturbance impacts, risk of crushing burrows, and bird strike hazards can be minimized by timing the eradication operation to avoid the seabird breeding season, and to target periods when migratory shorebirds are on breeding grounds in Alaska (Wegmann et al. 2014, Howald et al. 2005). Careful timing of the project can also minimize risks to avoid specific behavior or key windows for some species, such as avoiding the monk seal pupping period and avoiding the time period when large numbers of albatross chicks are still present and exposed to bait pellets on the ground. Albatross chicks have been observed to occasionally swallow small items that they pick up and manipulate referred to as pica behavior. “Pica” means to an indiscriminate preference for eating non-food items; in this case, albatross chicks pecking and ingesting rocks, sticks and other foreign objects (See “Pica” in Glossary of Terms). Other minimization and mitigation strategies may include captive holding and releasing animals after the risk period passes (e.g., Howald et al. 2005, Wegmann et al. 2014), hazing (scaring away from an area), as well as maintaining the antidote, Vitamin K, on hand if an individual animal is demonstrating signs of toxicosis and can be captured and held for treatment (e.g. Wegmann et al. 2014).

3-25 MIDWAY SEABIRD PROTECTION PROJECT FINAL ENVIRONMENTAL ASSESSMENT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

The Service will strive to implement all reasonable and prudent avoidance, minimization, and mitigation measures to reduce take of birds. The analysis of impacts, along with specific mitigation actions, will take into consideration the potential short-term impacts through mortality or injury of individuals due to mouse eradication activities, and will weigh these against the benefits of long-term species recovery and protection afforded by mouse eradication. Mitigation activities can minimize overall impacts, but in some cases, may not eliminate risks completely. The presence of wildlife on and around islands targeted for eradications present an inherent challenge. In essence, the operation must be able to deliver bait to every mouse on the island while minimizing availability of the rodenticide to other species. In addition to the toxicological risks, the operation may also impose disturbances and habitat alterations that could have negative impacts on the ecosystem. Although impacts to native species are only ecologically significant if they pose a population level effect, as a principle, any risks to wildlife should be avoided, minimized, or mitigated whenever reasonably possible. The long-term benefits of the eradication must outweigh the short-term environmental impacts and mitigation strategies must be considered in the trade-off framework (Broome et al. 2014).

3-26 FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

Figure 3.3 Monthly Species Abundance or Breeding Activity on Sand Island, MANWR

Source: USFWS (2017)

Although there are no species endemic to MANWR, the refuge is of high significance for several terrestrial and marine species, including:

• The critically endangered (non-migratory, resident) Laysan duck, extant on Sand and Eastern Island;

• The endangered short-tailed albatross that has bred in the recent past on Eastern Island and recently is represented by one pair of birds attending a site on Sand Island;

• High density breeding populations of Laysan albatross and black-footed albatross present for about 8-9 months of the year;

• Globally important populations of Bonin petrels and red-tailed tropicbirds;

• A resident population of spinner dolphins (Stenella longirostris) in the lagoon;

3-27 MIDWAY SEABIRD PROTECTION PROJECT FINAL ENVIRONMENTAL ASSESSMENT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

• Overwintering shorebirds, such as the bristle-thighed curlew (Numenius tahitiensis), a species of special concern among the shorebirds at Midway due to its small world population size;

• Threatened Hawaiian green sea turtles that use beaches for resting and nesting and Endangered hawksbill sea turtles;

• Endangered monk seals that use the beaches and foreshore vegetation for resting and pupping, and can be found on the island year-round;

• Two endangered plants present on Sand Island, the Pōpolo (Solanum nelsonii), a trailing shrub, and the Nihoa fan palm (Pritchardia remota), or Loulu. A general risk assessment to each species was developed to examine the likelihood of effects to the species from primary and secondary rodenticide exposure, risk of collision with the helicopter, risk of disturbance from the helicopter and ground personnel, risk of crushing burrows, risk of a ship strike for listed marine species, and significance of the risk relative to the global population of the species (see Table 3.10).

3.3.6.3 Environmental Fate of Brodifacoum in Soil and Water The chemical compound brodifacoum, the active ingredient in the Brodifacoum-25D bait, has low or extremely low solubility in water and bind tightly to organic matter in soil where the rodenticide would be degraded by soil micro- organisms and exposure to oxygen and sunlight. The solubility in water at 20°C is 0.0038 mgl-1, which is considered low. The typical persistence time (DT50) is listed as 84 days which is considered moderately persistent. The half-life in soil is ~84 to 175 days for brodifacoum. The rate of microbial degradation would be dependent on climatic factors such as temperature, light, humidity, and the presence of molds and soil microbes that potentiate degradation. Therefore, in general, degradation time would be more rapid in warm sunny places like NWHI than in colder climates (Eason and Wickstrom 2001, Eisemann and Swift 2006). Bait trials for the eradication of rats on the island of Lehua, Hawai’i indicated inert (placebo) pellets of a composition similar to the inert matrix of Brodifacoum-25D, in the terrestrial environment would break down and be undetectable in 35-40 days when under vegetation and around 65 days on rock or bare ground (Mazurek 2015). The results of a degradation study in soil indicate that only minor metabolites are formed (<3.5% of parent compound). In rats, no toxicologically relevant metabolites have been identified which could be introduced in soil via urine or feces (European Parliament and Council 2010). Because of the rapid breakdown of the rodenticide, the Preferred Alternative would be unlikely to directly or indirectly cause persistent contamination of soil on Sand Island. The potential presence of brodifacoum in groundwater is discussed in Section 3.3.2.

3-28 FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

Table 3.10 Preliminary Risk Assessment by Species, Consequence, and Potential Mitigation Disturbance Poisoning Risk Risk Global

Species Population Mitigation Measures

Significance

Primary Secondary BASH Risk Ground Air Laysan Albatross Low Low Med Low Low 64% n/a (adults) Black-footed Low Low Med Low Low 22% n/a Albatross (adults) Short-tailed Low Low Low Low Low 0.05% n/a Albatross (adults) For Short-tailed albatross, and only if chick is present from single pair Albatross (chicks; Low Low Med Low Low - attempting to breed on Sand Island - Implement regimen of checking for all spp) and removing bait pellets from around the nest or use bait boxes. Wedge-tailed Ground crews are: (i) to stay on established trails when possible; (ii) Low Low Low High Low 0.20% avoid walking close to nesting birds and exhibiting any actions that flush birds, especially those with eggs and young chicks; (iii) stay alert for bird Bonin petrel Low Low Low High Low < 0.01% burrows and avoid crushing them; (iv) if a burrow is crush in, gently dig out the bird. Black noddy Low Low Med Low Low < 0.01% n/a Brown noddy Low Low Med Low Low 0.60% n/a Gray-backed Low Low Low Low Low - n/a Sooty tern White tern Low Med High Low Low Low n/a Great frigatebird Low Med Med Med Low 0.01% n/a Red-tailed Low Low High Med Low 19% n/a Brown booby Low Low Low Low Low < 0.01% n/a Small shorebirds High High Low Low Low 0.15 - 1.1% n/a

3-29 MIDWAY SEABIRD PROTECTION PROJECT FINAL DRAFT ENVIRONMENTAL ASSESSMENT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

Species Poisoning Risk Disturbance Global Mitigation Measures Risk Population

Significance

Air BASH Risk Ground Primary Secondary Pacific golden Trap as many as is feasible; hold in pens; release after the risk period High Med Low Low Low 0.15-0.83% plover has passed. Trap as many as is feasible with 2 options: (i) hold in aviary (as opposed Bristle-thighed High High Low Low Low 2.42% to pens) and release after the risk period has passed; (ii) clip wing curlew and place on Eastern Island. Live capture a majority of the duck population (as possible) and: (i) hold in captivity on Sand Island and then Eastern Island after operation Laysan duck High High Low High Med > 50% commences until the risk period passes; (ii) clip wing feathers of birds on Sand Island and translocate to Eastern Island; (iii) a combination of the above strategies. Cattle egret Mitigation measures for the Atlantic canary due to its cultural Atlantic canary High High Low Low Low - significance. Trap and hold as many as feasible; release after the risk Common Myna period has passed. Measures to reduce pellets entering water; ground crews maintain a 100 ft. (30.5 m) buffer from Hawaiian monk seals; helicopters to avoid Hawaiian Monk Low Low n/a Low Low 5% hovering near seal basking and pupping areas, and to minimize seal distribution of pellets over seals on the beaches to the extent possible without sacrificing the chance of project success. Dolphin spp. Low Low n/a n/a Low - n/a Measures to reduce pellets entering water; ground crews maintain a 100 ft. (30.5 m) buffer from sea turtles; helicopters to avoid hovering Green sea turtle Low Low Low Low Low - near turtle basking areas and to minimize distribution of pellets over turtles on the beaches to the extent possible without sacrificing the chance of project success. Staff will be trained to recognize pōpolo and loulu plants, and to Pōpolo and Nihoa monitor their work areas at all times for the presence of these two fan palm (Loulu) n/a n/a n/a Low n/a - species. They will exercise extreme care when hand broadcast of bait is necessary or when servicing bait stations to minimize any damage to listed plants. Source: USFWS (2017)

3-30 FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

The amount of time bait pellets will be available on Sand Island plays a major role in risk of exposure to non-target species. Bait pellet longevity is influenced by many factors such as environmental conditions and bait uptake by mice and non-target species. Bait longevity may be high in some areas of Midway Atoll due to other factors (USFWS, Unpublished). Besides uptake by target species and exposure to environmental factors, bait pellet disappearance rates are also influenced by uptake by invertebrates like land crabs and cockroaches. Bait disappearance, in these cases, would vary depending upon the abundance and distribution of these invertebrates. While not driven by land crabs, the bait trials on Midway suggest that bait disappeared at high rates in some habitats (USFWS, Unpublished). Bait disappearance in some areas on Midway Atoll may be just as high as that found after bait operations for the eradication of rats on Palmyra Atoll, even though Midway Atoll has much lower populations of land crabs. The reasons for this occurring on Midway are unclear, but one possible explanation is that pellet disappearance on Midway was driven by cockroaches and other invertebrates. Moreover, the bait density proposed in this EA are based on data collected at Midway Atoll and tailored to Midway and its unique ecosystem. Bait pellets may inadvertently fall into the ocean, but small amounts entering the ocean is not considered a great risk. This is illustrated from monitoring conducted on 2 rodent eradication projects using Brodifacoum-25D. On Anacapa Island, divers and land-based observers monitored bait for entry at 7 separate locations (Howald et al. 2010). Sites were selected based on their probability of bait entering the water (e.g., near or under steep cliffs). The application rate on the project was 13.4 lbs./ac. (15 kg/ha) and no bait was observed to directly enter the water, though small quantities indirectly entered at 3 locations, and densities were estimated at 0.15 pellets/10.8 ft2 (1 m2). On Isabel Island, , where the application rate was 18.4 lbs./ac. (20.6 kg/ ha), divers monitoring the operation documented bait in the sub-littoral zone at <1 pellet/108 ft2 (10 m2 or 0.1/m2) (Samaniego-Herrera et al. 2014). Fisher et al. (2011) summarized the results of environmental monitoring for brodifacoum residues after rodent eradications in a fenced reserve at Maungatautari, New Zealand and on the offshore islands Little Barrier, Rangitoto and Motutapu, New Zealand. Brodifacoum was not detected in extensive fresh water monitoring at Maungatautari, or in fresh water samples from Little Barrier Island. Residual concentrations were present in soil samples from underneath degrading bait pellets on Little Barrier and decreased to near the limit of detection 100 days after application. No brodifacoum was detected in marine sampled from Little Barrier, Rangitoto or Motutapu. However, examples exist where large quantities of bait entered the marine environment. During the Palmyra rat eradication, where the documented bait application was high, 75.6 lbs./ac. (84.8 kg/ha) for the 1st application and 71.5 lbs./ac. (80.1 kg/ha) for the second application, the average density of bait entering the water was as high as 40 lbs./ac. (44.7 kg/ha) during the first application and 41 lbs./ac. (46.3 kg/ha) during the second (Engeman et al. 2013). A variety of factors are thought to have contributed to the high quantity of bait entering the marine environment at Palmyra, which include: an irregular coastline, baits drifting in the wind, pilot difficulty locating the shoreline due to overhanging palm trees, and an ineffective and broken bait deflector. The shoreline of Sand Island is much more regular, there are no trees overhanging shorelines or beaches with which the pilot must contend, some of the beaches are wide and bare of vegetation so that the pilot would have a clear view of the high tide line. However, the portion of the island with a sea wall has an abrupt edge between land and sea. Complete breakdown of pellets in the water would be quick, especially for the exposed shorelines on the southern side of Sand Island. For the northern

3-31 MIDWAY SEABIRD PROTECTION PROJECT FINAL ENVIRONMENTAL ASSESSMENT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

shorelines adjacent to the lagoon and protected by coral reefs, a channel was dredged by the military between Sand Island and Eastern Island into the lagoon, which allows a significantly larger tidal flux in and out of the lagoon and harbor than would otherwise occur. During the inert bait trials on Lehua in 2015, data collected show that pellets disintegrated within 30 minutes after application to seawater, and no pellets were found after 24 hours (Mazurek 2015). In trials on Kapiti Island, New Zealand, inert bait pellets were seen to disintegrate within 15 minutes (Empson and Miskelly 1999) and on Isabel Island, pellets “sank immediately and disintegrated by wave action within a few minutes” (Samaniego-Herrera et al. 2014). During a rat eradication project on Anacapa Island in southern California, bait that entered the ocean completely dissolved within 5 hours (Howald et al. 2010). Sampling of seawater 24- and 48-hours post-application, conducted in conjunction with the Anacapa project, tested negative for brodifacoum residues. In the process of breaking down, pellets would consist of suspended cereal grain flocculants and dissolution of the active ingredients into seawater. The solubility of brodifacoum in water is 3.2 x 10-5 oz./gallon (0.24 mg/l) (USEPA 1991). Any effect of salt water on solubility has not been reported. For the rat eradication on Mōkapu, Hawaiʻi and the first attempt on Lehua, seawater samples were collected 5 and 7 days after the last application of rodenticide baits. No diphacinone was detected in the seawater samples from either operation (Gale et al. 2008, Orazio et al. 2009). Because very little of the rodenticide is expected to enter the water, its low solubility in water, and the rapid breakdown in water, the Preferred Alternative would be unlikely to directly impact seawater on or around Sand Island.

3.3.6.4 Terrestrial Environment

3.3.6.4.1 Terrestrial Vegetation: Existing Conditions The first vascular plant inventory was conducted on Midway Atoll in 1904 by William Bryan who documented 13 species (Bryan 1905). Since that time, over a dozen visits have been made by botanists to the remote islands for the purpose of investigating the plant life occurring there (e.g., Christophersen and Caum 1931, Apfelbaum et al. 1983, Bruegmann 1998, Conant et al. 1983, Neff and DuMont 1955). In all, 389 species have been recorded, 350 of which are exotic (see Figure 3.4 below). Introductions of exotic species to Midway have occurred via a variety of means, beginning with the planting of gardens to provide food for residents of the Pacific Commercial Cable Station during the first decade of the 20th century. Other species were brought in accidentally. Not all plant species introduced to Midway have persisted or thrived. Less than half of the exotic plants once observed on Midway, 170 species, occur there today. Regular botanical inventories have been conducted at MANWR since 1999 (Starr and Martz 1999, Starr and Starr 2008, Starr and Starr 2015). As with earlier investigators, the bulk of the information produced has consisted of lists of the plant species observed, collection of voucher specimens for new occurrences, and locations of some of the plants observed with qualitative notes on distribution and abundance.

3-32 FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

Figure 3.4 Map of Vegetative Cover on Sand Island, MANWR

Source: USFWS, Draft Midway Atoll Habitat Management Plan (2014)

Annual weed monitoring has been conducted on MANWR since 2012 and has focused almost exclusively on detecting trends in cover of golden crown-beard (Verbesina encelioides), a highly invasive plant species that has been the subject of a refuge-wide eradication campaign. Past monitoring was conducted at 100 sites (50 sites each on Sand and Eastern islands) and entailed visually estimating the foliar cover of golden crown-beard within a 15.4 ft (4.7 m) radius plot. This monitoring effort served the refuge well in documenting the steep decline of golden crown-beard, which decreased from approximately 50% cover to <1% across a 5-year period.

3.3.6.4.2 Dominant Vegetation Types The land cover classes from the Draft Midway Atoll Habitat Management Plan offer a snapshot of vegetation communities and plant associations on MANWR in December 2014 (see Figure 3.4 and Table 3.11 below). Spatially, the classification includes areas managed for human use, such as the “town” area, as well as areas primarily managed for habitat. The classes are arranged based on canopy type, starting with the highest tree canopy (forest) and ending with unvegetated classes. When more than 1 canopy layer is present, the highest layer determines the cover class. For example, woodland includes a sparse (<25%) tree canopy layer and a field layer, but in some areas a secondary tree canopy and shrub layer are also present.

3-33 MIDWAY SEABIRD PROTECTION PROJECT FINAL ENVIRONMENTAL ASSESSMENT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

Table 3.11 Vegetation Communities and Plant Associations on Sand Island, MANWR FOREST: Casuarina Forest Association with overall tree canopy cover >25%. Tree canopy may range from open to closed. Casuarina seedlings are rarely present. The dominant species is Casuarina equisetifolia, although Casuarina glauca occurs in a few areas. Formerly, Pluchea carolinensis formed a sub-canopy in some areas, but P. carolinensis has largely been eliminated except for seedlings in areas where Casuarina has been removed. Adjacent to beaches, a thin shrub layer of Scaevola taccada may encroach a short distance into the forest. There is often no ground cover. When present, ground cover species include Verbesina encelioides and Euphorbia cyathophora. Euphorbia peplus, Coronopus didymus or Eustachys petraea may occur in forest openings. In areas underlain by pavement or cement, there are few petrel burrows; organic matter, needles and organic debris form the ground cover. Here, the trees are easily wind-thrown and there may be a tangle of downed trees. In dunes and sandy soils, petrel burrows abound and the soil shows weakly stratified layers of sand and organic duff. WOODLAND: Mixed Woodland Association with Casuarina tree canopy cover <25%. Usually distinguished from forest by understory and/or ground cover composition, the mixed woodland association is dominated by large diameter, mature Casuarina equisetifolia or Casuarina glauca trees. Coccoloba uvifera, Hibiscus tiliaceus, Terminalia catappa and various other non-native trees may also occur, forming dense thickets. In combination with Casuarina, these other species may increase overall tree canopy above 25%. Behind beaches, Scaevola taccada may form a discontinuous subcanopy. A range of herbland communities may be present, but common ground cover includes Cynodon dactylon (around the town area), Stenotaphrum secundatum, Boerhavia repens and various other herbaceous species or bare ground. The sandy soil does not have a visible layer of tree detritus or needles. The south half of Sector 4 and many areas around “town” are mixed woodland. The Casuarina forest also intergrades into woodland in some areas. SHRUBLAND: Scaevola Shrub Association with shrub canopy cover >25%. Scaevola taccada dominates the shrub association except in areas bordering beaches, where Tournefortia argentea may co-dominate. Both are considered widespread, halophytic, pioneer species, and they can be found in monotypic stands, or co-dominant and mixed forest (Niering 1963, Kepler and Kepler 1994). Below the canopy, very dense shrub stands eliminate most ground cover, resulting in bare ground covered with dead sticks. In Sector 4, beachside Scaevola encroaches on the neighboring Casuarina forest and woodland. Casuarina seedlings are scattered throughout the shrubland backing North and West Beaches on Sand Island. Eustachys petraea, Verbesina encelioides and an occasional Bidens alba may climb up through the shrubs. Conyza spp. also occur. Scaevola shrub association is found in small patches inland, usually where it has been planted for soil stabilization or cover. A mosaic of shrub and herbaceous associations occurs on the northeastern portion of Eastern Island. On Spit Island, Solanum nelsonii grows in the shrubland. VINES: Viney Perennial Association with >50% cover of viney perennials in the field layer. For landcover mapping purposes, both the Tribulus-Boerhavia Association and the Ipomoea Association are combined into the viney perennial association. Tribulus-Boerhavia Association: Tribulus-Boerhavia is the dominant cover class on Eastern Island. It intergrades with the Lobularia-Cynodon herbaceous association, making classification difficult depending on the season. The invasive mustard Brassica nigra is strongly associated with viney perennials on Eastern Island. Herbicide use in the fall and late winter significantly reduces Tribulus foliage, reducing competition with Lobularia maritima. During these times, the Lobularia-Cynodon association is prominent. Tribulus-Boerhavia vines persist from a woody base, and by early summer, account for >50% cover, masking underlying Lobularia plants. The vines also spread across abandoned runways, forming an important component of the partially vegetated runway class. Species common in this layer include Tribulus cistoides, Boerhavia repens, Lobularia maritima, Brassica nigra, Verbesina encelioides, Coronopus didymus, and occasional patches of Cynodon dactylon and other grasses. Ipomoea Association: >50% cover in the field layer; limited distribution on Sand and Eastern Islands. A few scattered patches of Ipomoea pes-caprae occur on or near beaches, often with Dactyloctenium aegyptium encroachment. Ipomoea indica occurs inland on Sand Island. An aggressive vine, it is usually associated with and overruns the Lobularia-Cynodon association, also occasionally climbing onto Eragrostis variabilis and various trees or shrubs nearby.

3-34 FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

HERBLAND: Lobularia-Cynodon Herbaceous Association where herbaceous vegetation accounts for >75% of cover but may be less, as long as it exceeds woody plant cover. The dominant non-forest cover class on Sand Island, this herbland is composed of forbs and low-growing grasses. Lobularia and Cynodon intergrade throughout the island, with up to 100% cover of Cynodon in the “town” area and 100% Lobularia in other locations. Predominant species include Lobularia maritima, Cynodon dactylon, Solanum americanum, Lepidium virginicum, Coronopus didymus, Eleusine indica and other low-growing non- native forbs and grasses that may change throughout the year. Small patches of the tall grass Stenotaphrum secundatum are included when completely surrounded by herbland. These are generally remnants of former Casuarina forest or woodland that has been removed and converted to herbland. On Eastern Island, the Lobularia- Cynodon Herbaceous Association occurs in low-lying swales backing the beach and shrub zones along the north shore, and around the building foundations and previously developed areas near the pier. Herbs are also an important component of the Tribulus-Boerhavia viney association and seasonally appear to be the dominant cover class. Predominant species include Lobularia maritima, Cynodon dactylon, Tribulus cistoides, Boerhavia repens, Coronopus didymus, Brassica nigra, and Verbesina encelioides. BUNCHGRASS: Eragrostis Bunchgrass Association is generally present in small patches spreading from out-planted nuclei on all 3 islands. Eragrostis variabilis has been the primary species used for restoring native plant cover. In recent years, the grass started seeding and some Eragrostis patches are now expanding on their own. Near beaches, Eragrostis is occasionally overgrown by Scaevola taccada. Other species include Lobularia maritima, Coronopus didymus, Conyza spp, and other herbaceous species. On Eastern Island, Tribulus cistoides may occasionally overrun Eragrostis plantings. Chenopodium oahuense and other native shrubs and forbs are being restored to this community. WETLAND: Wetland Cyperus Association with >50% cover of wetland-associated plants, primarily sedge. Runway 24 east extension on Sand Island is the main expression of this association, but small patches occur in low- lying areas that intermittently flood. The main species is Cyperus polystachyos. Other species include Cyperus laevigatus, Cyperus involucratus, Andropogon virginicus, Sesuvium portulacastrum, Heliotropium procumbens, Phyla nodiflora and Portulaca oleracea. This category usually does not apply to the areas surrounding most of the ponds and former seeps since their soils do not support hydrophytic species. Some facultative wetland species occur at Catchment Pond shoreline and the former Sunrise and Mauka-Makai seeps on Sand Island. BEACH STRAND: Includes herbaceous cover and open beach from the mean tide mark to a shrub belt or herb-covered terrace above the beach. While most of the beaches are unvegetated, a few species occur in patches or at low densities: Lepturus repens, Sesuvium portulacastrum, Ipomoea pes-caprae, and sparse Tournefortia argentea seedlings. Large, dense patches of Sesuvium portulacastrum surrounded by beach, such as the S. portulacastrum/Fimbristylis cymosa on Spit Island, are included in this class. BARREN: Where there is <5% vegetation due to management activities or erosion rather than artificially hardened substrate, excluding beaches. The substrate and transitory nature of barren areas distinguishes this class from the unvegetated class. The barren substrate will support plant growth, whereas the hardened, artificial substrate in the unvegetated class will not until the surface is broken. Barren areas are not vegetated due to recent removal of plant cover, usually for restoration purposes or contaminants remediation. For example, large scale ironwood removal projects result in barren areas until restored to another land cover type, such as an herbaceous plant community. In future landcover maps, barren areas will be classified into the appropriate vegetation association. On Eastern Island, naturally-occurring barren areas occur on the tops of some revetments, however normal erosion processes will eventually result in plant growth. Barren areas also occur in erosive features behind sea walls where vegetative cover is destroyed by waves. Note that barren and unvegetated are somewhat arbitrary distinctions and could be combined into a single unvegetated cover class for mapping purposes. PARTIALLY VEGETATED: Partially Vegetated Former Runway/Pavement/Foundations which are 5-25% covered with vegetation. On Sand Island, common species are Casuarina equisetifolia, Fimbristylis cymosa, Eustachys petraea, Andropogon virginicus, Dactyloctenium aegyptium, Sporobolus indica, Bidens alba, Pseudognaphalium sandwicensium and various herbaceous weeds. Many abandoned runways on Eastern Island are now nearly indistinguishable from off-runway vegetated areas and will usually contain a mix of species from the Tribulus- Boerhavia association as well as a number of incipient weeds.

3-35 MIDWAY SEABIRD PROTECTION PROJECT FINAL ENVIRONMENTAL ASSESSMENT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

UNVEGETATED: Unvegetated Structures/Runways with <5% vegetation. Runways, concrete, asphalt, buildings, piers, structures. Source: USFWS (2017)

Where noted, a few “orphan” plant types, generally remnants from previous land uses, are lumped into the cover class surrounding them or a structurally similar class. For example, the tall grass Andropogon virginicum, an upland species, intermixes with Cyperus polystachyos, the dominant wetland sedge, at the Catchment Pond area. In this case, it is classified as part of the wetland sedge association. Where A. virginicum has spread onto the adjacent abandoned runway, it is classified as “partially vegetated runway.” Some plant associations change dramatically with the seasons, such as viney associations and herblands on Eastern Island. These situations are noted and, while they may affect community composition at various times of the year, the overall effect on habitat structure is negligible. Because management activities and stochastic events create a dynamic system at MANWR, cover classes recognized in previous years have changed and, in some cases, disappeared. No longer is “Verbesina shrubland” a viable cover class, a testament to the success in controlling an invasive weed that caused significant degradation of seabird habitat.

3.3.6.4.3 Endangered Plants Kāmanomano is a short-lived rare perennial grass that has two recognized varieties: Cenchrus agrimonioides var. agrimonioides, found on the islands of O‘ahu, Lāna‘i, and Maui, and C. agrimonioides var. laysanensis, known historically from the Northwestern Hawaiian Islands of Laysan, Kure, and Midway (USFWS 2009b). Populations of C. agrimonioides var. laysanensis were last collected in 1973 on Kure Atoll, Midway Atoll, and Laysan (Wagner et al. 1999, Service 2009). The relatively isolated occurrences of kāmanomano in the northwestern Hawaiian Islands are negatively affected (on the low-lying islands) by nonnative plants and by stochastic events such as tsunami. An assessment by Fortini et al. (2013) concluded that kāmanomano is moderately vulnerable to the impacts of climate change. This plant was federally listed as endangered in 1996. Cenchrus agrimonioides var. laysanensis, a variety of this species endemic to the Northwestern Hawaiian Islands, became extinct in the 1980s (Bruegmann and Caraway 2003). Therefore, the species has not been seen on MANWR for years, even though botanical surveys have been conducted on Sand Island, so it is very likely no longer present here and will not be addressed further. Pōpolo is a sprawling or trailing shrub up to 3 ft (1 m) tall, in the nightshade family (Solanaceae). Pōpolo is a plant species listed as endangered throughout all of its range (USFWS 2016a). No critical habitat has been designated. Typical habitat for this species is coral rubble or sand in coastal sites up to 490 ft (150 m), in the coastal ecosystem (Symon 1999, TNCH 2007, HBMP 2010). Historically, pōpolo was known from the island of Hawai‘i, the island of Ni‘ihau, Nihoa Island, Laysan Island, Pearl and Hermes Reef, and Kure Atoll (Lamoreaux 1963, Clapp et al. 1977, HBMP 2010). Currently, pōpolo occurs in the coastal ecosystem, on the islands of Hawai‘i and Moloka‘i, and on the northwestern Hawaiian Islands of Kure, Midway (on Sand, Eastern and Spit Islands) (Klavitter 2013), Laysan, Pearl, Hermes, and Nihoa (Aruch 2006, in litt.; Rehkemper 2006, in litt.; Tangalin 2006, in litt.; Bio 2008, in litt.; Vanderlip 2011, in litt.; Conry 2012, in litt.; PEPP 2013). Currently at MANWR, there are 915 pōpolo plants on Sand Island, ten plants on Spit Island, and one plant on Eastern Island (USFWS unpublished). Propagation efforts have been extremely successful, and under management out-plantings proliferate (USFWS unpublished).

3-36 FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

The relatively isolated occurrences of pōpolo in the northwestern Hawaiian Islands are negatively affected (on the low-lying islands) by nonnative plants and by stochastic events such as tsunami. Climate change may result in alteration of the environmental conditions and ecosystems that support this species. Pōpolo may be unable to tolerate or respond to changes in temperature and moisture or may be unable to move to areas with more suitable climatic regimes (Fortini et al. 2013). The Nihoa fan palm, or loulu, is a federally endangered species of palm (USFWS 1998) endemic on the island of Nihoa, Hawaiʻi, and later transplanted to the island of Laysan. Loulu occurs in the dry climate of Nihoa but is distributed on the island that suggests the plant thrives in areas of greater water availability because many plants are found in valleys and near freshwater seeps by cliffs (USFWS 1998). It is listed as it is a smaller tree than most other species of Pritchardia, typically reaching only 13-16 ft. tall (4–5 m) tall and with a trunk diameter of 5.9 in. (15 cm). The loulu is a long-lived perennial tree and is among three species that are endemic to the Island of Nihoa that were listed under ESA in 1996. At the time of listing, the plant was limited to two extant populations on Nihoa. loulu stems found near sea level on Oahu suggest the plant was more widespread before lowland habitat was altered for human use (USFWS 1996). Loulu is represented in ex situ collections such as nurseries or arboretums (USFWS 1998, University of 2009) to ensure species viability. Recovery of the Nihoa populations include removing invasive weeds and monitoring seedling growth. Out-planting in other locations such as MANWR also safeguard against extinction. In 2009, approximately 300 seeds were brought to Midway Atoll to plant in the Service’s greenhouse for out-planting within the atoll; a few plants survive on both Sand and Eastern Islands. Currently, only five lo‘ulu plants exist on MANWR with most on Sand island (Goodale, K., pers. comm.). Conservation measures for pōpolo and loulu in MANWR include native plant propagation efforts, fencing or flagging areas of nursery out-plantings, monitoring established out-planted populations, and removing invasive nonnative plants. In 2017, planting and seeding of pōpolo was done in various locations on Sand and Eastern Islands (see Figure 4.1). Out-plantings are done to change the landscape, moving towards more native-dominated (and less nonnative) plant communities. Out-plantings of pōpolo, along with Kāwelu (Eragrostis variabilis), ʻĀweoweo (Chenopodium oahuense), and other native species, is a conservation strategy whereby these plants will take advantage of the open landscape (and slightly less competitive environment).

3.3.6.4.4 Probable Impacts: No Action Alternative Under the no-action alternative, the mouse population would not be eradicated, and the population size would continue to fluctuate within an annual cycle. It is likely that population levels would increase during the rainy season and decline during the dry season. Anecdotal evidence suggests that native plant communities (including the pōpolo and loulu plant species) and vegetation restoration activities on Sand Island would continue to be limited in their productivity from herbivory of seeds and young plants by mice. When collecting seeds from species to plant and grow in the greenhouse for restoration purposes, biologists have noted that seeds from these species are much more abundant on Eastern Island where there are no mice. In addition, mice are eating the seeds and damaging young plants grown in the greenhouse, which are used for the restoration plantings. Adoption of the No Action Alternative would not meet the management objective of restoring the MANWR ecosystem.

3-37 MIDWAY SEABIRD PROTECTION PROJECT FINAL ENVIRONMENTAL ASSESSMENT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

Under the no-action alternative, some vegetation would be exposed to the hand broadcasting of AGRID3 bait pellets for seabird and listed candidate plant protection, in areas where mouse predation of seabirds and mouse damage to listed or candidate plants is detected. (See Section 2.4 “Alternative 2: No Action” detailing current control efforts).

3.3.6.4.5 Probable Impacts: Preferred Alternative Primary Exposure Plants are not known to be susceptible to toxic effects from rodenticides. Brodifacoum is strongly bound on soil particles and is not taken up by plants (WHO 1995). Operational Hazard (Noise Disturbance and Trampling) Helicopters would be taking off and landing from paved staging areas, which are clear of vegetation. Plants, including the two-listed species above, would not be affected by helicopter operations since helicopters would be flying an average of 164 ft. (164 m) above the ground during baiting operations. There would be potential for trampling of some plants as a result of ground- based operations. However, ground-based operations are only expected to take approximately 240 personnel hours to complete over a 3-week period. Hand broadcasted bait will be distributed by a team who systematically walk on parallel transects stopping at predetermined intervals to disperse pellets as evenly as possible by hand. This will only be done in areas not covered by the helicopter bait drop and could include a small area buffer around the water catchment pond located between the runways; 2) a potential narrow strip of land between the coast and the airport runway (at the southeast end of the runway); 3) portions of a narrow sea wall that extents from the north end of the harbor southward and; 4) if necessary, in a narrow buffer around covered seeps and ponds. We will ensure that these technicians are well trained on how to identify loulu and pōpolo and how to work safely in areas that contain the plants. If these species occur in areas where bait will be hand- broadcasted, we will clearly mark them with flagging to alert technicians and avoid incidental trampling. However, given the low number of loulu plants on Sand Island (5 plants, well marked), and the location of pōpolo in restoration areas that will be covered by helicopter, we anticipate impacts to both species will be minimal. A biosecurity plan has been in place on MANWR for many years and any additional bio-security measures would only occur if there was a new incursion of a non-native species. There could also be some additional impact to terrestrial plants during implementation of the monitoring plan for the eradication project. New biosecurity measures enacted could include rapid response to an incursion by non-native invasive species and implementing the project monitoring plan would include follow-up observations and the collection of samples (see Appendix A and B). Potential impacts such as trampling of vegetation could occur with the additional monitoring activities. Given the minimization measures that will be in place to protect endangered plants, we expect any additional impacts to be minor and short-term. Implementing the biosecurity plan is only expected to have beneficial effects to listed plants. Positive indirect effects from the Preferred Alternative would be expected. Removal of mice may result in an increase in the number and diversity of native plants growing on Sand Island, a long- term indirect effect of eradicating mice. Positive effects of the action include the elimination of mice that might eat the seeds and parts of the pōpolo and loulu. The nature of the changes to the

3-38 FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

plant composition would be complex and it is uncertain how native plants would respond compared to non-native plants under potentially increased competition and reduced predation (Eijzenga 2011). It is possible that some species of plants being planted for vegetation restoration projects on Sand Island may spread faster without the negative effects of seed herbivory by mice. The most common invasive introduced taxa on MANWR include ironwood (Casuarina equisetifolia), golden crown-beard (Verbesina enceloides), wild poinsettia (Euphorbia cyanospora), haole koa (Leucaena leucocephala), sweet alyssum (Lobularia maritima), buffalo grass (Stenotaphrum secundatum), peppergrass (Lepidium virginicum), and Bermuda grass (Cynodon dactylon). Mitigation Staff will be trained to recognize pōpolo and loulu plants in the field, and how to work safely in areas that contain these plants; additionally, staff will monitor their work areas at all times for the presence of these two species. If these species occur in areas where bait will be hand-broadcasted, we will clearly mark the locations of plants with flagging to alert technicians and avoid incidental trampling. They will exercise extreme care when hand broadcast of bait is necessary or when servicing bait stations to minimize any damage to listed plants. By implementing these measures, the proposed project will avoid potential adverse effects to the pōpolo and loulu. As part of the project to restore plant communities on Sand Island and eliminate invasive weeds, vegetation restoration sites and invasive weed populations would be monitored post-mouse eradication. Annual weed monitoring has been conducted on Midway since 2012 and would continue, including detecting trends in the cover of golden crown-beard. Any significant changes to the success of these programs would be closely monitored and adaptive management actions taken if needed. All personnel visiting or working on MANWR would adhere to current biosecurity protocols to prevent any new botanical alien species from becoming introduced to Sand Island (Appendix A). In addition, the compressed grain bait pellets are manufactured to ensure that no active seeds are embedded into the baits, thereby preventing accidental introduction. The USFWS has a long-term commitment to successfully restore the MANWR ecosystem and monitor the results.

3.3.6.5 Seabirds MANWR provides a refuge for some 3 million seabirds (USFWS 2014). Twenty-one species are known to breed or roost on Sand Island (see Table 3.12 below). All species of seabirds at MANWR are protected under the MBTA. One species, the short-tailed albatross, is Federally listed as Endangered (Federal Register; July 31, 2000).

3-39 MIDWAY SEABIRD PROTECTION PROJECT FINAL ENVIRONMENTAL ASSESSMENT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

Table 3.12 Breeding Seabirds of MANWR (n = 21 species) Name IUCN-list Federal Status Common Scientific Hawaiian Laysan albatross Phoebastria immutabilis Mōlī Near Threatened - black-footed Phoebastria nigripes ka'upu Near Threatened - albatross short-tailed albatross Phoebastria albatrus - Vulnerable / Endangered Endangered wedge-tailed Puffinus pacificus ‘ua‘u kani Least Concern - shearwater Christmas Puffinus nativitatis - Least Concern - shearwater Bulwer’s petrel Bulweria bulwerii ‘ou Least Concern - Bonin petrel Pterodroma hypoleuca - Least Concern - Tristram’s storm- Hydrobates tristrami - Near Threatened - petrel black noddy Anous minutus Noio Least Concern - brown noddy Anous stolidus noio kōhā Least Concern - sooty tern Onychoprion fuscatus ‘ewa‘ewa Least Concern - little tern Sternula albifrons - Least Concern - least tern Sternula antillarum - Least Concern - gray-backed tern Onychoprion lunatus pakalakala Least Concern - white tern Gygis alba manu-o-ku Least Concern - great frigatebird Fregata minor ‘iwa Least Concern - white-tailed Phaethon lepturus koa‘e kea Least Concern - tropicbird red-tailed tropicbird Phaethon rubricauda koa‘e ‘ula Least Concern - red-footed booby Sula sula ‘ā Least Concern - brown booby Sula leucogaster ‘ā Least Concern - masked booby Sula dactylatra ‘ā Least Concern - Source: USFWS (1983, 2014)

Details on the distribution, population status and trends, ecology, and conservation concerns for these 21 species can be found in the Regional Seabird Conservation Plan, Pacific Region (USFWS 2005) and the International Union on the Conservation of Nature (IUCN) Red List for Endangered Species (IUCN 2017a). The greatest threats to the seabirds of MANWR include introduced and invasive species, fisheries interactions, ocean contaminants, marine debris, and climate change (USFWS 2005).

3.3.6.5.1 Existing Conditions: Albatrosses Three albatross species are found at MANWR, (i) Laysan albatross, (ii) black-footed albatross, and (iii) short-tailed albatross. Adults forage offshore; their diet consists primarily of fish, squid, and crustaceans. Albatross pairs are philopatric and mate retention is high (USFWS 2005). Albatrosses are not agile flyers, due to their long wingspans and heavy body weight. Chicks exhibit pica behavior and occasionally ingest soil, rocks, and paint chips when pecking the ground (Finkelstein et al. 2003). Adult albatross are subject to mortality at sea during interaction with commercial fishing operations. Adults and chicks are subject to mortality when on land during

3-40 FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

periods of low wind when birds die from overheating in the sun. Mortality can also occur from other causes like starvation or injury (USFWS, Unpublished b). Laysan albatross. The global population is estimated at 1.5 to 1.6 million individuals (Arata et al. 2009, 2017b, ACAP 2012b, USFWS, Unpublished b). Over 70% of the global population depends upon MANWR for breeding. Of MANWR’s population, over 60%, or approximately 719,530 birds, nest on Sand Island (USFWS, Unpublished b). Individuals measure about 31.0-32.0 in. (79- 81 cm) in length, have wingspans of 76.8–79.9 in. (195–203 cm), and on average weigh about 5- 8 lbs. (2,200-3,600 g) (Whittow 1993b). The species is present from November to July (USFWS, Unpublished b). Nests are built on bare ground and among vegetation; birds use surrounding grasses, shrubs, and dirt piled into large mounds as a nest cup into which a single egg is laid (USFWS 2005). Black-footed albatross. The global population of this albatross of the northern Pacific Ocean is estimated at 140,138 breeding individuals (Flint 2007, Naughton et al. 2007, ACAP 2012a). Thirty-seven percent of the global population or about 51,230 birds depend upon MANWR, with 22% of the refuge’s population (approximately 30,168 birds) using Sand Island (USFWS, Unpublished b). Individuals measure about 25-29 in (64-74 cm) in length, have wingspans of 76- 85 in (193-216 cm), and on average weigh about 5-9.5 lbs. (2200-4300 g) (Whittow 1993a). The species is present on MANWR from November through June (USFWS 2005). Nests are similar to Laysan albatross nests in construction but occur in more open areas with sparse vegetation; primarily near coastal beach areas (Reynolds et al. 2017). Short-tailed albatross. Federally listed as Endangered, this species historically occurred throughout the North Pacific Ocean (USFWS 2005). Currently, 85% of the global population breeds on Torishima Island in Japan (USFWS 2005). At MANWR, two birds, representing 0.05% of the global population of 4,200 birds (IUCN 2017a), arrived on Sand Island in recent years but have yet to successfully breed (USFWS, Unpublished b). Individuals measure about 84- 33-37 in (94 cm) in length, have wingspans of 84-90 in (213-229 cm), and on average weigh about 8-14.5 lbs. (3,700-6,600 g) (USFWS 2008c). The species is present November through June (USFWS 2008b, c). Nests are similar to Laysan albatross nests in both construction and placement (USFWS 2005). The primary threat to this species is from volcanic activity, which has the potential to destroy or disturb the Torishima Island colony.

3.3.6.5.2 Existing Conditions: Nocturnal seabirds Nocturnal seabirds include: (i) wedge-tailed shearwater, (ii) Christmas shearwater, (iii) Bonin petrel, (iv) Bulwer’s petrel, and (v) Tristram’s storm-petrel. All 5 of these species nest in subterranean burrows or in cryptic rock crevices or rubble; Christmas shearwater also nest underneath trees, shrubs and dense bunch grasses on the ground (Reynolds et al. 2017). These species fly to and from their colonies at night. Impacts to Tristram’s storm-petrel would be low because birds are absent from MANWR during the proposed baiting operation. Impacts to Christmas shearwater and Bulwer’s petrel would be low given that nesting by these species on Sand Island has not been documented to date (USFWS, Unpublished b). Of the 5 nocturnal species, only wedge-tailed shearwaters and Bonin petrels may be negatively affected by the operation. Wedge-tailed shearwater. This wide-ranging flyer is one of the most common seabirds in Hawai‘i. About 10,000 birds nest on Sand Island (USFWS, Unpublished b), representing about 0.2% of the

3-41 MIDWAY SEABIRD PROTECTION PROJECT FINAL ENVIRONMENTAL ASSESSMENT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

global population estimated at around 5,200,000 birds (Brooke 2004). Individuals measure about 16-18 in. (41-46 cm) in length, have wingspans of 38-41 in. (97-104 cm), and on average weigh about 0.86-0.9 lb. (390-404 g) (USFWS 2015a, Gross et al. 1963). Adult arrival is in late March and eggs are laid in June; hatching occurs from late July to late August and chicks fledge mid- October to mid-November (USFWS 2005). This species feeds mostly on fish, but also cephalopods, crustaceans and marine insects, catching prey mainly on the wing by dipping and surface-seizing or pursuit-plunging (IUCN 2017a). Bonin petrel. The Bonin petrel population at MANWR rebounded after rat eradication efforts from 1994-1997 but suffered a set-back from flooding caused by a 2011 tsunami (Pyle and Pyle 2017). The global population of this species is not known but estimated at around one million individuals (IUCN 2017a, Seto and O’Daniel 1999) with most of those occurring on Laysan, Lisianski and Sand Islands. These birds measure 11.8 in (30 cm) in length with wingspans of 24.8-28 in (63-71 cm), and a mean body mass of 0.45 lb. (203.7 g) (Seto and O’Daniel 1999). The nest is a shallow sandy burrow where a single egg is laid in mid-January; chicks fledge by June (Mitchell et al. 2005). This species feeds offshore and surface dips for prey like fish, squid, marine insects, and crustaceans (Seto and O’Daniel 1999). Bonin petrel leave Sand Island in June but return again in August-September (Grant et al. 1983), resulting in a short absence window (USFWS 2005). The proposed eradication period coincides with this period when the lowest numbers of Bonin petrels are expected to be present.

3.3.6.5.3 Existing Conditions: Noddies Both black and brown noddies are present on Sand Island year-round (USFWS, Unpublished b). Both species may be negatively affected by the operation. Black noddy. The global population of black noddy is estimated to be 3-4.5 million birds (IUCN 2017a), of which, 0.003% or about 12,000 birds, occur on Sand Island. Black noddy measure 14- 16 in. (35–40 cm) in length, with a wingspan of 25-28 in. (65–72 cm), and a body mass ranging between 0.2-0.3 lbs. (85–140 g) (Gauger 1999). In Hawai‘i, this species is an asynchronous and aseasonal breeder, meaning that eggs are laid year-round, the synchronous timing of egg-laying among individuals varies, and peak egg-laying occurs in different seasons in different years (USFWS 2005). Nests are built in tree and shrub canopies and 1 egg is laid per breeding attempt (Gauger 1999). Black noddy forage in the surf along beach shoreline (USFWS, Unpublished b), in lagoons of atolls, and brackish water of coastal ponds; most feeding is <6.2 mi. (<10 km) from shore (Gauger 1999). Food items include small fish, squid, and crustaceans that are seized from the surface of the water or just below it. Birds tend to concentrate over nearshore waters. Brown noddy. The global population of brown noddy is estimated to number 180,000-1,100,000 individuals (Delany and Scott 2006), with approximately 7,200 birds occurring on Sand Island, or 0.6% of the global population (USFWS, Unpublished b). Adults are about 16-18 in. (40–45 cm) in length with a body mass of about 0.35 lbs. (180 g) (Chardine and Morris 1996). Brown noddy breed synchronously, with peaks in breeding activity occurring in both the spring and summer months (USFWS 2005). Eggs are present from March through August; a single egg is laid in a nest built on the ground or in a tree. Flight habit is to concentrate over nearshore waters, but mainly feeds offshore on small fish and squid it catches at the sea’s surface by dipping or seizing (Chardine and Morris 1996).

3-42 FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

3.3.6.5.4 Existing Conditions: Terns On MANWR, these species include the: (i) white tern, (ii) little tern, (iii) least tern, (iv) gray-back tern, and (v) sooty tern. Population levels of all 5-tern species, except the white tern, are low on Sand Island, and the sooty tern only nests on Eastern Island. The operation would be unlikely to affect little, least and grey-backed terns because their flight behavior and nest locations make aircraft strikes unlikely. Of the 5 terns, only the white tern may be negatively affected by the operation. White Tern. The common white tern has a distribution across the tropics of the world, found year- round on islands in the south Atlantic Ocean, the , and the western and central Pacific. This species is difficult to census accurately, because of the protracted breeding season, cryptic coloration of eggs and chicks, and inaccessibility of cliff nesting sites. The population at MANWR has increased greatly in past 50 years, owing to introduction of ironwood trees that have provided suitable nesting habitat (Niethammer and Patrick 1998). Population in Northwestern Hawaiian Islands was estimated at 80,700 individuals: 50,840 (63%) non-breeders and 29,860 (37%) breeders (USFWS 1983). The global population of white tern is not known but is thought to be stable (IUCN 2017a). Approximately 49,420 birds occur on Sand Island (USFWS, Unpublished b). As the name implies, this tern is all-white and is 11-13 in. (27.5–33 cm) in length, has a wingspan of 27.6-34.2 in (70–87 cm), and a body mass ranging between 0.17-0.35 lbs. (77–157 g) (Niethammer and Patrick 1998). White tern on Sand Island breed year-round, and most nests are on branches of ironwood (Casuarina equisetifolia) trees or on buildings and other human-made structures at MANWR (Howell 1978). No nest is made; instead, it lays its single egg, often delicately balanced, on a branch, building, or rock. Newly hatched chicks have well-developed feet and claws with which they cling to their perilous nest sites. Adults carry small fish and squid in their bills, often several at a time, to feed chicks. Adults seem tolerant of disturbance, often rearing their young near human activity (Niethammer and Patrick 1998). Most egg-laying occurs February to June (USFWS 2005). White terns forage primarily in the surf along the beach shoreline, along shoals and banks, and offshore for fish and squid (Niethammer and Patrick 1998).

3.3.6.5.5 Existing Conditions: Great Frigatebird The great frigatebird comes to Sand Island not to breed but to roost, and do so in relatively small numbers (USFWS, Unpublished b). The global population is estimated to be between 500,000- 1,000,000 individuals (Gauger et al. 2002, USFWS 2005), with approximately 0.01%, or 140 birds occurring on Sand Island. Individuals are 33.5-41.3 in. (85–105 cm) in. length, have a wingspan of 81-90 in. (205–230 cm), and a body mass of 2.2-4 lbs. (1-1.8 kg) (Gauger et al. 2002). The bird is an aerial acrobat, foraging for and squid at the ocean’s surface over deep, offshore waters (Gauger et al. 2002, USFWS 2005). The great frigatebird may be at some risk during the operation.

3.3.6.5.6 Existing Conditions: Tropicbirds Both red-tailed and white-tailed tropicbirds are present at MANWR. At MANWR, the white-tailed tropicbird is a rare breeder on Sand Island, and represented by only a few pairs (USFWS, Unpublished b). The red-tailed tropicbird is more numerous and would be breeding on Sand Island

3-43 MIDWAY SEABIRD PROTECTION PROJECT FINAL ENVIRONMENTAL ASSESSMENT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

during the operation. Therefore, some negative impact on this species is possible. Both species behaviorally make showy aerial displays around midday at Sand Island, flying 98-164 ft. (30-50 m) in the air (Fleet 1974). Additionally, both species feed far offshore primarily on fish and squid (IUCN 2017a, USFWS 2016b). Red-tailed tropicbird. The global population of the red-tailed tropicbird is estimated to number 60,000-80,000 individuals (Schreiber and Schreiber 2009), with approximately 19%, or 15,000 of birds occurring on Sand Island. The streaming red central rectrices are a striking characteristic for this species that is about 17-18.5 in. (44-47 cm) in body length and has a body mass ranging from 1.4-1.7 lbs. (650-780 g). Red-tailed tropicbirds are aseasonal breeders on Sand Island, and nests occur in a variety of available substrates. Eggs are typically laid on the ground in small depressions in the sand, soil, humus, or fine coral rubble under vegetation, or tucked into cliff crevices (Morrell and Aquilani 2000, USFWS 2005).

3.3.6.5.7 Existing Conditions: Boobies Three booby species, red-footed, brown, and masked, are present at MANWR year-round. There are approximately 475 breeding pairs of red-footed booby on Eastern Island. Some red-footed booby individuals temporarily roost on Sand Island. The masked booby only nests in extremely low numbers on Eastern Island and are too few in number to be affected by the operation. There are generally fewer than 5 masked booby pairs nesting per year on Eastern Island, with no historical information indicating they were previously more numerous (USFWS, Unpublished b). The brown booby, once considered the most common booby at MANWR in the 1930s until rats were introduced, now number only 2 pair nesting on Eastern Island. This species is the only 1 of the 3 that forages in nearshore waters. Given that all 3-species nest on Eastern Island, with low numbers roosting on Sand Island, impacts would be minor.

3.3.6.5.8 Probable Impacts: No Action Alternative Under the No Action Alternative, mice would not be eradicated on Sand Island and would continue to impact seabird colonies and the vegetation they use for nesting and cover. Over 3 million seabirds, encompassing 21 different native species, nest on MANWR’s 3 islands and many of them are susceptible to predation by mice. If mice can prey upon adult Laysan albatrosses, one of the largest seabirds on Sand Island, then other ground nesting and burrowing seabirds, adults and chicks, are at risk from similar impacts. Introduced mice would continue to prey on nesting seabirds on the island, preventing them from reaching their full population potentials, and if the predatory behavior spreads at the rate observed in 2016/2017, then the predation would likely contribute to accelerated declines in affected seabird populations. Brooke et al. (2017), drawing on data regarding eradication of non-native mammals on islands from around the world, examined the population growth rates of 181 seabird populations (of 69 seabird species) following successful mammal eradication projects. This analysis included the recovery of Bonin petrels following the eradication of the black rat at MANWR in 1996. After successful eradication, the median population growth rate (lambda) was 1.119 and seabird populations, with positive growth rates greatly outnumbering those in decline. The study confirmed that invasive mammal eradication is usually followed by growth of seabird populations. Under the No Action Alternative, Sand Island’s mouse population would not be the subject of a targeted eradication project, but mouse control efforts to protect breeding albatross, which were

3-44 FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

started in 2016, would continue and expand as deemed appropriate. (See Section 2.4 “Alternative 2: No Action” detailing current control efforts). Eradication of the house mouse from MANWR would facilitate the protection and restoration of native plants present in the refuge. Beyond the predation of seabirds, mice likely have deleterious effects on many other components of the ecosystem of MANWR. No action would result in the continued presence of mice and a continuing long-term adverse impact on biological resources (see Section 1.2 and Section 1.3). Rodents are known to be a disease vector (de Bruyn et al. 2008, Avenant and Smith 2003), feed opportunistically on plants, and alter the floral communities of island ecosystems (Campbell and Atkinson 2002). They also sometimes compete with native species and degrade the quality of nesting habitat for birds that depend on the vegetation (Wegmann 2009, Young et al. 2010). Rodents can also prey on intertidal invertebrates and reptiles and thus affect their abundance and age structure (Navarrete and Castilla 1993), along with negative effects to marine algae abundance (Kurle et al. 2008). The presence of rodents has also been shown to lower total soil carbon, nitrogen, phosphorous, mineral nitrogen, marine-derived nitrogen, and pH relative to rodent-free islands (Fukami et al. 2006).

3.3.6.5.9 Probable Impacts: Preferred Alternative Primary Exposure The potential for exposure of adult seabirds to rodenticides on Sand Island would be limited. Consumption of bait pellets by adults would not be a risk, as all species are carnivorous, feeding on fish, squid, eggs, or crustaceans. Almost all of the 21 seabird species reported from MANWR (Table 3.12) typically forage many miles away from the island in deep water; the exceptions include black noddy, brown booby and white tern, which are assessed for secondary poisoning in the next section. Primary poisoning would be a potential risk for some species of seabird chicks. Ten species of seabirds have been documented to nest on Sand Island: Laysan albatross, black-footed albatross, short-tailed albatross, wedge-tailed shearwater, Bonin petrel, black noddy, brown noddy, white tern, white-tailed tropicbird, and red-tailed tropicbird. Of these, 4 nest on the surface of the ground and might have young in the nest during the period when bait would be on the ground. These species include the Laysan albatross, short-tailed albatross, brown noddy, and red-tailed tropicbird. The majority of black-footed albatross chicks fledge from Sand Island by late June, prior to when bait would be on the ground. However, only Laysan albatross and short-tailed albatross are at risk of ingesting non-food objects on land. These species nest on the ground, and their chicks are known to peck at or sometimes swallow objects found on the ground such as rocks, sticks, and foreign objects. This tendency is known as pica behavior. This behavior would only be detrimental to the young if they actually pick up a bait pellet and then ingest it. Both the red-tailed tropicbird and brown noddy were nesting on Palmyra during the 2011 rat eradication project and neither species was detected as having ingested brodifacoum after the operation. In addition, there are no descriptions of the red-tailed tropicbird or brown noddy manipulating or swallowing foreign objects on land. In a study of the effects of chemical weapons incineration on a nesting colony of red-tailed tropicbirds on Johnston Atoll in the central Pacific, Schreiber et al. (2001) found that the ingestion of heavy metals from soil was not considered to be a problem since red-tailed tropicbirds did not feed on land and did not tend to pick up objects with their bill while on land. The method of feeding at sea by red-tailed tropicbirds is very different from that of albatross that

3-45 MIDWAY SEABIRD PROTECTION PROJECT FINAL ENVIRONMENTAL ASSESSMENT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES do pick up and sometimes ingest items on land (chicks only) (Work and Smith 1996). Albatross adults sit on the water and pick up potential food items and it is normal for chicks to do this on land instinctively as they learn to use their bill. However, red-tailed tropicbirds dive from the air to feed at sea and thus neither the adults or chicks appear to ingest particles on land and no soil is found in their stomach contents (Ashmole and Ashmole 1967, Schreiber and Schreiber 1993). The likely start date for the bait drop on Sand Island is early July. In a detailed study by Fisher and Fisher (1969) of the nesting chronology of Laysan albatross on MANWR, researchers using study plots found that the number of adult and juvenile Laysan albatross on June 11 was 19% of peak counts in April and only 7.1% of the peak population by June 29. Nestlings begin to leave the colony about June 20 and those that survived were essentially gone by August 2 or 3. In the 3 years of study 10-26% left before July 1. The period of most rapid departure was July 5 to 25, but 50% of the young albatrosses were gone by July 15 in 1961, July 16 in 1962, and by July 9 in 1963 (Fisher and Fisher 1969). Therefore, 50% of young Laysan albatross would likely be gone after the first 2 weeks of the bait drop. In addition, many of the young albatross that do not leave the island by late July often succumb to exposure to the heat and/or starvation, as the adults are no longer returning to feed them. For primary poisoning risks associated with pica behavior, objects within about 3 ft. (1 m) of the nest would be accessible to a Laysan albatross chick. At later ages, however, chicks wander even farther from the nest, to find shade for example. For brodifacoum, an albatross chick would need to ingest about 0.73 oz. (20.8 g), or 21 pellets, for a potential lethal dose (Table 3.13). This is 11.1% of its expected daily food intake. A chick would need to ingest about 0.07 oz. (2 g), or 2 pellets, for a sub-lethal dose. This would be unlikely given that the primary diet of albatross chicks is fish and squid brought by the parents. This latter point of primary diet also applies to chicks that wander further from their nest so that it would be unlikely for them to consume 21 pellets or more via pica behavior. However, as stated in Section 3.3.6.1, there are many indications that LD50 values for rodenticide products have underestimated their true toxicity and may not reflect true risks in the field (Mineau 2018). Regardless of the formula used to assess toxicity of brodifacoum to seabirds, it is clear that the risk of lethal and sublethal primary poisoning to Laysan albatross chicks is very high if pellets are picked up and ingested. For the short-tailed albatross, only 1 pair is present on Sand Island and they have not had an active nest to date. Therefore, it is unlikely an active nest with a chick would be present during the bait drop. However, the pair is monitored each year, and if a chick is present, all bait pellets within reach of this chick would be collected and removed as a mitigation strategy. The above risk of primary exposure for albatross may be overestimated, primarily because most bait pellets from the first application would be rapidly consumed or cached by mice and cockroaches. Subsequent applications would occur after most albatross chicks have fledged. Also, not all birds would pick up the bait, and of those that do, not all would ingest significant amounts because of their main diet of fish and squid provided by the parents. Therefore, it would also be unlikely that any Laysan albatrosses would be killed or sub-lethally affected by primary poisoning from brodifacoum.

3-46 FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

Secondary Exposure While there would be potential for secondary poisoning of black noddy, brown booby, white tern and great frigatebird, that risk would be low. Of the 21 species of seabirds on Sand Island, only the black noddy, brown booby and white tern sometimes forage in the nearshore marine environment around Sand Island. This foraging habit puts these species at risk of potentially ingesting fish contaminated with brodifacoum, and potential exposure to persistent residues post- operation. The great frigatebird could be at risk of secondary poisoning from scavenging dead mice.

Table 3.13 Primary and Secondary Toxicity of Brodifacoum Primary toxicity presented as grams of bait and secondary as grams of fish. “N/A” indicates that there is no feeding behavior for this pathway and no risk from poisoning. The mg of active ingredient is abbreviated as a.i.; LD50 is that the dosage (D) of a toxin that is lethal (L) to 50% of animals of the species in laboratory tests. Brodifacoum Daily %(5) of (4) Food LD50 (mg LD50 2 Daily (2) (3) Body Intake LD50 of LD50 (g Food Species Wt (g) (g)(1) (mg/kg) a.i./bird) (g of bait) of fish) Intake black noddy 110 26.9 0.26 0.029 N/A 24.7 91.7 brown noddy 180 37.4 0.26 0.047 N/A N/A N/A brown booby 1,356 144.7 0.26 0.353 N/A 303.9 210.0 great frigatebird 1,400 147.8 0.26 0.364 N/A 56.4(6) 38.2 white tern 110 26.9 0.26 0.029 N/A 24.7 91.7 Laysan albatross 2000 187.7 0.26 0.520 20.8 N/A 11.1 chick black-footed 3,400 267.9 0.26 0.884 35.4 N/A 13.2 albatross chick short-tailed 6,400 409.3 0.26 1.664 66.5 N/A 16.3 albatross chick red-tailed 660 89.3 0.26 0.172 6.9 N/A 7.7 tropicbird chick Notes: 1. Daily food intake calculated from allometric equation ( Rate = 0.059 x (W 0.67) (USEPA 1995). 2. LD50 of brodifacoum based on mortality of mallard ducklings at the lowest dose, 0.26 mg/kg body weight = 0.00026 mg/g (Ross et al 1980, EPA 1991). 3. This is also the number of bait pellets that would need to be consumed to be lethal to 50% of the animals. 4. Based on mortality of fish at 1.16 mg/kg = 0.00116 mg/g (Pitt et al 2015). 5. Calculated as LD502 grams of bait or prey/daily food intake (g). 6. For the predatory great frigatebird, this is grams of whole rat. Contamination based on highest reported residues in rat carcasses (6.45 mg/kg liver = 0.00645 mg/g) (Pitt et al. 2015).

Black noddy forage within a few yards of the shore and so could be at risk of secondary poisoning if they consume small intoxicated reef fish. Surveys indicate approximately 12,000 birds could be present on Sand Island in July and August. However, black noddy feed on fish at the surface of the water, which are species less likely to consume any bait that falls in the water compared to bottom feeding fish (mullets, etc.). Although it is possible that black noddy could forage on fish exposed to rodenticide, the overall likelihood that this would occur is low. The relatively small quantity of bait that could enter the water and rapid degradation and dispersion of the rodenticide would combine to greatly reduce the probability of a noddy being exposed to an intoxicated marine

3-47 MIDWAY SEABIRD PROTECTION PROJECT FINAL ENVIRONMENTAL ASSESSMENT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

fish or invertebrate. This is supported by results from post-application sampling of the near shore marine environment from 2 eradication projects in Hawaiʻi. The results showed no detectable levels of diphacinone in fish, invertebrates, or seawater (Gale et al. 2008; Orazio et al. 2009). However, during the second 2017 Lehua rat eradication effort, fish were caught from shore and gut contents examined for signs of bait material and the pyranine biomarker fluorescence. Bait material and/or the biomarker were observed in some specimens of fish but not others (see details in Section 3.3.6.14.3). During rat eradication on Palmyra Atoll, the carcasses of 12 seabirds were recovered including 2 sooty terns, 6 red-footed boobies, and 1 black noddy. None of these were confirmed to have died from ingesting brodifacoum, which is noteworthy considering the permitted application rate was 6 times higher than normal due to the rapid consumption of the bait by numerous land crabs (Pitt et al. 2015). In addition, an individual bird the size of a black noddy (110 g) would have to consume 0.9 oz. (24.7 g) of toxic fish to receive a potential lethal dose (Table 3.13) of rodenticide, which is 91.7% of its daily food intake. Because very little of the rodenticide is expected to enter the water, the rapid breakdown in water (see impacts to physical environment section), and the large % of a bird’s daily intake required for a potential lethal dose, the Preferred Alternative would be unlikely to result in the mortality of any black noddy adults or chicks. Brown booby could also be at risk of secondary poisoning since they forage for a variety of fish and squid either at sea or in the surf along the beach shoreline (Schreiber and Norton 2002) and are often observed loafing atop the channel buoys on MANWR. However, only 2 pairs of brown booby nest on Eastern Island and no breeding occurs on Sand Island. In addition, an individual adult booby would have to consume 10.7 oz. (304.0 g) of toxic fish to receive a potential lethal dose (Table 3.13) of rodenticide, which is 210% of its daily food intake. Due to the low number of birds on MANWR, the large % of a bird’s daily intake required for a potential lethal dose, and the low likelihood of encountering toxic fish, the Preferred Alternative would be unlikely to result in the mortality of any brown booby adults or chicks. White tern could be at risk from secondary poisoning since they forage primarily in the surf along the beach shoreline and along shoals and banks as well as offshore for fish and squid (Niethammer and Patrick 1998). However, an individual adult white tern would have to consume 0.87 oz. (24.7 g) of toxic fish to receive a potential lethal dose (Table 3.13) of rodenticide, which is 91.7% of its expected average daily food intake. Because very little of the rodenticide is expected to enter the water, the rapid breakdown of bait pellets in water (see impacts to physical environment section), and the large % of a bird’s daily intake required for a potential lethal dose, the Preferred Alternative would be unlikely to result in the mortality of any white tern adults or chicks. Frigatebird are at risk of secondary poisoning from scavenging dead mice and a maximum of 140 birds have been reported on Sand and Eastern Islands. These birds are known to take unattended seabird chicks, thus there exists potential for scavenging or predation of mice that are dead or dying from ingestion of rodenticide. However, several factors reduce the risk of the frigatebird preying or scavenging on intoxicated mice. First, exposure risk would be relatively low, because most mice dying from anticoagulant rodenticide often do so in areas inaccessible to avian predators (Lindsey and Mosher 1994). Second, while the behavior of foraging on land (i.e, preying on chicks) is documented, it does not appear to be exhibited by all members of the population (Megyesi and Griffin 1996). In addition, based on the body weight of this species and the LD50 value, it would take the consumption of 2.0 oz. (56.4 g) of whole mice contaminated with

3-48 FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

brodifacoum for an individual bird to reach a potential lethal dose. This is 38% of its estimated daily food intake. Therefore, it is extremely unlikely that any individuals would be at risk from secondary poisoning. Based on the above analysis, it is unlikely that black noddy, brown booby, white tern, or great frigatebird would be killed or sub-lethally affected by secondary exposure to brodifacoum from the Preferred Alternative. Operational Hazard (Bird Strike and Noise/Human Disturbance) Crushed Burrows. There is a risk that ground crews that are hand broadcasting bait or conducting monitoring can accidentally collapse active seabird burrows containing adults, eggs, or chicks. There are 4 burrow-nesting seabirds on Sand Island: wedge-tailed shearwater, Bonin petrel, Bulwer’s petrel, and Tristram’s storm-petrel. All 4 species nest in subterranean burrows or cryptic rock crevices, and adults make flights to and from their nests at night. On Sand Island, nesting Tristram’s storm-petrels would not be present during the operation and no nests of Bulwer’s petrel have been documented (USFWS, Unpublished b). Bonin petrel are not expected to be impacted because a majority of these birds would have left Sand Island and few, if any, would be left during the operation. On MANWR, in 1993 and 1994, mean fledging dates for Bonin petrel was 6 Jun ± 5.2 d SD (range 24 May–17 Jun, n = 36) and 4 Jun ± 5.0 d SD (range 26 May–15 Jun, n = 47), respectively (Seto 1994). The main species of concern is the wedge-tailed shearwater, which would be present and breeding during the operation. There are an estimated 5,000 pairs breeding on Sand Island (USFWS, Unpublished b). During the targeted period of July-August, the majority of birds would be at the egg stage and into the beginning of the chick rearing stage of breeding (USFWS 2005, 2015). Hand broadcast application of bait and monitoring on Sand Island would be conducted by personnel skilled in recognizing seabird burrows and trained in mitigation measures. Hand broadcasting is being limited to a few sites such as limited portions of narrow shorelines, ponds and freshwater seeps that cannot be covered or protected in other ways, and around certain structures (see section 2.3.1.2). Bait is typically distributed by a team who systematically walk on parallel transects stopping at predetermined intervals to distribute pellets as evenly as possible. For Sand Island, an estimated 240 total person-hours would be needed to hand broadcast for the entire operation (3 scheduled drops, 8-person team, working 10-hour days). However, after the initial hand broadcast operation, crews would be walking the exact same transect lines for the second and third bait applications. Therefore, it is less likely that any additional burrows would be crushed after the first walk-through. This initial hand broadcast operation is expected to take a total of 80 personnel hours. The success rate of rescuing birds from a crushed burrow is high if mitigation protocols described below are followed. The exceptions are very deep sandy burrows for which it is difficult to find both passageways as a person digs or when a person actually steps on the chamber and crushes the egg or kills the bird. Regardless of whether an adult or chick is pulled out or a damaged burrow is repaired that has an occupant, it would still be counted as a “take” on project monitoring reporting forms. No data is available to gauge exactly how many burrows would be at risk. However, assuming a worst-case scenario, that 1 burrow is crushed for every 3 hours of personnel time, this would result in the risk of 26 burrows being damaged or crushed during the 1st bait drop. However, not all of

3-49 MIDWAY SEABIRD PROTECTION PROJECT FINAL ENVIRONMENTAL ASSESSMENT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

these burrows would be active nest sites and may not be in use. In addition, for active nest burrows, some of the birds would be rescued by having personnel dig out the collapsed burrows. On Lehua, nocturnal shearwaters did not seem bothered by personnel on the ground and reacted by moving a few feet away from the person (USFWS and HDLNR 2017). Given the limited number of sites being hand treated, the low number of people on the ground, the few active shearwater burrows per acre, and the protocol and mitigation strategy in place, there would be a low risk of crushing a large number of wedge-tailed shearwater burrows with adults, eggs, or chicks. Ground crews have been very successful at minimizing impacts to surface and subterranean nests using the additional avoidance and mitigation protocols described below. By adhering to these protocols during ground operations, it is possible that up to 26 shearwater burrows could be collapsed containing an egg, chick, or adult resulting in injury or death. Due to the low number of anticipated mortalities relative to their population size on Sand Island, direct impacts would be minor. Effects would be detectable, but localized, small, and of little consequence shearwater populations on MANWR or globally. We expect there to be some additional impact to burrow nesting seabirds should the need arise to implement the rapid response portion of the refuge’s biosecurity plan and during implementation of the project monitoring plan. A biosecurity plan has been in place on MANWR for many years and any additional bio-security measures would only occur if there was a new incursion of a non- native species. Biosecurity measures enacted would include rapid response to an incursion by non- native invasive species and implementing the monitoring plan would include follow-up observations and sample collecting (see Appendix A and B). Potential impacts such as trampling of burrows and crushing birds or eggs in the burrows could occur along with additional human disturbance. Given the minimization measures that will be in place to protect burrow nesting seabirds, we expect any additional impacts to be minor, localized, and short-term. Implementing the biosecurity plan is only expected to have beneficial effects to seabirds. Mitigation Measures for Burrow Nesting Seabirds Ground-based personnel would be instructed to avoid walking over known wedge-tailed shearwater burrows or other visible burrows. If a burrow is accidentally collapsed by personnel, it would be excavated to re-open the nest entrance to allow adults access to chicks. Burrows would be rebuilt as best as possible to provide chicks or eggs protection from the elements. To avoid impacting nest burrows, biologists would, whenever possible, stay on established trails. Airstrike Hazard Bird airstrike data from airplanes landing and taking off from Henderson Airfield on Sand Island were summarized from 2004 to 2010 (USFWS, Unpublished b). Albatross were the greatest hazard to aircraft movements at Midway because of their body sizes (wingspan approximately 6.5 ft. or 2m), abundance, and lack of maneuverability. Albatross, frigatebird and booby would be the most dangerous to aircraft, because of their large body masses and wingspan. Birds with wingspans of 6.5 ft. (2 m) are large and may be difficult to avoid (Klavitter 2004). To minimize bird strikes, all aircraft are now required to land and takeoff during darkness when albatross are present at Midway from November to July. This can sometimes cause nocturnal seabirds to be struck and killed. However, the practice of landing at night has been an effective measure to mitigate albatross and other bird strikes, since albatross typically do not fly during hours of darkness. However, in

3-50 FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES emergencies, and at other times of the year, aircraft do sometimes land at Henderson Field during the day. In addition to albatross, approximately 900,000 Bonin petrels (wingspan about 18 in. or 46 cm) are found at Midway from August to May. The petrel is most active at night, but typically have lower numbers around the airfield (Klavitter 2004). All species of seabirds on MANWR are diurnal, except for those listed in Section 3.3.6.5.2 (Existing Conditions: Nocturnal seabirds). From 2004 to 2010 there were a total of 59 bird strikes during 1,280 aircraft movements at Henderson Field. This is 1 strike for every 21.7 movements at Midway, or 460 strikes per 10,000 aircraft movements. This strike rate is 81 times that of commercial airlines, which as of 2010 was 5.7 strikes per 10,000 movements (Klavitter 2004). The higher strike rate at Henderson Field is due to the high populations of nesting seabirds on Midway compared to any other airfield in the nation. Sixty-three % of the strikes were from Laysan and black-footed albatrosses (Table 3.14). Species contributing to more than 10% of the total strikes also included Bonin petrel and brown noddy (Table 3.14). Two Department of Defense helicopter flights occurred on Midway during this period and no strikes were observed to these species.

Table 3.14 Number and Species of Birds Reported Struck by Airplanes at Henderson Field, MANWR: 2004-2010 Species Number Struck % of Total Reported Strikes Black-footed albatross 7 11.9 Bonin petrel 6 10.2 brown noddy 6 10.2 great frigate 1 1.7 Laysan albatross 30 50.8 Unknown 3 5.1 white tern 5 8.5 yellow canary 1 1.7 Totals 59 100 Source: USFWS 2017, unpublished.

Timing of the eradication on Sand Island in the July-August period would place it outside of the peak breeding season for most seabirds, limiting collision risks to these species. Operations during the months of July and August would also occur before the majority of shorebirds arrive on MANWR, thus minimizing bird airstrike hazard (BASH) to the helicopters and risks to these species. The helicopters broadcasting the rodenticide would fly at speeds ranging from 29–58 mph (46-93 km/hr.) and at an average expected altitude of approximately 164 ft. (50 m) above the ground. The bait hopper would be on a long-line 20-30 ft. (6-9 m) slung below the helicopters. Therefore, the bait hopper would be approximately 134-144 ft. (41-44 m) above the ground during the rodenticide drop. Consequently, except for landings and take-offs on the runway, the helicopters would fly at an average altitude of 164 ft. (50 m) and the bucket would be >134 ft. (41 m) above the ground most of the time. The height of the helicopters would likely be higher when above forested areas or buildings. These numbers are an estimate and may change as needed to complete the baiting operation safely and effectively.

3-51 MIDWAY SEABIRD PROTECTION PROJECT FINAL ENVIRONMENTAL ASSESSMENT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

Due to the significantly lower air speed of the helicopters compared to planes and jets landing at Midway, the risk of bird collisions with a helicopter would be much lower. In addition, all helicopter operations would only occur during daylight, when visibility for birds and the pilot are optimal. Therefore, nocturnal seabirds would not be at risk, which include the wedge-tailed shearwater, Christmas shearwater, Bonin petrel, Bulwer’s petrel, and Tristam’s storm petrel. Further, impact to Tristram’s storm-petrel would be low because birds are mostly absent from MANWR from July to October, a period coinciding with the baiting operation; impacts to Christmas shearwater and Bulwer’s petrel would be low given that nesting by these species on Sand Island has not been documented to date (USFWS, Unpublished b). Although 6 Bonin petrels were struck by aircraft at Henderson Field on Sand Island (Table 3.14), the majority of these birds were struck after 8:00 p.m. or in the early morning hours during night landings. Black-footed albatross would have finished nesting and are unlikely to be present on Sand Island during the baiting operation. Population levels of all 5 tern species, except the white tern, are low on Sand Island, and sooty tern only nest on Eastern Island. Therefore, only the white tern is at a high risk during the operation. At MANWR, the white-tailed tropicbird is a rare breeder on Sand Island and only the red-tailed tropic bird would be at high risk of collisions. Three species of booby are present at MANWR year-round, red-footed booby, masked booby, and brown bobby. The red-footed booby does not nest on Sand Island and the number of total pairs is low, however some of the red-footed booby will temporarily roost on Sand Island. The masked booby and brown booby only nest in extremely low numbers on Eastern Island. Therefore, booby are unlikely to be at risk from the helicopter operations. In addition, with the helicopters flying at an average altitude of 164 ft. (50 m) and the bucket >134 ft. (41 m) above the ground, these heights would allow most birds to flush below the aircraft and temporarily disperse. Great frigatebird are expected to be at risk due to their large body size and habit of soaring above the island although their populations are small on Sand Island. Three bait drops are planned on Sand Island, using 2 helicopters to maximize the chance of completing each drop in a single day, but a reduction to 1 helicopter is possible and the same effort would occur over multiple days. These drops would be spaced 7 to 10 days apart. It is expected each drop would take a total of 20 flight hours or 60 total flight hours for the entire operation. Total flight times could be longer. Using helicopter strike data from the Palmyra rat eradication, and the number of hours of helicopter time used on Palmyra, we calculated the average number of small bodied <1.8 lbs. (<800 g) and large bodied >1.8 lbs. (>800 g) birds killed per hour on Palmyra. Small bodied birds (sooty tern and 2 unidentified birds) were struck at a rate of 0.085 birds per hour while large birds (red-footed booby) were struck at a higher rate of 0.119 birds per hour. We then applied these collision rates to each of the remaining species of seabirds at risk of strike hazard on Sand Island, but to be conservative, we assumed each species in the 2 body size classes was at equal risk to the 2 species killed on Palmyra, thus likely overestimating risk. These included large seabirds (Laysan albatross and great frigatebird) and small seabirds (black noddy, brown noddy, red-tailed tropicbird, and white tern). After calculating the number of individuals at risk of a strike for each species based on the total helicopter time expected on Sand Island, we adjusted the numbers up or down based on the difference in total population numbers of each species of large or small bodied seabirds at Palmyra versus Sand Island. This adjustment thus accounted for the total number of birds expected to be in the air, based on their population size on Sand Island during the baiting period. For a small seabird, if the population of the white tern on

3-52 FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

Sand Island was 1/3 that of the sooty tern population on Palmyra, final mortality estimates were adjusted lower by 1/3. Although Laysan albatross populations are large on Sand Island and potentially a high strike hazard, in a detailed study by Fisher and Fisher (1969) of the nesting chronology of albatross on MANWR, researchers found that as the chicks mature in June and July, adults frequent the island even less often and for shorter periods of time. In the 60-day interval of mid-May to mid-July, when a number of the older young have already fledged, each parent probably visits the island at intervals of about 3 or 4 days, as indicated by the frequency of the feeding of the young. Therefore, after mid-May the parents may make 15 round trips to MANWR, and only spend 3% of the time in the colony. Over the entire year, males and females only spend 18% and 12% of their total time on land, respectively. To rear a chick successfully during the total nesting period of 230 days requires only 160 trips to the colony by both parents combined (Fisher and Fisher 1969). In addition, adult breeding birds start to leave in mid-June, and by the first of August few of them may be seen in the colony. Adult albatross are also known for the brief duration of their visits to feed older chicks. In their study plots, Fisher and Fisher (1969) found the number of adult and juvenile Laysan albatross on June 29 was only 7.1% of the peak population. After August 7, adults of any category are seldom observed on Midway. Therefore, for Laysan albatross, we calculated that only 7.1% of adults would still be present on Sand Island at the start of the operation in early July. During the 2015 aerial bait trials on Lehua, field crews noted that seabirds were not disturbed (e.g., flushing) due to helicopter operations (Mazurek 2015). Seabirds and shorebirds are also somewhat conditioned to planes landing at Midway, with hundreds of landings and take-offs occurring per year. However, air operations would be expected to kill some seabirds. Using the results of our calculation above, based on the data from the eradication operation on Palmyra, there would be likelihood that 6 species of seabirds could be killed from air operations on Sand Island under the Preferred Alternative. This mortality includes a maximum of 15 Laysan albatross, 1 black noddy, 1 brown noddy, 1 red-tailed tropicbird, 2 white terns, and 1 great frigatebird. These numbers represent a worst-case scenario, based on maximum numbers of birds expected to be on Sand Island, and high probabilities of collision assumed with the helicopter. Due to the low number of anticipated mortalities relative to their overall population levels (see Table 3.10), the Preferred Alternative could cause a direct impact, but those impacts would be minor. Effects would be detectable but localized, small and of little consequence to their global populations. There are no specific mitigation measures that would be expected to reduce the amount of the impact from helicopter operations. Eradication of mice would be expected to directly benefit several species of seabirds, including the Laysan albatross, short-tailed albatross, black- footed albatross and other species. Elimination of risk from predatory mice would allow these species to increase their population sizes. The eventual population-level benefits to seabirds from eradication of predatory mice on Sand Island outweigh the small number of potential mortalities that might occur through implementation of the Preferred Alternative.

3-53 MIDWAY SEABIRD PROTECTION PROJECT FINAL ENVIRONMENTAL ASSESSMENT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

3.3.6.6 Shorebirds

3.3.6.6.1 Existing Conditions MANWR supports overwintering, year-round, and vagrant populations of shorebirds (Table 3.15). Species include bristle-thighed curlew, Pacific golden-plover, wandering tattler, ruddy turnstone, and sanderling, and occasionally grey-tailed tattler, long-billed dowitcher, wood sandpiper and others. While some young birds stay at MANWR year-round, the bulk of shorebirds that had spent the summer breeding elsewhere arrive around August/September for the winter, and depart around April/May, making the preferred window to carry out the application of bait on Sand Island in the July/August period (Figure 3.3). Due to their habit of breeding elsewhere, the summer population of shorebirds on MANWR consists of juveniles and non-breeding birds.

Table 3.15 Ducks, shorebirds, and wading birds of MANWR (n = 7 species) Common Name Scientific Name Status, IUCN-list Laysan duck Anas laysanensis Critically endangered ruddy turnstone Arenaria interpres Least concern wandering tattler Tringa incana Least concern Pacific golden plover Pluvialis fulva Least concern bristle-thighed curlew Numenius tahitiensis Vulnerable sanderlings Calidris alba Least concern cattle egret Bulbucus Least concern Source: USFWS (2017)

There are only 4 species of shorebirds present on MANWR in any significant numbers during the summer months of July and August, when the application of rodenticide is expected to occur. These include 4 species of concern, the bristle-thighed curlew, Pacific golden plover, wandering tattler, and ruddy turnstone. The bristle-thighed curlew and Pacific golden-plover are designated as species of high conservation concern in USFWS National and Regional Shorebird Plans, including the U.S. Pacific Islands Regional Shorebird Conservation Plan (Engilis Jr. and Naughton 2004). Both of these species overwinter at Midway, with low numbers of juvenile or non-breeding birds also present through the summer months. All of the shorebird species using MANWR demonstrate a pattern of greater abundance during the winter months, with very few or no individuals remaining throughout the summer when adult birds return to the breeding grounds in the Arctic (USFWS, Unpublished a). More details on the bristle-thighed curlew can be found in Section 3.3.6.7.

3.3.6.6.2 Pacific golden plover This species has been sighted regularly on the shoreline of all 3 islets at MANWR, in the forested areas, and in open, grassy areas (USFWS, Unpublished b). This species has an extremely variable body mass range, about 0.2-0.4 lbs. (100-200+ g) within an annual cycle (Johnson and Connors 2010). These plovers spread out while foraging; they are visual foragers that feed by running and seizing prey. At MANWR, they prefer open spaces, low vegetation fields, roadsides, sandy beaches, and mudflats. Their diet consists primarily of terrestrial invertebrates (and some freshwater and marine invertebrates), and includes berries, leaves, and seeds. Although these plovers do not nest in the tropical Pacific, they may return to the same winter foraging island each year (Pratt et al. 1987). Therefore, breeding birds are absent at MANWR from about May to

3-54 FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

August, but low numbers of juvenile and non-breeding birds are present throughout the summer months (USFWS, Unpublished b). The global population estimates range from 185,000 to 250,000 (Delaney and Scott 2006) with one estimate of 1,000,000 birds (Bamford et al. 2008, Johnson and Connors 2010). The highest documented count of this species on Sand Island from 1993 to 2001 was 1,540 birds, or 0.15- 0.83% of the global population.

3.3.6.6.3 Ruddy turnstone Ruddy turnstone occurs on Sand Island, with a majority on Eastern Island. Breeding birds are absent at MANWR from about May to August, but low numbers of juvenile and non-breeding birds are present throughout the summer months (USFWS, Unpublished b). Individuals measure 8-10 in. (21-26 cm) in length, have a wingspan of 20-22 in (50-57 cm), and a body mass ranging 0.2-0.4 lbs. (84-190 g) (Nettleship 2000). This stocky, orange-legged shorebird forages on sandy and rocky beaches, coral reefs, tidal pools, and mudflats for marine insects, crabs, clams, and mussels. Ruddy turnstone are common throughout the Pacific Islands, but the wintering population in this region is small compared to the global total (Engilis and Naughton 2004). Estimates of wintering populations range from 259,000 to 544,000 birds (Nettleship 2000, Van Gils and Wiersma 1996). The highest documented count of this species on Sand Island from 1993 to 2001 was 880 birds, or 0.16-0.34% of the global population.

3.3.6.6.4 Wandering tattler Wandering tattler occur on Sand Island. Breeding birds are absent at MANWR from about May through July, but low numbers of juvenile and non-breeding birds are present throughout the summer months. Individuals measure 10-12 in. (26–30 cm) in length, with a wingspan of 19-21 in. (50–55 cm), and a mean bill length 1.3-1.7 in. (3.4–4.2 cm) and an average body mass of 0.2- 0.3 lbs. (100–140 g). Wandering tattlers wade while they actively forage, making jerky bobbing movements, and repeatedly return back to the same spot over very short time intervals. They feed on rocky coasts, exposed reefs, sandy beaches, and mudflats (Gill et al. 2002), for crustaceans and marine worms, but they also eat insects (only when breeding) and small vertebrates. The global population is estimated to range from 10,000 to 25,000 individuals (Engilis and Naughton 2004). The highest documented count of this species on Sand Island from 1993 to 2001 was 108 birds, or 0.4-1.1% of the global population.

3.3.6.6.5 Sanderling Sanderlings occur in very low numbers on Sand Island. No individuals of this species were recorded on MANWR during avian surveys conducted in June, July, August and September 2017. Breeding birds are absent at MANWR from about May to September, but low numbers of juvenile and non-breeding birds can be present throughout the summer months. Individuals measure 7-8 in. (18-20 cm) in length with a body mass of 0.09-0.22 lbs. (40–100 g) (Macwhirter et al. 2002). These plump shorebirds are faithful to mudflats, beaches, and open marshes, and are primarily found on sandy stretches of coast lines, the outer reaches of estuaries, and along rocky shores (Van

3-55 MIDWAY SEABIRD PROTECTION PROJECT FINAL ENVIRONMENTAL ASSESSMENT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

Gils and Wiersma 1996). Sanderling feed on small mollusks, marine crustaceans, polychaete worms and insects in adult, larval and pupal stages (e.g. Diptera, Coleoptera, , Hemiptera and Hymenoptera), as well as occasionally fish and carrion. This species is considered of “limited importance” in the Pacific Islands since the vast majority of birds winter in other parts of the world. The global population is unknown (Engilis and Naughton 2004). Other shorebird species occasionally visit MANWR. These species can include the grey-tailed tattler, long billed dowitcher, black-bellied plover, wood sandpiper, and others. Such vagrants for example, can be birds that are blown off their normal migration routes. Vagrants do not breed at MANWR, nor do they use the refuge as annual wintering grounds. Because these 4 shorebird species do not breed at Midway, but use it as non-breeding habitat in the boreal winter, one simple method for reducing the risk to shorebirds during broadcast application of rodenticide bait is to time the dates of application to coincide with the period of lowest shorebird presence on Sand Island, specifically, the summer months of June, July, and August (see Figure 3.3 and Figure 3.5). This timing does not eliminate the exposure potential because in any given year, a certain proportion of the population may not migrate north. The subset of the population that remains on MANWR is made up of younger, non-breeding birds and those that may not be in good enough physical condition to make the journey north.

Figure 3.5 Mean counts for years 1993-2001 and 2017 of bristle-thighed curlew, Pacific golden plover, wandering tattler, and ruddy turnstone combined by month on MANWR

Note: Data were sometimes influenced by pulses of greater numbers that may represent waves of birds coming from further south and stopping to rest or forage on MANWR.

Mean counts of bristle-thighed curlew, Pacific golden plover, wandering tattler, and ruddy turnstone at MANWR were derived from various sources. Data from 1993-2001 compiles counts gleaned from annual systematic surveying efforts, like the Christmas Bird Count, and from literature sources like MANWR trip reports and incidental counts by Refuge staff and visiting biologists. During 2017, monthly shorebird surveys were conducted simultaneously and

3-56 FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES systematically on all 3 islands (Sand, Eastern, and Spit) to count every shorebird and for 3 hours in the morning (0700-1000); observers walk/bike assigned units to ensure complete coverage of each island, and to record time, species, location, and direction of travel if birds are in flight (Goodale 2017). The highest recorded numbers from 1993 to 2001 for each species during June, July, and August, respectively, were bristle-thighed curlew (12, 168, 242); Pacific golden plover (300, 202, 1,071), wandering tattler (12, 6, 67), and ruddy turnstone (186, 355, 633). The average number of each species over June, July, and August during monthly systematic shorebird counts on Sand, Eastern, and Spit Islands in 2017 were 106 bristle-thighed curlew, 115 Pacific golden plover, 297 ruddy turnstone, and 8 wandering tattlers (USFWS, Unpublished b). These numbers exclude surveys conducted on 8/30/17 when numbers for each of the 4 species increased significantly (see Figure 3.6 below) and outside of the time when the operation is scheduled to occur. Bristle-thighed curlew, Pacific golden plover, wandering tattler, and ruddy turnstone are a concern because of their numerical presence at MANWR, their foraging habits, and documentation of their vulnerability during similar operations at ecologically comparable sites (Engilis and Naughton 2004). The Pacific Islands function as an essential migratory habitat for maintaining many global shorebird populations. All of the shorebird species using MANWR demonstrate a seasonal pattern of greater abundance during the fall and winter months beginning in August/September, and very few individuals of some species remain throughout the summer when the breeders are elsewhere (Engilis and Naughton 2004).

3-57 MIDWAY SEABIRD PROTECTION PROJECT FINAL ENVIRONMENTAL ASSESSMENT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

Figure 3.6 Monthly counts on MANWR conducted in 2017 for bristle-thighed curlews (BTCU), Pacific golden plovers (PGPL), wandering tattlers (WATA), and ruddy turnstones (RUTU)

Source: USFWS (2017)

3.3.6.6.6 Probable Impacts: No Action Alternative Under the No Action Alternative, brodifacoum would not be used island-wide to eradicate mice on Sand Island, except for targeted control with the less toxic AGRID3. Since the rodenticide AGRID3 used under the no action alternate is not known to be harmful to shorebirds, there would be no effects to these species under the no-action alternative.

3.3.6.6.7 Probable Impacts: Preferred Alternative Primary and Secondary Exposure As summarized above, the Pacific golden-plover, wandering tattler, and ruddy turnstone are regularly seen on Sand Island. At MANWR during 2017, monthly shorebird surveys have been conducted simultaneously and systematically on all 3 islands (Sand, Eastern, and Spit) to count every shorebird and passerine for 3 hours in the morning (0700-1000); observers walk/bike

3-58 FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES assigned units to ensure complete coverage of each island, and to record time, species, location, and direction of travel if birds are in flight (Goodale 2017). The highest recorded numbers for each of these species during July and August 2017 surveys (excluding surveys conducted on 8/30/17 because the proposed action is not scheduled for this time) were 278 Pacific golden-plover, 351 ruddy turnstone, and 15 wandering tattlers. Significantly higher numbers of each of these species were recorded on surveys conducted on August 30 and September 15, 2017 with maximum counts of 1,419 Pacific golden plover, 880 ruddy turnstone, and 84 wandering tattler (Figure 3.6) (USFWS, Unpublished b). These 3 species primarily consume a wide range of invertebrate and possibly vertebrate prey, which makes secondary poisoning a potential risk. Each is also documented to consume foods comprised of vegetable matter (natural and man-made); however, this is usually a minor component of the diet, making primary poisoning a relatively low risk. Table 3.16 indicates that to ingest a lethal dose of brodifacoum, any of these 3 shorebird species would only need to eat between 0.3 to 0.39 oz. (9-11 g) of contaminated invertebrates on a single day. This is only 35 to 37% of their estimated daily food intake. In addition, the ingestion of 2 pellets of bait (0.07 oz. or 2 g) would constitute a potential lethal dose for an individual of each species. This would amount to only 7% or more of each species daily normal food intake of non- animal matter (Table 3.16). However, as stated in Section 3.3.6.1, there are many indications that LD50 values for rodenticide products have underestimated their true toxicity and may not reflect true risks in the field (Mineau 2018). Regardless of the formula used to assess toxicity of brodifacoum to birds, it is clear that the risk of primary and secondary poisoning to shorebirds is very high.

3-59 MIDWAY SEABIRD PROTECTION PROJECT FINAL ENVIRONMENTAL ASSESSMENT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

Table 3.16 Primary and secondary toxicity of Brodifacoum to shorebirds on MANWR along with the Laysan duck and cattle egret Primary toxicity presented as grams of bait and secondary as grams of invertebrate prey. Mg of active ingredient is abbreviated as a.i. LD50 is that the dosage (D) of a toxin that is lethal (L) to 50% of animals of the species in laboratory tests. Brodifacoum (3) (5) Body LD50 LD50 % of Daily (2) (4) Wt Daily Food LD50 (mg of (g of LD502 Food Intake Species (g) Intake (g)(1) (mg/kg) a.i./bird) bait) (g of prey) (prey) Pacific 110 29.9 0.26 0.029 1.14 9.4 34.9 golden plover ruddy 130 30.1 0.26 0.034 1.35 11.1 36.8 turnstone wandering 120 28.5 0.26 0.031 1.25 10.2 35.9 tattler bristle- 497 73.9 0.26 0.129 5.17 42.4 57.4 thighed curlew cattle egret 370 60.6 0.26 0.096 3.85 14.9(6) 24.6 Laysan duck 500 74.2 0.26 0.130 5.20 42.6 57.5 Notes: 1. Daily food intake calculated from allometric equation (Bird Feeding Rate = 0.059 x (W 0.67) (USEPA 1995). 2. LD50 of brodifacoum based on mortality of mallard ducklings at the lowest dose, 0.26 mg/kg body weight = 0.00026 mg/g (Ross et al. 1980, EPA 1991). 3. This is also the number of bait pellets that would need to be consumed to be lethal to 50% of the animals. 4. Cockroach contamination based on highest reported value of 3.05 mg/kg = 0.00305 mg/g (Pitt et al. 2015). 5. Calculated as LD50 2 grams of prey/daily food intake (g). 6. For the predatory cattle egret, this is grams of whole rat. Contamination based on highest reported residues in rat carcasses (6.45 mg/kg liver = 0.00645 mg/g) (Pitt et al. 2015).

In a study of mortality of shorebirds after the Palmyra rat eradication using brodifacoum, Pitt et al. (2015) found that all bristle-thighed curlew, Pacific golden plover, ruddy turnstone, and the wandering tattler mortalities collected had detectable brodifacoum residues. Six bristle-thighed curlew, 2 Pacific golden plover, 2 ruddy turnstone, and 1 wandering tattler were found dead after the rodenticide application. Bristle-thighed curlews had the highest average brodifacoum residue relative to the other birds analyzed. Results from Palmyra Atoll indicate that non-target of shorebirds are a concern (Pitt et al. 2015). The number of shorebird mortalities from the eradication attempt were likely underestimated because some avian mortalities probably went undiscovered. Although the USFWS (2011) estimated that more than 20,000 rats were killed during the Palmyra eradication, only a few dozen rat carcasses were found during the project. To estimate the total maximum number of shorebirds that may have been killed due to poisoning, Pitt et al. (2015) used the low tide shorebird survey counts conducted prior to the operation (June 3, 2011) and compared these counts to the numbers of birds counted after completion of the bait drop, but prior to the arrival (July 30, 2011) of post-breeding birds in autumn. Maximum estimates of mortality Pacific golden plover, ruddy turnstone, and the wandering tattler using this method were 28, 8 and 10 birds respectively. Using these data, we can calculate the minimum and maximum % mortality of shorebirds from the Palmyra operation. Using 2010 survey data, the maximum population estimates of these 3 species on Palmyra were 62 Pacific golden plover, 35 ruddy turnstone, and 48 wandering tattler. The minimum and estimated maximum of the total population of each shorebird

3-60 FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

species sub-lethally impacted or killed was 3.2-45.2% for Pacific golden plover, 5.7-22.8% for ruddy turnstone and 2.1-20.8% for wandering tattler. Based on the above analysis, after the application of brodifacoum on Sand Island, it would be likely that 9-125 plovers, 20-80 turnstones, and 1-3 tattlers, would be sub-lethally impacted or killed by either primary or secondary poisoning from brodifacoum during the operational window. This estimate assumes that the maximum count of all shorebirds recorded from all 3 islands would all be present on Sand Island, which is unlikely. Mortality estimates for bristle-thighed curlew are presented in Section 3.3.6.7.3 below. Brodifacoum can persist in the environment for 6 months to a year and have impacts beyond the proposed operational window (Rueda et al. 2016; Pitt et al. 2015), such that any returning shorebirds would be at risk from secondary poisoning. Therefore, additional shorebirds could be killed in the months following the application of brodifacoum. Taking the estimated maximum % of each species killed on Palmyra and using the highest count of each species recorded on MANWR since 1993, we calculated the potential additional mortality of shorebirds over the fall and winter after the bait drop. There could be an additional 584 Pacific golden plover, 182 ruddy turnstone, and 17 wandering tattler at risk over the next 12 months after the baiting operation. This is a worst-case scenario based on conservative LD50 toxicity values, maximum expected number of birds on Sand Island, and maximum expected mortality rates. However, because of the high degree of site fidelity for birds on their wintering grounds, it would be likely that mortality of shorebirds from rodenticide poisoning would decline in subsequent months. In addition, some shorebirds would be expected to forage on Eastern Island and Spit Island, and therefore not experience primary or secondary exposure to the rodenticide. Combining the potential estimated mortality from the initial bait drop and over the subsequent year, the total estimated mortality of Pacific golden plover (709), ruddy turnstone (262), and wandering tattler (20) from primary or secondary poisoning is only 0.37%, 0.10%, and 0.2% of their estimated minimum global populations and therefore would not have a measurable effect on those species’ populations. Operational Hazard (Bird Strike and Noise/Human Disturbance) It is unlikely ground operations associated with the Preferred Alternative would impact any shorebirds, as no nesting occurs on the island. Air operations have the potential to disturb shorebirds, but observations from field personnel during pre-project surveys indicate this disturbance would be minor and may cause birds to temporarily flush from their location. No shorebirds were known to be killed from collisions with the helicopter during the Palmyra operation, however, based on the data of small bodied birds (sooty terns and unidentified birds) killed from helicopter collisions on Palmyra (0.085 birds per hour), and using the estimated hours of helicopter time for the Sand Island operation and expected population of shorebirds during the drop, the estimated mortality is less than 1 bird for each of the 3 species of shorebirds. Calculations are based on the maximum number of birds observed in 2017 during the planned drop period. We expect there could be some additional impact to shorebirds should the need arise to implement the rapid response portion of the refuge’s biosecurity plan and during implementation of the project monitoring plan. A biosecurity plan has been in place on MANWR for many years and any additional bio-security measures would only occur if there was a new incursion of a non-native species. Biosecurity measures enacted would include rapid response to an incursion by non-native invasive species and implementing the monitoring plan would include follow-up observations and

3-61 MIDWAY SEABIRD PROTECTION PROJECT FINAL ENVIRONMENTAL ASSESSMENT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

sample collecting (see Appendix A and B). Potential impacts, such as causing shorebirds to flush, could occur from the additional human disturbance. However, we expect any additional impacts to be minor, localized, and short-term. Mitigation Several mitigation options have been proposed that could minimize exposure of shorebirds to bait containing the rodenticide during the eradication campaign on Sand Island. These include: • Timing implementation of the eradication for when the population is lowest before the majority of birds arrive at MANWR from their breeding grounds (July/August); • Keeping bait pellets away from shorebird feeding areas to the extent possible by not broadcasting in intertidal zones; • Trapping as many individuals as possible and keeping them in captivity until risk of exposure is eliminated or greatly reduced. As many shorebirds as possible would be trapped and either kept in captivity or their feathers would be clipped and they would be released on Eastern Island before the rodenticide is applied to Sand Island. The above estimates of mortality do not consider this mitigation action, as shorebirds can be very difficult to capture. Combined potential mortalities from primary and secondary poisoning, along with possible mortalities from airstrikes, represent insignificantly small numbers relative to overall global populations of these species, and represent a worst-case scenario based on maximum numbers of birds documented on Sand Island and high probabilities of collision with the helicopter. Due to the low number of anticipated mortalities relative to their overall global population levels, the Preferred Alternative would be unlikely to adversely impact shorebird populations on MANWR. The impacts would be readily detectable and localized with likely measurable consequences to shorebirds, but not readily detectable or measurable beyond the immediate area of impact.

3.3.6.7 Bristle-thighed curlew

3.3.6.7.1 Existing Conditions The bristle-thighed curlew is a medium-sized curlew with a wingspan of approximately 15.7-17.3 in. (40-44 cm) and blue-gray legs (see Figure 3.7). The birds have a moderately long and decurved bill, which is flesh-colored at the base, turning to brown near the tip, and a dark lateral crown is present on the head along with eye stripes. Their name comes from the bristle-like extensions at the base of their legs, although these are generally inconspicuous. Females are heavier than males and have longer wings and a shorter bill. Mean weight of adults was 1.09 lbs. (497.4 g) and ranged from 0.84-1.29 lbs. (393-584 g) for 61 unsexed birds caught in the Northwestern Hawaiian Islands from October through November (Marks 1993). Juveniles are similar to adults except for the presence of larger cinnamon-buff spots on the upperparts, and virtually unstreaked underparts (Marks et al. 2002).

3-62 FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

Classified as Vulnerable on the IUCN Red List Figure 3.7 Bristle-thighed curlew 2012 (BirdLife International 2016b), bristle- thighed curlew are considered species of high conservation concern because of their small population size and anthropogenic pressures on the wintering grounds, which include habitat loss, habitat degradation, and introduced mammalian predators. The bristle-thighed curlew has also been designated as a Bird of Conservation Concern by the USFWS at the regional and national scale (USFWS 2008b) due to limited breeding and non-breeding distributions, low relative abundance, and threats during the non-breeding season. Although the birds’ tundra breeding grounds have remained largely undisturbed, some researchers believe that bristle-thighed curlew are negatively impacted by introduced mammalian predators, including rats, on their wintering grounds and atolls, including Palmyra Atoll (Marks et al. 1990). The risk of a steep and sudden population decline is a serious threat

Source: USFWS (2017) because these birds are long-lived and site- faithful. Bristle-thighed curlew occur on Sand Island. Breeding birds are absent at MANWR from about May through the end of July, but low numbers of juvenile and non-breeding birds are present throughout the summer months (USFWS, Unpublished b). During the breeding season, bristle-thighed curlew are found in the remote mountainous regions of western Alaska in the Andreafsky Wilderness Area north of the Yukon River mouth and on the central Seward Peninsula (McCaffery and Peltola Jr. 1986, Kessel 1989, Gill et al. 1990, Marks et al. 2002). During the non-breeding season this species is found on remote Pacific Ocean islands and atolls (Marks et al. 1990) including the Hawaiian Islands. Sub-adults may remain in the Pacific until they are nearly 3 years old (Collar et al. 1992). Site fidelity on both the breeding and wintering grounds is high, with many birds returning to the same location on an island for multiple years (Marks and Redmond 1996). Habitat preference in the Northwestern Hawaiian Islands includes beaches and shorelines with coral ledges, but birds are often found inland among grass- and forb- dominated areas (Woodward 1972, Ely and Clapp 1973, Amerson et al. 1974, Clapp and Wirtz II 1975, Marks et al. 2002). In fall of 1988, Laysan Island populations were predominantly inland (68% of 2,521 sightings), and only 1% on beaches (Marks et al. 2002). Birds walk as they pick up items from the ground, also probing in soil or mud with their long bills, to feed on crustaceans, small fish, grubs, and snails. This species has also been observed catching and eating mice. This species has been sighted regularly on the shoreline of all 3 islets at MANWR, in the forested areas, and in open, grassy areas (USFWS, Unpublished b). Comprehensive surveys of known breeding range from 1988 to 1992 yielded about 3,200 breeding pairs (Marks et al. 2002), while Engilis and Naughton (2004) estimate that the global population

3-63 MIDWAY SEABIRD PROTECTION PROJECT FINAL ENVIRONMENTAL ASSESSMENT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

is 10,000 birds and is limited by restricted breeding and non-breeding distributions, a low relative abundance, and threats during the non-breeding season. Numbers may be declining, although data on population trends are not available (Marks et. al. 2002). In total, about 800 birds are thought to winter in the Northwestern Hawaiian Islands during the same period (1980 to 1990), which included 300–350 birds on Laysan Island, 300–400 on Lisianski Island, and 100 on MANWR (Marks and Redmond 1994). A survey has been employed on MANWR to count shorebirds, including bristle-thighed curlew. The age of birds was not recorded because adult and sub-adult curlews can be somewhat difficult to distinguish. However, birds remaining through the boreal summer are most likely sub-adults, which do not migrate back north until they are approximately 34 months old (Marks 1993, Marks and Redmond 1996). Bristle-thighed curlew counts on MANWR from June to September 2017 ranged from 11 to 242 birds, with the maximum number of birds counted on August 30 (Figure 3.6). Although survey methods varied, using avian survey data collected from 1993 to 2001, along with the 2017 surveys, counts from January through December on MANWR ranged from 12 to 242 birds with the lowest counts in June. The maximum counts on MANWR during the expected drop period in July/August was 168 birds in 2017. The highest documented count of this species on Sand Island from 1993 to 2001 was 242 birds, or 2.42% of the global population. Since there are no marked birds, it is not known how many of these birds are winter residents versus migrants stopping before they head further south.

3.3.6.7.2 Probable Impacts: No Action Alternative Under the No Action Alternative, rodenticides would not be used island-wide to eradicate mice on Sand Island, except for targeted control with the less toxic AGRID3. Since the rodenticide AGRID3 used under the No Action Alternate is not known to be harmful to shorebirds, there would be no effects to the bristle-thighed curlew under the No Action Alternative.

3.3.6.7.3 Probable Impacts: Preferred Alternative Primary and Secondary Exposure The mouse eradication project places non-resident migratory shorebirds, including bristle-thighed curlew, at risk of both primary and/or secondary exposure to rodenticide bait that would be broadcast across the island. Research has shown that shorebirds may consume rodenticide bait pellets directly (Pierce et al. 2008) or may sustain secondary or tertiary exposure to rodenticides after ingesting smaller organisms that previously consumed bait. In the non-breeding season, bristle-thighed curlew forage primarily in terrestrial habitats, consuming spiders, land crabs, insects, seabird eggs, , and carrion (Marks 1993). Stomach contents of 14 curlews collected in Polynesia contained vegetation, crustaceans, insects, gastropods, and scorpions (Johnsgard 1981). Given the diverse diet of this species there are several potential secondary/tertiary exposure pathways for the bristle-thighed curlew including: • Feeding on crabs that have consumed bait or scavenged mice carcasses; • Feeding on insects that have consumed bait or scavenged mice carcasses;

3-64 FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

• Feeding on lizards/geckoes that have consumed insects that have ingested the bait; and/or • Feeding directly on live mice or mice carcasses that have consumed bait. Research conducted on Palmyra Atoll using remote cameras indicated bristle-thighed curlew showed little interest in the bait placed on the ground or in the carcasses of rats put out by the researchers (USFWS 2011). Curlew were ever-present throughout the atoll during these studies and were frequently captured in photographs by the motion sensing cameras that were monitoring bait stations, tracking tunnels, and rat carcasses. Though frequently sighted, curlew, plover, wandering tattler, and booby seemed more interested in the camera itself than in the bait pellets or carcasses. Birds were occasionally photographed looking in the direction of a carcass or bait station, yet no close inspection or contact was made. Although it was estimated that more than 20,000 rats were killed during the eradication project on Palmyra, only a few dozen rat carcasses were found during the project after the bait was applied (Pitt et. al. 2015). It is expected that this same result would occur on Sand Island. Thus, it is expected that there will be a few carcasses available for bristle-thighed curlew to find. A June 2011 rat eradication effort on Palmyra Atoll using brodifacoum reported 6 bristle-thighed curlew individuals found dead after the bait drop and Pitt et al. (2015) estimated as many as 68 could have been killed after the application of rodenticide. Using these numbers, 3.3 to 37.4% of the island’s population may have been sub-lethally impacted or killed by either primary or secondary poisoning from brodifacoum during the operational window. However, bait densities were very high for this effort and included 75.6 lbs./ac. (84.8 kg/ha) applied in the first drop and 71.5 lbs./ac. (80.1 kg/ha) applied on the second drop, with a third application in some areas. This is 10-16% higher bait density than that proposed for the MANWR Preferred Alternative. Two of the curlews were confirmed to have been exposed to bait prior to capture and were successfully treated on Palmyra. Other eradication projects reported non-target mortality of bristle-thighed curlew (Pierce et al. 2008). Pierce et al. (2008) reported 2 dead curlew found 13 days after the first spread of brodifacoum bait on Rawaki Island, Republic of Kiribati, in the Central Pacific, while 8 other live curlews were also recorded after the bait drop, which is a mortality rate of 25%. However, surveys for dead or dying shorebirds were not comprehensive for most eradication projects (USFWS 2011). It is also likely that most carcasses are rapidly consumed by scavengers, making estimations of mortalities resulting from rodenticide bait difficult to predict. Merton et al. (2002) showed high shorebird mortality in brodifacoum broadcast eradications on islands in the . However, these eradications were timed to coincide with the lull in the tourist season rather than a seasonal low in shorebird numbers, so birds were present on the islands in great numbers during the eradication. For the Asiatic whimbrel, a related species of the bristle-thighed curlew, 20% mortality was observed on one of the islands. For the ruddy turnstone, the mortality rates ranged from 20% to 90% on the 4 islands, with exceptional mortality rates possibly enhanced due to the birds being fed by people (Merton et al. 2002). The individuals over-summering on MANWR are the young non-breeders. Using data from bristle-thighed curlew mortality rate on Palmyra in 2011 (3.3 to 37.4%), the mortality documented on the Seychelles for the closely related whimbrel (20%) and the bristle- thighed curlew mortality reported from Rawaki Island (25%), and the number of birds expected to be on Sand Island during the July/August drop period, we calculated the minimum and maximum % mortality to bristle-thighed curlew from the Sand Island rodenticide application. It is anticipated

3-65 MIDWAY SEABIRD PROTECTION PROJECT FINAL ENVIRONMENTAL ASSESSMENT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

that the bait application could sub-lethally affect or incidentally kill 5-52 bristle-thighed curlew on MANWR without mitigation (trapping and holding birds in captivity or clipping wings and releasing birds on Eastern Island). Birds that remain on Eastern Island during the drop would be unlikely to experience any primary or secondary exposure. However, without trapping and clipping the wings of birds, individuals would be able to fly from Eastern to Sand Island. In addition, because it is not a closed population and because breeders would not be taken, it is unlikely there would be any negative long-term population-level effects to this species from this alternative. Brodifacoum could persist in the environment for 6 months to a year and have impacts beyond the proposed operational window if bait were available to foraging curlews (Rueda et al. 2016, Pitt et al. 2015), such that any returning shorebirds would be at risk from secondary poisoning. Therefore, additional bristle-thighed curlew could be killed in the months following the application of brodifacoum. Taking the estimated maximum percent of bristle-thighed curlew killed on Palmrya (Pitt et al. 2015) and using the highest count of this species recorded on Sand Island since 1993, we calculated the potential additional mortality of bristle-thighed curlew over the fall and winter after the bait drop. There would be an additional 71 bristle-thighed curlew at risk over the next 12 months after the baiting operation. Combined with the estimated morality from the initial bait drop, this is 50% of the highest count of birds on MANWR. This is a worst-case scenario based on the maximum expected number of birds on MANWR, maximum expected mortality rates, and no mitigation. However, as stated in Section 3.3.6.1, there are many indications that LD50 values for rodenticide products have underestimated their true toxicity and may not reflect true risks in the field (Mineau 2018). Regardless of the formula used to assess toxicity of brodifacoum to bristle- thighed curlew, it is clear that the risk of primary and secondary poisoning to these birds is very high. However, because of the high degree of site fidelity for birds on their wintering grounds it would be likely that mortality of shorebirds from rodenticide poisoning would decline in subsequent months. Combining the potential estimated mortality from the initial bait drop and over the subsequent year, the total estimated mortality of bristle-thighed curlew (n=123) from primary or secondary poisoning is 1.23% of the estimated global population. A population viability analysis (USFWS 2011) was completed for the Palmyra rat eradication project to explore the implications of incidental poisoning mortality and other mortalities of bristle- thighed curlew due to activities associated with the proposed application of brodifacoum. The model investigated the likely effects on global bristle-thighed curlew populations from the mortality of 10, 50, and 150 bristle-thighed curlew individuals on Palmyra Atoll. They assumed, given the timing of the rat eradication, that the majority of mortalities would be sub-adult birds remaining at MANWR during the breeding season, but they conservatively assume that 50% of the mortalities would be adults. For all scenarios, a second conservative assumption of an additional 50 bristle-thighed curlew lost due to a simultaneous but unrelated one-time event was added, again assumed to be 50% adults. They then modeled population trajectories over 50 years and reported the extinction risk and projected median and lower 90th percentile of population size. Therefore, the mortality scenarios they considered on Palmyra included 60 to 200 individuals which represented a 3-8% increase in adult mortality and a 13-31% increase in sub-adult mortality of the global population for one year. They concluded that these one-time mortality events appeared to have a minor and diminishing effect on projected future populations. The one-time mortality events did not alter future population growth rate and did not appear to put the population in danger of increased risk of stochastic extinction over a 50-year period. Therefore, extrapolating

3-66 FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

from that model, the expected mortality of 5-123 bristle-thighed curlew from the MANWR mouse eradication project, even without the mitigation described below, is not expected to have long- term adverse effects to the global population. Operational Hazard (Bird Strike and Noise/Human Disturbance) It would be unlikely ground operations associated with the bait drop would impact any bristle- thighed curlew, since no nesting occurs on the island. Air operations have the potential to disturb bristle-thighed curlew but observations from field personnel during pre-project surveys on Palmyra indicate this disturbance would be minor and may cause birds to temporarily flush from their location (USFWS 2011). Bristle-thighed curlew could be at risk of collisions with the helicopter during bait drop operations. However, no shorebirds were known to be killed from collisions with the helicopter during the Palmyra operation, and no bristle-thighed curlew have been recorded killed by aircraft at Henderson Airfield on Sand Island. In addition, shorebirds including bristle- thighed curlew are not concentrated in locations on Sand Island but scattered across a larger area. Most are also fast and maneuverable, and thus better able to avoid helicopters, particularly as the helicopters will be flying at approximately 50 meters above ground. We know of no records of bristle-thighed curlew colliding with a helicopter (USFWS, Unpublished b), although the occasional plover may be hit. Therefore, the risk of bristle-thighed curlew colliding with a helicopter are very low. However, based on the data of small-bodied birds (sooty terns and unidentified birds) killed from helicopter collisions on Palmyra (0.085 birds per hour), and using the estimated hours of helicopter time for the Sand Island operation and expected population of bristle-thighed curlew during the drop, the estimated mortality is less than 1 bird for the curlew. Calculations are based on the maximum number of birds observed in 2017 during the planned drop period. We expect there could be some additional impact to bristle-thighed curlews should the need arise to implement the rapid response portion of the refuge’s biosecurity plan and during implementation of the project monitoring plan. A biosecurity plan has been in place on MANWR for many years and any additional bio-security measures would only occur if there was a new incursion of a non- native species. Biosecurity measures enacted would include rapid response to an incursion by non- native invasive species and implementing the monitoring plan would include follow-up observations and sample collecting (see Appendix A and B). Potential impacts, such as causing birds to flush, could occur from the additional human disturbance. However, we expect any additional impacts to be minor, localized, and short-term. Mitigation Several mitigation options have been proposed (see the shorebird protection plan in Appendix C) that could minimize exposure of bristle-thighed curlew to bait containing the rodenticide during the proposed eradication campaign on Sand Island. These have been described above for other shorebirds but for the bristle-thighed curlew these would also include: • Trapping as many individuals as possible and keeping them in captivity or clipping their flight feathers and moving them to Eastern Island until risk of exposure is eliminated or greatly reduced and the birds grow new feathers. Supplemental food and water would be supplied to ensure adequate carrying capacity on Eastern Island.

3-67 MIDWAY SEABIRD PROTECTION PROJECT FINAL ENVIRONMENTAL ASSESSMENT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

For the Palmyra Atoll rat eradication feasibility study (Howald et. al. 2004), the assessment team concluded that the most effective mitigation for this species was to conduct the eradication between early June and mid-July, when most adult bristle-thighed curlew are still on their breeding grounds in Alaska. Similarly, the lowest counts of bristle-thighed curlew on MANWR are in June. Timing during the summer period would also minimize the risk of short-term rodenticide exposure to the majority of curlew overwintering on MANWR. Any birds remaining on the atoll at this time of year are likely sub-adults (aged 1 to 3 years) that remain on the tropical Pacific islands until they mature. In addition, Pierce et. al. (2008) noted that it was easier for bristle-thighed curlew to ingest rodenticide bait after it had been softened by rainfall. However, the eradication on Sand Island will occur during the driest time of the year. In addition, the rodenticide bait used for Sand Island would be dyed a green color to reduce the likelihood of birds picking up the bait (Day and Matthews 1999, Oppel et al. 2016). The Palmyra Atoll eradication used blue-green dye, which was a color known to be least visible/preferred to non-target species including shorebirds (U.S. National Park Service 2000). Live-trapping bristle-thighed curlew prior to the eradication of mice and holding the birds in captivity on Sand Island, or releasing birds with clipped wings on Eastern Island, would eliminate the risk of primary and secondary rodenticide exposure (see the shorebird protection plan in Appendix C). Although feasible, it is unlikely that a significant portion of the summer population, the majority of which are likely sub-adult birds (1-3 years old), could be captured before the bait drop. Before the bait drop on Palmyra in 2011 for the rat eradication, biologists were able to trap 5-9% of the bristle-thighed curlew population present (USFWS 2011). If this same rate of capture were possible on MANWR, approximately 12 to 22 birds could be captured and held in captivity or released with clipped wings on eastern Island. However, there would also be additional risks associated with capturing birds, including physical injury and physiological stress that could result in the death of some individuals. Thirteen bristle-thighed curlews were caught and successfully held in captivity on Palmyra without any injuries or mortalities reported, but several different trapping methods would need to be employed on MANWR to maximize success, as bristle-thighed curlew tend to be extremely wary.

3.3.6.8 Laysan duck

3.3.6.8.1 Existing Conditions The Laysan duck (Anas laysanensis) is currently on the Federal Endangered Species List due to its restricted distribution and small wild population. The Laysan duck (see Figure 3.8) was extirpated across the Hawaiian archipelago with one extant population persisting only on the island of Laysan. The translocated population of Laysan ducks is a species considered “critically endangered” (CR) by the IUCN as well as one of the most imperiled waterfowl in the northern hemisphere (BirdLife International 2016a). A second and third population were established via translocation of wild birds from Laysan Island to Midway Atoll in 2004 and from Midway Atoll to Kure Atoll in 2014 (Reynolds and Klavitter 2006, Reynolds et al. 2015). The Laysan duck is largely nocturnal and sedentary. Body weight fluctuates significantly with season and reproductive status (Moulton and Weller 1984). In spring, females are heavier than males and can exceed 1.3 lbs. (600 g), while males are at less than optimum mass because of courtship activities. In summer, males are generally heavier than females. Females with young

3-68 FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

broods are lightest of all mature birds, often <0.88 lbs. (<400 g). Having evolved with avian rather than mammalian predators, Laysan ducks are more likely to walk than fly, and freeze rather than flush, when startled. Similar to other waterfowl, Laysan ducks molt all of their feathers at the same time, during which they become incapable of flight, and thus more vulnerable to predators at this time. On Laysan Island, molting typically occurs between July and August for males and between July and September for females. On MANWR, males start molting in June and the molt for females with brood may extend into October.

Figure 3.8 Laysan duck family The Laysan duck feeds almost solely on macroinvertebrates, primarily insects (Moulton et al. 1996). These include Dipteran brine fly, adults, larvae, and pupae taken in and around lakes and ponds, along with brine (Artemia spp.) They also eat the larvae and pupae of noctuid ( dislocate) under low vegetation in upland areas. Habitat requirements of the Laysan duck include vegetative cover, an invertebrate prey base, a source of fresh water, and protection from mammalian predators (USFWS 2009a). On Laysan and Midway, Source: USFWS (2017) ducks use all available habitats: upland vegetation, ephemeral wetlands, freshwater seeps, mudflats, the hyper-saline lake, and coastal areas. Nests are on the ground and well-concealed, usually within a base of vegetation, especially bunchgrass (Eragrostis variabilis), but sometimes in Cyperus or Heliotropium, on vegetated portions of the island (Moulton and Weller 1984). Egg-laying can begin as early as February and occur as late as November (Reynolds et al. 2007), but typically occurs from April to August; initiation and duration of egg-laying varies from year to year. Duckling activities are concentrated near sources of fresh water with nearby food and cover. Extirpation of the Laysan duck from the main Hawaiian Islands most likely was caused by a combination of predation by introduced mammals, especially rats, hunting by humans, and habitat destruction and degradation. Current threats to Laysan ducks vary. A small population size and extremely limited distribution make this species highly vulnerable to demographic fluctuation and chance of adverse environmental influences from droughts, severe storms, epizootics, predators, and invasive species (Mitchell et al 2005, Pyle and Pyle 2017, USFWS 2009a). Alien species (mice, invasive weeds, and possibly predatory insects) can alter the Laysan duck’s habitat (USFWS 2009a). Habitat degradation and loss within PMNM may be intensified by increased storm severity and sea level rise associated with global climate change (Vanderwerf 2012). Mass mortality from avian botulism is also a threat, especially for the MANWR population of Laysan ducks that appear more sensitive to botulism outbreaks (USFWS 2009a). Dead mice and other dead non-target species could exacerbate the botulism issue by promoting a maggot cycle. The last severe outbreak of botulism at MANWR occurred in 2016 (K. Goodale, USFWS Unpublished).

3-69 MIDWAY SEABIRD PROTECTION PROJECT FINAL ENVIRONMENTAL ASSESSMENT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

The duck population on MANWR was founded with 42 wild birds from Laysan Island during 2004–2005 and grew to 661 adult and juvenile birds (95% CI 608–714) in 2010. A population decline of 38% was observed between 2010 and 2012 after the 2011 Tōhoku Japan earthquake- generated tsunami inundated 41% of the duck’s nesting and terrestrial foraging habitat (Work et al. 2010, Reynolds et al. 2017). An avian Botulism type C (Clostridium botulinum) outbreak also occurred following the tsunami flooding event, contributing to overall mortality documented in 2011 (Reynolds et al. 2017, USFWS unpublished data). After a severe botulism outbreak during 2015, the population again experienced a 37% decline. Monitoring data indicate that the MANWR population, like the Laysan Island population, is susceptible to catastrophic population declines (Work et al. 2010, Reynolds et al. 2017). Botulism has emerged as a problem for the Laysan duck on MANWR, and mortality has occurred even with intervention by the USFWS (capture and treatment, hold and release, and habitat modification). In an analysis of Laysan duck population sizes on MANWR, Reynolds et al. (in press) used 8 years of monitoring data to estimate population size and test the validity of an index to accurately monitor the abundance of this species. For her analysis, Reynolds fitted 587 Laysan ducks with unique markers from 2004 to 2015 and she recorded 21,309 re-sightings until March 2016. She also conducted standardized survey counts from 2007–2015. Then, a modified Lincoln-Petersen mark-re-sight estimator and statistical (ANCOVA) models were developed to test the relationship between survey counts and population abundance. Her results showed strong, positive correlations between the seasonal maximum counts and population estimates. The statistical models supported the use of standardized bi-monthly counts of unmarked birds as a valid index to monitor population trends among years and within a season at MANWR. In 2015, Reynolds estimated there were between 314 and 435 Laysan ducks (95% CI for population estimate) on MANWR. The point estimate was 375 individuals (Reynolds et al. In Press). This estimate of Laysan ducks at MANWR is approximately 50% of the global population. In comparison, a population estimate on Laysan Island in 2012 was 339 individuals (95% CI: 265– 413). The most recent estimate of the population is 581 adult birds (95% CI 503–682) on Laysan Island (Reynolds and Citta 2007). Twenty-eight Laysan Ducks were translocated from Kure Atoll from Midway Atoll by a team including USGS, USFWS, and DOFAW in the fall of 2014; these had produced at least 19 fledglings by May 2015 and totaled 35-42 birds by 2016 (Reynolds et al. 2015), with continued successful breeding (Pyle and Pyle 2017). The Reynolds model was then used to estimate the current population of Laysan ducks on MANWR using data from March 2017 to March 2018 that includes numbers from the non- breeding season (October-February). Using a maximum count of 372 ducks, the model returns a population estimate of 600 ducks with a range of 526-685 (95% CI) individuals (K. Goodale, Pers. Comm.).

3.3.6.8.2 Probable Impacts: No Action Alternative Under the No Action Alternative, rodenticides would not be used island-wide to eradicate mice on Sand Island, except for targeted control with the less toxic AGRID3. Since the rodenticide AGRID3 used under the no action alternate is not known to be harmful to birds, there would be no adverse effects to Laysan ducks under the no-action alternative from the use of this rodenticide. However, any adverse impacts of predatory mice to nesting Laysan ducks, eggs, and their young would continue. Mice could also be competing with ducks for invertebrate food resources and mice are

3-70 FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

known to be a seed predator on Eragrastis spp., and important grass and plant cover for Laysan ducks. Therefore, any negative effects on Laysan ducks from the presence of the mice would continue under the no-action alternative.

3.3.6.8.3 Probable Impacts: Preferred Alternative Primary and Secondary Exposure Initial tests at sites on MANWR where non-toxic bait piles were put out and monitored, indicated that Laysan ducks would readily consume bait pellets. Thus, there is a clear, primary route of exposure to the rodenticide, as bait consumption by ducks is assumed. Since the ducks also consume invertebrates, there is a likely secondary exposure. The consequence of that exposure is presumed to be substantial and without mitigation, a large number of individual ducks present on the island during the eradication would very likely succumb to the toxic effects of the rodenticide. The use of brodifacoum to eradicate mice on Sand Island poses risks to Laysan ducks that must be mitigated to ensure that the benefits of the proposed action outweighs the risks it poses to non- target species, a fundamental principle of rodent eradications. The toxicity of the compound is determined to a significant extent by species' intrinsic susceptibility, and the toxicokinetics of the compounds used (Eason et al. 2002). Brodifacoum is known to be highly toxic to birds (Eason et al. 2002) and toxicity is a fixed element of risk. Therefore, reducing or eliminating the pathways of exposure would be a key mitigation strategy as described in the protection and mitigation measures below. A female Laysan duck’s average weight during incubation is approximately 1.1 lbs. (500 g). The -6 lethal dose (LD50) of brodifacoum for birds is 4.1 x 10 oz./lb. (0.26 mg/kg). This is the dose the dosage (D) of brodifacoum that is likely to be lethal (L) to 50% of birds ingesting the toxin. For primary exposure to the rodenticide, at an average weight of 1.1 lbs. (500 g), an individual duck would need to ingest only 5 bait pellets to receive a potential lethal dose of brodifacoum. For secondary exposure, an individual duck would need to ingest 1.5 oz. (42.6 g) of contaminated invertebrate prey, which would be 57.5% of a bird's daily food intake (Table 3.16). However, as stated in Section 3.3.6.1, there are many indications that LD50 values for rodenticide products have underestimated their true toxicity and may not reflect true risks in the field (Mineau 2018). Regardless of the formula used to assess toxicity of brodifacoum to Laysan ducks, it is clear that the risk of primary and secondary poisoning to this species is very high. Therefore, based on Laysan duck feeding habits, primary and secondary exposure to the rodenticide are both likely sources of mortality. Thus, without mitigation, the impact could be very high to a significant proportion of the Laysan duck population. Without mitigation measures, the worst-case scenario is that the entire population on Sand Island could be at risk of mortality. However, birds that remain on Eastern Island would not be exposed to the bait and have little likelihood of being exposed to any contaminated invertebrate prey. Setting aside the worst-case scenario, using the maximum population estimate of 600 ducks for MANWR, it is possible that 20-80% (120-480 birds) of the population could be at risk from primary or secondary poisoning without mitigation. This range was derived from mortality data of shorebirds from the Palmyra rat eradication project (USFWS 2011, Pitt et al. 2015), other small mammal eradication projects documenting mortality to waterbirds (Pierce et al. 2008, Morton et al. 2002), and a worst-case mortality rate of 80% reported from a study of passerine mortality from an island rat eradication project using brodifacoum (McClelland 2002). Other instances of high avian mortality from rodent eradication projects

3-71 MIDWAY SEABIRD PROTECTION PROJECT FINAL ENVIRONMENTAL ASSESSMENT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES include severe population impacts to ravens on Langara Island, British Columbia (Howald et al 1999), 80-90% mortality of Stewart Island weka, a in New Zealand (Eason and Spurr 1995), and mortality to bald eagles on Rat Islands, Alaska, where forty six bald eagles died, which exceeded the known population of 22 bald eagles on the island (The Ornithological Council 2010). Laysan ducks could be injured or killed during each phase of the operation including during capture, translocation, handling, captivity, or after bait is dropped and available on the ground on Sand Island (Table 2.1). The maximum numbers of injury/mortality estimated in Table 3.17 represent a worse-case scenario and are unlikely to occur. The primary source of injury and mortality risk to adults is expected to occur to ducks that are not captured before the bait is dropped (the aerial and hand-broadcast baiting phase) or if released back onto Sand Island too early (release phase), or to ducks kept in captivity (Table 3.17). The estimated worst-case mortality to adult ducks is not expected to exceed 33% of the total population. For subadults and ducklings, natural mortality is high, and so there will be some natural mortality to birds that are wing-clipped but free on Eastern Island. Since it will be difficult to distinguish natural injury and mortality of these age classes from injury/mortality indirectly due to project operations on Eastern (overcrowding of birds, aggression, etc.), the numbers in Table 3.17 also include expected natural injury and mortality to these age classes. It should also be noted that a duck could be injured, recover, and receive some other form of injury, or later die from another phase. Thus, the total estimated number of ducks injured or killed cannot be simply summed between all the phases of the operation. No mortality or injury to ducklings and eggs will occur during the release phase of the operation since these age classes are expected to be subadults by that time. The protection and mitigation strategy for Laysan ducks will be in effect to minimize adverse effects and impacts throughout the project (Appendix C, D and E). Spatially, brodifacoum will be available to Laysan ducks in different ecological compartments such as bait pellets, invertebrates, and soil. Temporally, brodifacoum will remain available at biologically significant levels for an unknown amount of time after the bait application. Forecasting the potential pathway and risk of exposure over time for Laysan duck on Midway is challenging without a comprehensive understanding of the food web, and a priori knowledge of the persistence of brodifacoum. Thus, risk must be inferred from toxicological and ecotoxicological data from other rodent eradication projects (Pitt et al. 2015, Pierce et al. 2008, Merton et al. 2002, McClelland 2002, Howald et al. 1999, Eason and Spurr 1995, The Ornithological Council 2010).

3-72 FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

Table 3.17 Estimated number of injuries and mortalities to each Laysan duck age class from the proposed project Estimated Range of Laysan Duck Injury and Mortality Adults Subadults Ducklings Eggs Project Phase Max Max Max Max Max. Injury Mortality Injury Mortality Injury Mortality Mortality Min Max Min Max Range Range Range Range Range 0 30 0 6 Capture1 5 5 10 10 20 0.0% 5.0% 0.0% 1.0% 0 30 0 6 Translocation 5 5 10 10 20 0.0% 5.0% 0.0% 1.0% 0 30 0 6 Handling2 5 5 0 0 N/A 0.0% 5.0% 0.0% 1.0% 0 120 0 60 Captivity3 30 30 75 10 80 0.0% 20.0% 0.0% 10.0% 30 120 30 60 Aerial and 60 60 30 30 20 Hand-Broadcast4 5.0% 20.0% 5.0% 10.0%

6 60 6 60 Release5 50 30 N/A N/A N/A 1.0% 10.0% 1.0% 10.0% Totals/Ranges 36 390 36 198 155 135 125 60 140 Total % 6.0% 65.0% 6.0% 33.0% N/A N/A N/A N/A N/A 1Includes the number of ducks that may be injured or killed during capture operations and possible breakage of eggs. 2This only includes the number of ducks that may be injured/killed during flight clipping and botulism vaccination procedures. 3Includes ducks that may be injured/killed from being kept in aviaries and for ducks being held on Eastern Island (due to crowding and aggression around food or water guzzlers, botulism, etc.) and any natural mortality on Eastern Island. On Eastern Island, it will be hard to discriminate between natural mortality and mortality due to the operation. 4Includes the number of ducks that may be injured/killed from bait being available on the ground from the aerial drop and from disturance from hand-broadcasting operations.

5Includes ducks that could molt on Eastern Island and fly back to Sand Island before the environment is considered safe for release along with other ducks that are released and could be affected by any remaining brodifacoum residues.

Spatial Risk from Brodifacoum Without mitigation, Laysan ducks would have several potential exposure pathways to brodifacoum once bait is delivered to Sand Island and moves through different ecological compartments. Initial tests with non-toxic bait pellets on MANWR, indicated that Laysan ducks would highly likely consume bait pellets, and thus, would be at high risk of primary poisoning. Based on feeding trials conducted on Midway, it is likely Laysan ducks (i.e., adults, juveniles, and ducklings) remaining on Sand Island while bait is available on the ground will consume a lethal dose of brodifacoum.

3-73 MIDWAY SEABIRD PROTECTION PROJECT FINAL ENVIRONMENTAL ASSESSMENT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

Invertebrates are known to feed on bait pellets, and although generally not negatively affected by anticoagulant rodenticides (Booth et al. 2003, Hoare and Hare 2006) (but see Booth et al. 2003 and references therein), may accumulate brodifacoum and function as secondary pathways of exposure to animals that feed on them. Laysan ducks are known to consume invertebrates as part of their diet and therefore, a secondary path of exposure to the rodenticide is possible and highly likely. Without mitigation, it is possible that Laysan ducks (i.e. adults, juveniles, and ducklings) would accumulate lethal doses through secondary exposure and succumb to mortality. Moreover, sub-lethal effects (e.g. lethargy, subcutaneous, intramuscular, and internal hemorrhaging, piloerection, diarrhea, bloody diarrhea, and anorexia) are possible. Temporal Risk from Brodifacoum Unconsumed bait pellets containing brodifacoum will be available on the ground for an unknown amount of time on Sand Island. Degradation time of bait pellets will be dependent on weather conditions, most notably rain and humidity given that decomposer organisms such as molds, fungus, and microbes will potentiate degradation by physically breaking down bait pellets and rodenticide. During bait trials on Midway, bait pellets in ironwood forest appeared as fresh pellets after at least 17 days of relatively dry conditions (Wes Jolley, pers. comm.). On Desecheo Island, bait pellets remained in palatable form for at least 20 days after the first application (Shiels et al. 2017). On Lehua Island, bait pellets containing Diphacinone, but with similar inert matrix to B25D, remained relatively intact for at least 24 days, degrading slowly and consistently (Island Conservation 2017). Finally, on Kahoolawe Island, inert bait pellets placed in exclusion cages on hardpan surfaces remained for at least five months (Island Conservation, unpublished). Once bait pellets are applied to Sand Island, the project team has no control over how long it will take for bait pellets to degrade. A primary path of exposure to Laysan ducks will exist for as long as bait pellets are available on the ground. Once bait pellets degrade, brodifacoum residues bind tightly to organic matter in soil where the rodenticide is degraded by soil micro-organisms and exposure to oxygen and sunlight. The half- life in soil is ~84 to 175 days, depending on the soil type and aerobic vs. anaerobic soil conditions. The rate of microbial degradation is dependent on climatic factors such as temperature, light, humidity, and the presence of molds and soil microbes that potentiate degradation. Therefore, in general, degradation time will increase in colder climates and decrease in warm sunny places. The rate of brodifacoum degradation is unknown on Midway. On Anacapa Island in southern California, only one of 48 soils samples taken six months post bait applications tested positive for brodifacoum (Howald et al 2010). Furthermore, concentration of brodifacoum in bait pellets left in exclusion cages had decreased by 55% in six weeks and 92% in six months (Howald et al 2010). Invertebrates are known to consume bait pellets containing brodifacoum and the amount of time that residues will be present in invertebrates that are part of the Laysan duck diet is unknown. Therefore, there is inherent uncertainty in the temporal scale for which brodifacoum residues will remain available at biologically significant levels after the operation is completed through a secondary/tertiary path of exposure. Laboratory studies have found brodifacoum at low but detectable levels in cockroaches 42 days after exposure (Brooke et al. 2013). Cockroaches, analyzed by whole body, had brodifacoum concentration of 45 ng/g (ppb) four years after an eradication attempt on Desecheo Island (Shiels et al. 2017). On Palmyra, brodifacoum residues peaked after bait application and then decreased, but were still detectible in hermit crabs, fiddler crabs, cockroaches, ants and 46-52 days after the last application of bait (Pitt et al 2012).

3-74 FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

Moreover, there is the potential that Laysan ducks could bio-accumulate brodifacoum residues over time through the consumption of invertebrates contaminated at very low levels. Although brodifacoum will eventually move out of the Laysan duck’s food web into the soil matrix, where it will be degraded by bacteria, molds, and fungus to base components of C02 and H20, it may be remobilized by invertebrates into the food web over time, with uncertain ecological consequences (see examples: cockroaches - Shiels et al. 2017 – and Lizards - Rueda et al. 2016). Operational Hazard and Disturbance (Bird Strike, Noise/Human Disturbance, Transport, Captive Care, Disease) Since Laysan ducks are more likely to walk from place to place rather than fly, and tend to freeze when disturbed rather than flush, they are at a very low risk for collision with the helicopters during bait drop operations. In addition, Laysan ducks molt all their feathers at the same time and individuals are incapable of flight between July and August for males and between July and October for females. Therefore, many individuals would naturally be incapable of flight during the bait drop period in July and August. Disturbances from helicopter noise and hand-broadcasting of bait could disturb ducks. Disturbance to individual ducks during the rodenticide treatment period, would aid the mitigation plan by helping locate and capture any remaining ducks. Laysan ducks are upland nesters and sensitive to ground-disturbing activities. If present, nests and eggs can be accidentally trampled by ground crews, or adults can be flushed off the nest, resulting in injury and death of adults, subadults, and ducklings, and destruction of eggs and/or ducklings in the nest; if adults are flushed off the nest, the disturbance would only be temporary, but injury or death could result. Ground crews can also accidentally separate broods, which could result in injury and death to the adult and ducklings. If bait applications occur when Laysan duck nests are present, and where hand-broadcasting of bait requires ground-based personnel to traverse across or apply bait in duck nesting habitat, staff will be instructed to remain vigilant for birds and nests, and to call in the capture team if one is detected; this disturbance would only be temporary but injury or death of adults, subadults, ducklings, and eggs could result. As part of the mitigation to protect this species, a significant number of birds would be captured and kept in captivity during the bait drop period, thus protecting them from possible helicopter collisions, disturbances from helicopter noise and humans distributing bait, and accidental trampling of eggs or chicks from distributing bait by hand broadcast methods. In addition, Laysan ducks are accustomed to human presence on MANWR and can be approached within a few feet before being disturbed. Three previous translocations of Laysan ducks have demonstrated a tolerance for boat travel in transport boxes very well and for much longer trips than across MANWR’s lagoon to Eastern Island. Reynolds and Klavitter (2006) successfully transported 42 ducks from Laysan Island to Sand Island in 2004 and 2005 (20 and 22 birds, respectively) via a 50-hour boat trip each year. In 2014, 28 ducks were successfully boat-transported from Midway to Kure Atoll (USFWS 2014, Reynolds et al. 2015). To successfully transport ducks from Sand Island to Eastern Island aviaries prior to the first bait application, protocols for boat transport would follow those of previous efforts (Reynolds et al. 2004, 2014, Reynolds and Klavitter 2006, Reynolds et al. 2015). While Laysan duck translocation operations have had no mortality in captivity, there are risks from captive care including enhanced disease transmission risk, behavioral disruption, prevention of subordinates from getting to food dispensers, and risk of injuries from fighting. Additionally, the higher density of seabird presence, such as Laysan Albatrosses, can increase risk of harm to smaller

3-75 MIDWAY SEABIRD PROTECTION PROJECT FINAL ENVIRONMENTAL ASSESSMENT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

Laysan ducks. Finally, there is a higher risk of botulism outbreaks when there is a high density of ducks using a few water sources. Mitigation measures to address behavioral issues and avian botulism are addressed in Appendices C, D and E. If there was a new incursion of a non-native invasive species to MANWR during the Proposed Action, rapid response measures in the Biosecurity Plan would be immediately enacted (Appendix A). These measures have the potential to impact ducks. Impacts include flushing ducks off their nest, trampling of eggs and nests, trampling of vegetative cover, and increased human noise and presence. These impacts could result in injury and death of adults, subadults, and ducklings, and destruction of eggs and/or ducklings in the nest. Given that the Biosecurity Plan’s prevention measures will be adhered to stringently, it is expected that rapid response measures will not need to be enacted. Therefore, associated impacts are unlikely and implementing the biosecurity plan is only expected to have beneficial effects to Laysan ducks. The accidental introduction of mice on Eastern Island also has the potential to jeopardize the proposed mitigation strategy. Biosecurity measures for the transport of gear, supplies, and personnel between Sand and Eastern Islands will be stringent to prevent the accidental introduction of mice to Eastern Island. Protocols will include using mouse-proof containers, performing rigorous inspections, employing bait stations and/or traps at both boat landings, and using mouse detection gear deployed on Eastern Island. Monitoring efforts have the potential to affect Laysan ducks through added human disturbance. Monitoring duck behavior and health will include follow-up field observations, setting and retrieving cameras on nests to track productivity, and collecting prey samples for determining residue levels of bait in the duck’s food source. These monitoring methods do not require ducks to be handled, minimizing stress levels in ducks and potential for injury. Given the mitigation measures that will be in place to protect Laysan ducks, we expect any additional impacts to be minor and short-term. Mitigation The protection, mitigation, and impact minimization strategies proposed for the Laysan duck focus efforts on eliminating or minimizing exposure of Laysan ducks to brodifacoum whenever possible (Appendix C). It addresses both spatial and temporal risks to Laysan ducks resulting from the proposed action. Given the inherent uncertainty of this type of intervention, the proposed strategy identifies and addresses uncertainty, as well as creates redundancy to produce a successful outcome. Moreover, it identifies and addresses potential cumulative risks that could result from the proposed mitigation actions. The goal of the proposed protection, mitigation, and impact minimization strategy is to ensure a viable long-term population of Laysan ducks on MANWR. Success is defined as all the surviving captive Laysan duck individuals released back to Sand Island once the operation is concluded and the return of the remaining ~400 wing-clipped birds on Eastern Island when their feathers molt and they regain flight capability to return to Sand Island. For reference, the population currently present on Midway is the result of a translocation effort that included 42 birds; in 6 years, this population grew more than 15-fold to 661 birds. Translocating ducks off of Midway Atoll as an additional protection measure was not possible due to current funding, regulatory and compliance constraints. Eliminating or Minimizing Exposure

3-76 FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

Measures to minimize rodenticide exposure to Laysan ducks include avoiding aerial broadcast near wetlands and open water bodies that cannot be covered and hand broadcasting in these areas when necessary thus keeping bait out of fresh water ponds and seeps. Many of these wet areas will not contain water since the July application period is the driest time of the year for MANWR. In addition, Laysan ducks would be removed from Sand Island and temporarily translocated to Eastern Island where they would not be exposed to bait pellets and have a low likelihood of exposure to contaminated invertebrates. Given the limited carrying capacity of Eastern Island (85 - 102 Laysan ducks) (SWCA 2017), enhanced shelter, food and water sources for Laysan ducks is required. These enhancements would increase the carrying capacity of Eastern Islands to approximately 400 ducks. The proposed action would implement the Laysan duck protection and mitigation strategy (Appendix C) which involves two redundant levels of protection to ensure the survival of as many healthy individuals as possible (~600) as described below: 1) The first set of actions would include capture of 200 Laysan ducks (40% males and 60% females, and 60% adults), and placement in aviaries on Sand Island until implementation, then transporting all ducks to Eastern Island and placing them in aviaries there until the risk window has passed as assessed by the residue monitoring strategy described later in this document. Individuals included in this strategy will be monitored more closely and their return to Sand Island is completely under the control of the project team. 2) The second set of actions would include capturing Laysan ducks, holding them for 10 days in order to give them two doses of botulism vaccine, clipping flight feathers, marking with color bands, and transporting to Eastern Island for release. Individuals that are part of this strategy are expected to be unable to fly until the next year’s molt and remain on Eastern Island. Given the limited carrying capacity of Eastern Island, we would provide enhanced shelter, food and water sources to increase the carrying capacity. Monitoring of movements, behavior, breeding, and survival would be done using a combination of observations at a distance using spotting scopes to minimize disturbance. Capturing of ducks (all life stages; adults, juveniles, ducklings, eggs) will start in Phase 1 of the operation and will continue until the residue monitoring strategy determines that the risk window has passed (Appendix C – Section 4.5). Any ducks not captured that are subsequently exposed to the rodenticide and demonstrating signs of bait ingestion or toxicosis such as colored feces, lethargy and weakness or bleeding would be captured and treated with the antidote Vitamin K by a veterinary professional to offset the negative effects of the rodenticide (Appendix C – Section 4.5). Recent studies with rats suggest that even if re-captured and antidoted for the haematological effects of brodifacoum, individuals may suffer longer-term neurological effects (Kalinin et al. 2017); although these results have not been demonstrated in birds. Ducks showing signs of exposure would be held in temporary holding pens under observation and any decision-making would occur based on the duck’s on-going health prognosis. Laysan ducks that survive after treatment will be wing-clipped and transported to Eastern Island for captive care there until Sand Island was deemed safe (Appendix C – Section 4.6). The existing duck lab in the USFWS building has the appropriate facilities, equipment, and medications to handle duck health emergencies and will be used as an avian clinic. It is estimated that approximately 8-15 personnel will be needed to handle these activities (Appendix C – Table 2).

3-77 MIDWAY SEABIRD PROTECTION PROJECT FINAL ENVIRONMENTAL ASSESSMENT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

Laysan duck detectability on MANWR varies across the different seasons and parts of the reproductive cycle. During the proposed capture period, detectability is highest in February, then decreases through the month of April and is low from May to June (Appendix C – Figure 1). Detectability of ducks then starts increasing again in July and August. Intensive searching of all areas of Sand Island by experienced duck capture experts will ensure that most animals are located. By starting early, we will have the capability to cover every square meter of Sand Island with duck search efforts. Any ducks that evade detection will likely succumb to brodifacoum poisoning so extreme effort will be made to locate all ducks. Only specifically-trained and federally-permitted personnel will handle Laysan ducks. Personnel will include ornithologists, avian rehabilitators, aviculturists, and licensed veterinarians. A veterinarian will be on site for the duration of the initial capture period, bait application period, and post application holding period (Appendix C – Table 2). Transport of the ducks is described in the Laysan Duck Mitigation Strategy (Appendix C – Chapter 4.4). In general, transport will consist of numerous boat trips across the approximately one-mile route via several boats available for transit to Eastern Island with as many birds as can be carried safely at a time, with one bird in each transport box. Subgroupings of captive ducks that were established in the Sand Island aviaries will be maintained to avoid a new round of social disruption and hierarchy establishment. Aviculturists will alternately rotate shifts on Eastern Island where a small camp will be established, to ensure care for captive animals regardless of weather and sea conditions. Three previous translocations of Laysan ducks have demonstrated that they tolerate boat travel in transport boxes very well and for much longer trips than across the lagoon. There has never been a loss due to transporting Laysan ducks by boat using small transport boxes or cat carriers as described in the Laysan duck Mitigation Strategy, which supports the project’s expectation that there will be minimal duck mortality due to transport (Table 3.1.7). Monitoring and Assessment of Spatial and Temporal Risk Bait degradation and disappearance will be monitored on Sand Island using pre-established bait plots in which persistence is checked regularly until all pellets are gone. Brodifacoum residues would be monitored in relevant ecological compartments including soil, and invertebrates starting at three weeks after the last application and at an appropriate frequency to inform risk mitigation response thereafter. Sentinel diet monitoring will be implemented using species known to be part of the Laysan duck diet (e.g. cockroaches) at appropriate frequency to inform risk mitigation response. Brodifacoum residue testing would be used to test temporal risk assumptions and would inform mitigation responses. Brodifacoum residues will be assayed and quantified by a qualified laboratory. Residue Monitoring and Duck Release Strategy Sampling and analysis of brodifacoum levels will be used to test assumptions regarding the expected values in different ecological compartments based on best available data and risk analysis and will inform Laysan duck release strategy (e.g. timeline for initial release). Typically, what is observed following bait application is a pulse of brodifacoum residues into the ecosystem and a subsequent decline of residue levels over time. Although presence of residues in the food web represents a risk, Laysan duck will tolerate a low level of environmental residues when the rate of ingestion equals or is less than the rate of metabolism. Forecasting that threshold and identifying

3-78 FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

the biologically safe residue level a priori relies on estimates from toxicological and ecotoxicological data and is supported from empirical (real time) data. (Figure 3.9).

Figure 3.9 Expected brodifacoum degradation curve following an eradication operation.

Red indicates biologically significant levels of brodifacoum residues and green indicates safe levels of brodifacoum residues. The uncertainty lies in identifying what are the thresholds for biological significance and the amount of time that the system will take to degrade brodifacoum to those safe levels.

Toxicity is most frequently represented as the LD50. LD50 is the chemical dose where 50% of the test animals died and is usually administered as a single dose. However, there is sufficient evidence to suggest that laboratory-derived LD50 values are not appropriate for predicting field safety to non-target species in response to the use of second-generation anticoagulant rodenticides (Mineau, 2018). For the Laysan duck risk assessment, we used an approached proposed by Thomas et al. (2011) and improved on by Mineau (2014). This approach provides a relationship between liver residues and the probably of evident coagulopathy at necropsy (Thomas 2011) and uses data on the proportion of a single dose that is retained in the liver over time (Mineau 2014). These two sets of information allow for a calculation of a “field-derived LD50” or more accurately ED50 (dose at which half of the test population will show an effect – not necessarily death) (Mineau 2018).

However, to be protective at the individual level for a rare or threatened species, ED50 is not appropriate, and for this analysis we used an ED5 (5% of the exposed population will be affected by coagulopathy – not necessarily death). If we assume a 500g Laysan duck, its liver weight will be approximately 11g (after relationships in broiler – Deaton et al. 1969). Using the approach described above, the calculated threshold whole body ED5 is 0.47-1.8 µg total intake of brodifacoum for an average Laysan duck (Mineau 2018). Because brodifacoum can bio-accumulate through multiple exposures, expressing the risk as a proportion of daily food intake is not appropriate (Mineau 2018). Instead, we used an

3-79 MIDWAY SEABIRD PROTECTION PROJECT FINAL ENVIRONMENTAL ASSESSMENT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

‘accumulation window’ of about 200 days (the mid-point estimate for the half-life of brodifacoum in rat liver tissue) (Horak et al. 2018). Pitt et al. (2015) found that residues in cockroaches averaged 2.3 µg/g 10-15 days after bait drop, and 0.9 µg/g about 45 days later. It is reasonable to assume that, over time, a certain amount of brodifacoum will ‘escape’ the insect food chain to be lost to soil and the abiotic environment more generally. Fitting these two points to an exponential curve (a common curve type when describing residue degradation or loss from a system) and defining the peak residues as day 1 (even though it took 10-15 days for the residues to maximally contaminate the cockroaches) yields the following: Residue level (µg /g) = 2.3496 * exp (-0.0213 * days after peak residue) Predicting forward, cockroach residues at 200 days (our chosen period of residue accumulation) post peak would be approximately 0.03 µg/g. We calculated the cumulative residue intake over a 200-day period which produces a recommended prolonged withholding time for the ducks of 13- 15 months needed to drop the risk level below an ED5.

It is important to note that we are being very protective by choosing an ED5 when the outcome is coagulopathy rather than mortality, representing a conservative approach. It is also reasonable to assume that an increasing part of the duck’s diet over time will be made up of unexposed cockroaches and other invertebrates, so the suggested holding times may be over-estimated. Therefore, a prudent, and conservative estimate, the first release of Laysan ducks may be at the 4- 6-month post bait application, with continued surveillance and monitoring of the ducks (i.e. if any negative effects are detected, adaptive management can be enacted to revert it). The first cohort of released birds would be monitored and followed for at least 6 months to confirm if they are showing symptoms of accumulating residues (Mineau 2018). Furthermore, field application of this proposed model will require real time estimates of remaining residue levels in the Midway ecosystem. Sentinel diet sampling and analysis would be performed at intervals appropriate to inform the mitigation strategy. It will be important to look at the overall loss rate of residues from the food-chain in the early days, weeks and months following the bait drop to characterize the residue decline curves. In addition, we will release canaries (Serinus canaria) and common myna (Acridotheres tristis) and will monitor their health status prior to releasing any Laysan ducks while also monitoring wild shorebirds on Sand Island. 5.1.3.4 Cumulative Risks The proposed mitigation actions have the potential to create additional and potentially cumulative risks to the Laysan duck population on Midway. There is inherent risk when handling critically endangered wildlife. There is the possibility of injury to the head, bill, neck, eye, leg, foot, and wing and even mortality of ducks during the capture effort. Capture methods are detailed in the Laysan duck protection strategy and will minimize injury whenever possible. There have been no injuries among ducks handled in this way in the Monument so it is impossible to estimate likelihood of injury. Maintaining Laysan ducks in captivity poses risks to individual ducks and there is the possibility of injury or mortality among captive-held birds. Aviaries will be designed and maintained to safely hold Laysan ducks in captivity for extended periods of time (>5 months). To minimize risks posed

3-80 FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

to captive ducks, precautions will be implemented that include a using a flooring that prevents ducks from developing foot and leg issues and is easy to clean, predator control measures around and within the aviaries, and closed-circuit camera monitoring of duck behavior (i.e., no human interference). Aviaries will be designed to keep group sizes small enough to be able to detect food exclusion, fighting problems, and signs of disease. Skilled staff will be available onsite to address emergency situations. Botulism outbreaks have occurred on Midway with severe negative impacts to the Laysan duck population. Personnel will be trained to prevent (Appendix D), as well as prepared to handle botulism outbreaks (Appendix E). All initially captured ducks will be held for at least 10 days to receive two doses of the Botulism vaccine. With implementation of the protection and mitigation strategies described, only short-term effects such as interruption of a breeding season and stress to individual ducks due to capture and captive holding are expected. The original population of Laysan ducks at MANWR (2004-2005) was 42 birds; in 6 years this population grew more than 15-fold to 661 birds despite several setbacks caused by botulism outbreaks. There are likely to be some short-term adverse impacts to the population of ducks on Sand Island, but it should recover quickly, especially if a direct food competitor, the house mouse, is removed. Thus, the action is not expected to have long-term adverse effects to the Laysan ducks at the refuge. Laysan duck populations appear to thrive with appropriate management and conservation efforts. The original population on Laysan Island grew to the island’s carrying capacity of about 400-600 (Seavy et al. 2009) in the decades after rabbits were eliminated in 1924. The translocated population of ducks in 2004-2005 to MANWR grew rapidly to 471−535 post-fledglings by 2008 (Reynolds et al. 2012, 2013). These translocated ducks on MANWR appear to breed successfully at an earlier age and produce larger clutches than birds on Laysan, probably owing to more food and a lower population density. Birds at MANWR breed in their first year and produce an average of seven eggs whereas birds on Laysan nest in their second year, producing an average of 3.3 eggs (Reynolds et al. 2008, Walters and Reynolds 2013). An estimated decline of about 42% on MANWR in 2011 to 2012 occurred after the Tōhoku earthquake-generated tsunami and associated botulism outbreak (Reynolds et al. 2015). The last botulism outbreak occurred in Fall 2016 and reduced the population to 350 individuals. As of February 2018, the population had increased to 600 individuals (K. Goodale, pers. comm.). This is an increase of 71% over ~18 months. Therefore, the Laysan duck population on MANWR is expected to recover quickly after any mortality occurs from the Proposed Action (2-3 years or less, depending on actual mortality rates). Since the Laysan duck is also a ground nesting species, the removal of mice would likely contribute to maintaining Laysan ducks on MANWR. Thus, the eradication project is not expected to have long-term adverse effects to the population. To implement any of these mitigation strategies would require USFWS-Refuges to consult under Section 7 of the Endangered Species Act. Any proposal must have no net negative impact to this species. The overall goal would be to maintain a healthy population of ducks on MANWR during and after the mouse eradication.

3-81 MIDWAY SEABIRD PROTECTION PROJECT FINAL ENVIRONMENTAL ASSESSMENT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

3.3.6.9 Land Birds

3.3.6.9.1 Existing Conditions Three non-native landbird species are known to breed at MANWR: the Atlantic canary (Serinus canaria), the common myna (Acridotheres tristis), and the cattle egret (Bubulcus ibis) which is discussed in the next section. Both the Atlantic canary and the common myna are at risk of poisoning as they are present all year-round on Sand Island. Therefore, the timing of the rodenticide application is not a viable mitigation option. The Atlantic canary is known worldwide simply as wild canary and is a small passerine bird in the family, Fringillidae. It is also called the island canary, canary or common canary. It is native to the Canary Islands, the Azores, and Madeira. Wild birds are mostly yellow-green, with brownish streaking on the back. The wild Atlantic canary can range from 3.9 to 4.7 in (10 to 12 cm) in length, with an average weight of around 0.53 oz. (15 g). The breeding population, which is confined to , is estimated to number 1,500,000 to 2,520,000 pairs, which equates to 3,000,000-5,050,000 mature individuals (BirdLife International 2016c). This species is often kept as a pet and is referred to as the domestic canary. Selective breeding has produced many varieties which differ in color and shape. The species is widely kept in captivity in most areas of the world and is also widely used in scientific research. There are no known significant threats to this species (BirdLife International 2016c). Island-wide monthly avian surveys, conducted from June through September 2017, counted 502 to 3,985 canaries on Sand Island (Figure 3.10). Using the maximum count of 3,985 birds from July 2017, the MANWR Atlantic canary population represents less than 0.078% of the global wild population.

3-82 FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

Figure 3.10 Results of island-wide monthly avian surveys conducted from June through September 2017 for land birds

Source: USFWS (2017)

For the Atlantic canary, a cup-shaped nest is built 3 to 20 feet (1–6 m) above the ground in trees or bushes. Nests are made of twigs, grass, moss and other plant material and lined with soft material. Eggs are laid between January and July in the Canary Islands. A clutch contains 3 to 4, or occasionally 5, eggs, and 2-3 broods are raised each year. The Atlantic canary typically feeds in flocks, foraging on the ground or amongst low vegetation. It mainly feeds on seeds such as those of weeds, grasses and figs. It also feeds on other plant material and small insects (Snow and Perrins 1998). In 1909, D. Morrison, superintendent in charge on Midway Island, purchased a mated pair of canaries from sailors on a ship in Honolulu Harbor. He brought them to Midway and kept them separate for 8 months until he put them in a breeding cage in January 1910. Soon after this, additional birds were brought from Hawaiʻi and all birds were released on MANWR. By the end of 1910 they were breeding on Sand Island in the Casuarina trees and by 1922 there were approximately 1,000 canaries at Midway (Fisher 1949). The common mynas native range covers a large swath of south and central Asia, from Turkmenistan eastward through all countries of the Indian subcontinent, including , Pakistan,

3-83 MIDWAY SEABIRD PROTECTION PROJECT FINAL ENVIRONMENTAL ASSESSMENT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

Nepal, Bangladesh, Bhutan, , and . It also exists in much of southeast Asia south to . The species has been introduced and established in many other parts of the world (Kannan and James 2001). It is also abundant in urban areas of the Hawaiian Islands. It is one of the most common and most widespread avian species in all of the human-inhabited islands. Sexes alike, but male median weight of 4.5 oz. (126.5 g) is slightly larger than the female median weight of 4.0 oz. (113.5 g). Baker et al. (1999) estimated the population on MANWR in 1996 to be 750- 850 individuals. Island-wide monthly avian surveys, conducted from June through September 2017, counted 278 to 490 birds for this species on Sand Island (see Figure 3.10). Using the maximum count of 490 birds from August 2017, the MANWR common myna population represents an insignificant proportion of the global population. In the Hawaiian Islands, the common myna nests in palms and trees. They are also known to nest in any place that will hold a large pile of leaves and twigs, including, air-conditioners, drain pipes, open-ended steel rafters, traffic lights, and under eaves of buildings, etc. (Kannan and James 2001). The common myna feeds while on ground and in trees and is omnivorous. The primary diet consists of fruits, grain, insects, and grubs. Its diet can also include kitchen scraps, tidbits from refuse dumps, bird eggs, small animals (young mice, frogs, lizards, crabs), and flower nectar. It has also been recorded to prey on seabird eggs including those of the wedge-tailed shearwaters in the Hawaiian Islands (Byrd 1979), terns (Sterna spp.) and noddies (Anous spp.) in and the Seychelles, and the eggs of gulls (Larus spp.) in New Zealand (Kannan and James 2001).

3.3.6.9.2 Probable Impacts: No Action Alternative Under the No Action Alternative, limited mouse control would continue using the less toxic AGRID3 rodenticide. Therefore, there would be no effects to the Atlantic canary or common myna under the no-action alternative since there are no known adverse effects to birds from the use of AGRID3. Any potential predation of seabird eggs by the common myna would continue.

3.3.6.9.3 Probable Impacts: Preferred Alternative Primary and Secondary Exposure. The data in Table 3.18 indicate that both the Atlantic canary and the common myna would be at risk of primary and secondary poisoning from brodifacoum, since birds would only need to eat a fraction of their normal daily food intake of bait to receive a lethal dose. For a primary pathway, 1-2 pellets of bait (0.03 to 0.07 oz. [1 -2 g]) containing 0.025 mg of brodifacoum each, if ingested, would be enough to kill an individual canary or myna bird (Table 3.17). For secondary toxicity from consuming invertebrate prey, it would only take 18.1% and 36.5% of their daily food intake to reach a lowest lethal dose for the canary and myna bird respectively (Table 3.18). However, as stated in Section 3.3.6.1, there are many indications that LD50 values for rodenticide products have underestimated their true toxicity and may not reflect true risks in the field (Mineau 2018). Regardless of the formula used to assess toxicity of brodifacoum to birds, it is clear that the risk of primary and secondary poisoning to these two species is very high. McClelland (2002) reported that 80% of fernbirds (Bowdleria punctata wilsoni), a warbler that forages on invertebrates, disappeared from a 74 ac (25ha) test plot that was used to assess risk of

3-84 FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

brodifacoum on non-target species prior to implementation of a rat eradication operation on Codfish Island, New Zealand. Robin (Petroica australis australis) populations were monitored in South Island New Zealand before and after aerial treatment of brodifacoum (Talon 20 P, 20 ppm) for control operations targeting stoats (Mustela erminea) and ship rats (Rattus rattus) in a forest environment (Brown 1997). Repeated observations of banded and radio-tagged territorial adult robins were used to monitor survival from 6 to 8 weeks after poisoning. Where poison was freely broadcast, only 52.2% (12/23) of marked robins were known to survive (47.8% mortality rate). At the non- treatment site 85.7% (18/21) of marked robins survived. This study concluded that individual robins were at risk from poisoning from exposed brodifacoum on the forest floor. In summary, brodifacoum would present a lethal risk to the Atlantic canary and common myna. Based on the above data and other avian mortality studies reviewed, it would be possible that 996 to 3,188 (25-80% of the MANWR population) Atlantic canaries and 122 to 392 (25-80% of the population) common mynas could be killed by either primary or secondary poisoning from brodifacoum without mitigation. These maximum numbers represent worst-case scenarios based on maximum numbers of birds documented on Sand Island during the drop window.

Table 3.18 Primary and secondary toxicity of Brodifacoum to Passerines on MANWR Primary toxicity presented as grams of bait and secondary as grams of invertebrate prey. Active ingredient (mg) of brodifacoum is abbreviated as a.i. LD50 is that the dosage (D) of a toxin that is lethal (L) to 50% of animals of the species in laboratory tests. Brodifacoum Body Daily Food LD50 Species Wt. (2) (3) (4) (5) Intake LD50 (mg of LD50 (g LD502 (g % of Daily (g) (g)(1) (mg/kg) a.i./bird) of bait) of prey) Food Intake Atlantic 15.0 7.1 0.26 0.0039 0.156 1.3 18.1 canary Common 126.5 29.5 0.26 0.0329 1.316 10.85 36.5 myna Notes: 1. Daily food intake calculated from allometric equation (Bird Feeding Rate = 0.059 x (W 0.67) (EPA 1995). 2. LD50 of brodifacoum based on mortality of mallard ducklings at the lowest dose, 0.26 mg/kg body weight = 0.026 mg/g (Ross et al. 1980, EPA 1991). 3. This is also the number of bait pellets that would need to be consumed to be lethal to 50% of the animals.4. Cockroach contamination based on highest reported value of 3.05 mg/kg = 0.00305 mg/g (Pitt et al 2015). 5. Calculated as LD502 grams of prey/daily food intake (g).

Operational Hazard (Bird Strike and Noise/Human Disturbance) It would be unlikely ground operations associated with the Preferred Alternative would impact either the Atlantic canary and common myna. Both species either exclusively or preferentially nest in trees or shrubs, so trampling of nests would not be a concern. Helicopter operations have the potential to disturb passerines, but since planes land and take off from Sand Island regularly, this disturbance would be minor and might cause birds to temporarily flush from their locations. It would also be unlikely that helicopter operations would cause any collision hazard with these 2 species. During the helicopter operations for the Palmyra rat eradication, all helicopter collisions were with seabirds and shorebirds. One canary was recorded killed by aircraft at Henderson Field from 2004-2010.

3-85 MIDWAY SEABIRD PROTECTION PROJECT FINAL ENVIRONMENTAL ASSESSMENT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

The potential mortalities that would be expected to result from the Preferred Alternative represent insignificantly small numbers relative to overall global populations of both these species, but the Preferred Alternative would have a moderate negative affect for each of the Sand Island populations. Adverse effects would be readily detectable and localized with measurable consequences to the local population, but not readily detectable or measurable beyond the immediate area of impact. Mitigation measures would only be instituted for the Atlantic canary due to its cultural importance. Indirect effects from the Preferred Alternative would not be expected. We expect there to be no additional impacts to the Atlantic canary and common myna should the need arise to implement the rapid response portion of the refuge’s biosecurity plan and during implementation of the projects monitoring plan. A biosecurity plan has been in place on MANWR for many years and any additional bio-security measures would only occur if there was a new incursion of a non-native species. Biosecurity measures enacted would include rapid response to an incursion by non-native invasive species and implementing the monitoring plan would include follow-up observations and sample collecting (see Appendix A and B). We expect such additional ground-based human activities would not impact tree-dwelling species like the Atlantic canary and common myna. Mitigation Atlantic canaries on Sand Island would be caught and kept in aviaries during the bait drop period (see protection plan in Appendix C), and not released until post-monitoring of brodifacoum residues show decreased chance of primary or secondary exposure. A combination of mist-nets and walk in traps would be used to capture a subset of the resident flock adequate to protect the genetic variability of the population. However, this activity would receive a lower priority than protection of Laysan ducks or migratory shorebirds. In a capture test in 2017, 4 birds were caught in 2 morning hours of mist-netting. Walk-in traps have not been tested but are known to be effective in trapping ground feeding passerines. The Atlantic canary is easy to hold in captivity, and any mortality of birds while being held should be minimal. After release, populations should grow quickly since this species can raise 2-3 broods each year, with the average nest containing 3 to 4 eggs.

3.3.6.10 Cattle Egrets

3.3.6.10.1 Existing Conditions Invasive non-native cattle egrets (Bubulcus ibis) have colonized MANWR by island hopping through the Northwestern Hawaiian Islands. They have been documented breeding on Eastern Island, and the cattle egret regularly use both Eastern and Sand Island. Island-wide monthly avian surveys, conducted from June through September 2017, counted 0 to 83 birds for this species on MANWR (Figure 3.10). The European population is estimated at 76,100-92,300 pairs, which equates to 152,000-185,000 mature individuals (BirdLife International 2015). The global population is larger than this, although no estimate is available. Using the maximum count of 83 birds from September 2017, and the estimate of the minimum number of individuals in Europe, the cattle egret population on MANWR represents 0.05% of the European population.

3-86 FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

The population of this species is actively managed on MANWR due to their predatory behavior on nesting seabirds. The USFWS has issued a federal migratory bird control order at 50 CFR 21.55 which allows for the take of cattle egrets (and barn ) on the main Hawaiian Islands, the Northwestern Hawaiian Islands, and on Midway, which authorizes the Service and personnel from certain other agencies to lethally take cattle egrets to protect native species without a permit (50 CFR 21.55 [or 82 FR 34419]). Prior to this Control Order, control operations were authorized by permit, and from March to August in 2017, MANWR staff destroyed 80 egret nests, treated 303 eggs, and dispatched 10 chicks. Since they only breed on Eastern Island, and a control program is already underway for this invasive species, helicopter disturbance or disturbance from ground personnel is not considered an issue.

3.3.6.10.2 Probable Impacts: No Action Alternative Under the No Action Alternative, Sand Island’s mouse population would not be the subject of an island-wide eradication project, but limited mouse control efforts would continue using the less toxic AGRID3 to protect nesting seabirds and for human health and safety. Therefore, there would be no effects to the cattle egret under the no-action alternative.

3.3.6.10.3 Probable Impacts: Preferred Alternative Primary and Secondary Exposure The likelihood of primary exposure to rodenticides would be very low. However, the egret is a predator species. These birds would be expected to consume small mammals, birds, lizards, and invertebrates. They would also likely scavenge mice or other animals that succumbed to rodenticide. Because of their diets, secondary exposure to the rodenticide could be a risk to these birds. This species would need to eat 52.0% or more of their normal daily food intake to receive a potential lethal dose of brodifacoum by consuming invertebrates. For mammalian prey, only 0.52 oz. (14.9 g) of contaminated prey needs to be consumed for a potential lethal dose to an individual, which is only 24.6% of their estimated normal daily food intake (Table 3.16). Therefore, mortality from secondary exposure of this species could be high. However, although it was estimated that more than 20,000 rats were killed during the eradication project on Palmyra, only a few dozen rat carcasses were found during the project after the bait was applied because the animals often returned to their burrows before they died. (Pitt et. al. 2015). It is expected that this same result would occur on Sand Island, especially since mice are significantly smaller than rats and thus more difficult to detect. Thus, there may be few carcasses available for cattle egrets to find and consume. Based on the above data, it is possible that 20-80% of the egret population (17-67 birds) would be killed or receive sub-lethal doses by consuming brodifacoum intoxicated prey. There would also be a potential that brodifacoum would persist in the environment for at least a year and have impacts beyond the proposed operational window (Rueda et al. 2016; Pitt et al. 2015), such that any returning egrets to Sand Island would be at risk from secondary poisoning. Operational Hazard (Bird Strike and Noise/Human Disturbance) It would be unlikely ground or air operations would impact cattle egrets. Egrets are not known to breed on Sand Island and this species does not nest on the ground. Air operations have the potential

3-87 MIDWAY SEABIRD PROTECTION PROJECT FINAL ENVIRONMENTAL ASSESSMENT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

to disturb roosting egrets, but observations from field personnel during pre-project surveys indicate this disturbance would be minor and could cause birds to temporarily flush from their location. In addition, egrets do not fly very high, and thus have a negligible risk for collision potential with the helicopter. Calculations of mortality are based on the maximum number of birds observed in 2017. The potential mortalities that would be expected to result from Preferred Alternative represent insignificantly small numbers relative to overall populations of this species. We expect there could be some additional impacts to cattle egrets should the need arise to implement the rapid response portion of the refuge’s biosecurity plan and during implementation of the projects monitoring plan. A biosecurity plan has been in place on MANWR for many years and any additional bio-security measures would only occur if there was a new incursion of a non- native species. Biosecurity measures enacted would include rapid response to an incursion by non- native invasive species and implementing the monitoring plan would include follow-up observations and sample collecting (see Appendix A and B). Potential impacts such as causing birds to flush could occur from the additional human disturbance although this bird is a non-native invasive species. However, we expect any additional impacts to be minor, localized, and short- term. Mitigation No mitigation measures are planned for this species since a lethal control program is in place on MANWR.

3.3.6.11 Terrestrial Mammals

3.3.6.11.1 Existing Conditions No native terrestrial mammals are known from MANWR. No domestic animals or livestock are present on the island. Rats were previously documented on the 3 islands but were successfully removed and the islands were declared rat free in 1997 (USFWS, Unpublished b). Personnel from Island Conservation were unable to confirm the presence of rats using remote cameras and bait on either Sand or Eastern Island. If somehow rats have been re-introduced to Sand Island and are present in low numbers at the time of the eradication, the design of the mouse eradication would eliminate rats as well. House mice are presently the only known non-native mammal on the island.

3.3.6.11.2 Probable Impacts: No Action Alternative Under the No Action Alternative, Sand Island’s mouse population would not be the subject of an island-wide eradication project, but limited mouse control efforts would continue using the less toxic AGRID3 to protect nesting seabirds and for human health and safety. Therefore, impacts to the islands ecosystem, including negative impacts to vegetation and seabird populations, would continue.

3.3.6.11.3 Probable Impacts: Preferred Alternative Under Preferred Alternative, every mouse on the island would receive a primary or secondary exposure to the rodenticide brodifacoum and the house mouse would be eradicated from Sand Island.

3-88 FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

3.3.6.12 Herpetofauna

3.3.6.12.1 Existing Conditions The only known terrestrial reptiles or amphibians present on Sand Island are non-native and include the House (Hemidactylus frenatus), Indo-Pacific Gecko (Hemidactylus garnotii), Mourning Gecko ( lugubris), Penny Skink (Lampropholis delicata), and Blind Snake (Ramphotyphlops braminus).

3.3.6.12.2 Probable Impacts: No Action Alternative Under the No Action Alternative, Sand Island’s mouse population would not be the subject of an island-wide eradication project, but limited mouse control efforts would continue using the less toxic AGRID3 to protect nesting seabirds and for human health and safety. Therefore, there would be minor effects to any herpetofauna under the no-action alternative. Any negative effects of mice on herpetofauna populations on Sand Island would continue.

3.3.6.12.3 Probable Impacts: Preferred Alternative Primary Exposure No mortality of herpetofauna associated with the use of brodifacoum has been recorded to date, and one study indicates reptiles appear to be relatively insensitive to anticoagulant rodenticides (e.g., brodifacoum) compared to birds or mammals (Weir et al. 2015). This finding is supported by evidence from New Zealand, where brodifacoum has been extensively used in areas occupied by a wide range of herpetofauna, with no reports of reptiles or amphibians poisoned with brodifacoum (Eason and Spurr 1995). However, on Round Island , Telfair’s skinks (Leiolopisma telfairii) that ate rain-softened brodifacoum pellets broadcast in an eradication project were found dead (Merton 1987 in Eason and Spurr 1995). Telfairs’s skinks are known to eat seeds and fruit. Analyses of the skink carcasses revealed brodifacoum concentrations in the liver of 9.6 x 10-6 oz./lb. (0.6 mg/kg). Neither the Penny skink nor the house gecko would be likely to consume pellets. Both species eat invertebrates and the gecko also consumes sap or nectar. Because of the apparent relative insensitivity of reptiles to anticoagulants, it would be unlikely skinks, geckos or snakes would be killed by rodenticide poisoning. However, the situation with brodifacoum is less clear compared to diphacinone. However, when the above data are taken in total, the risk of killing skinks, geckos or snakes from brodifacoum poisoning would be low. If any mortality were to occur, population level effects are highly unlikely. Secondary and Tertiary Exposure While negative impacts to the skinks, geckos or snakes are unlikely, it is expected that these species would consume invertebrates that have fed on the pellets, which would then pose a tertiary poisoning risk to other animals that eat these animals as part of their diet. Because of reptiles’ relative insensitivity to anticoagulant rodenticide, they may be able to accumulate relatively high sub-lethal residues. On Pinzón Island, native lava lizards (Microlophus duncanensis) were found to maintain brodifacoum residues in their liver more than 800 days after the last application of a rat eradication project (Rueda et al. 2016). This resulted in the deaths of several raptors.

3-89 MIDWAY SEABIRD PROTECTION PROJECT FINAL ENVIRONMENTAL ASSESSMENT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

The Preferred Alternative would be unlikely to affect the herpetofauna on the island, and there are no mitigation measures proposed to reduce this impact. Any impacts would likely be minor, localized, and of little consequence to the populations. We expect there to be some additional impact to herpetofauna should the need arise to implement the rapid response portion of the refuge’s biosecurity plan and during implementation of the projects monitoring plan. A biosecurity plan has been in place on MANWR for many years and any additional bio-security measures would only occur if there was a new incursion of a non-native species. Biosecurity measures enacted would include rapid response to an incursion by non-native invasive species and implementing the monitoring plan would include follow-up observations and sample collecting (see Appendix A and B). Potential impacts such as trampling vegetation or the crushing of animals underfoot could occur with the additional human activities along with collection of some animals to test for brodifacoum residues; although these species are non-native. However, we expect any additional impacts to be minor, localized, and short-term.

3.3.6.13 Terrestrial Invertebrates

3.3.6.13.1 Existing Conditions Midway’s terrestrial invertebrate have been greatly altered by more than a century of human occupation. The majority of what we know about the terrestrial invertebrate community has focused on describing the species assemblage on Midway, and much of that assemblage work has focused on Arthropoda, and to a lesser extent Mollusca (Conant et al. 1983, Nishida and Beardsley 2002). Five hundred and fifty-two species of terrestrial invertebrates have been recorded thus far, of which 81% are introduced (Nishida and Beardsley 2002) and fill a wide range of habitat and functional niches in Midway’s terrestrial ecosystem, as well as food resources for some of the avifauna. The number of taxa observed on the atoll has grown rapidly over the past century and is speculated to have been aided by an increase in plant diversity and ease of migration (Nishida and Beardsley 2002). In the main Hawaiian Islands, Shiels et al. (2013) found that over half the diet of feral mice consisted of , with caterpillars making up the vast majority of the arthropod prey (~94%), followed by ants, fly larvae, burrowing bugs, and spiders. These results were in contrast with larger rodents (rats) at the same locations, which fed on a wider variety of arthropods, including cockroaches, Katydids, and honey bees (Shiels et al. 2013).

3.3.6.13.2 Probable Impacts: No Action Alternative Under the No Action Alternative, Sand Island’s mouse population would not be the subject of an island-wide eradication project, but limited mouse control efforts would continue using the less toxic AGRID3 to protect nesting seabirds and for human health and safety. Therefore, there would be minor effects to terrestrial invertebrates under the no-action alternative. Any negative effects of mice on invertebrate populations on Sand Island would continue.

3.3.6.13.3 Probable Impacts: Preferred Alternative Terrestrial invertebrates can accumulate anticoagulant rodenticide residues. It has been suggested that anticoagulant rodenticides, such as diphacinone and brodifacoum, are not likely to affect

3-90 FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

invertebrates, because their blood clotting mechanisms are different from those of vertebrates (Shirer 1992 in Eason and Spurr 1995). However, the toxicity of anticoagulants may differ across groups of invertebrates. Brodifacoum is highly toxic to Daphnia magna, an aquatic invertebrate, with an EC50 of 0.98 ppm after 48 hours (USEPA 1991). Most hard-bodied terrestrial invertebrates (e.g., crabs, cockroaches, beetles) appear to be relatively insensitive (Booth et al. 2003, references therein; Morgan et al. 1996; but see Pitt et al. 2015). There is evidence that brodifacoum may be more toxic to soft-bodied terrestrial invertebrates than hard-bodied species (Booth et al. 2003). In a laboratory study, 2 species of land snail (Pachnodus silhouettanus, Achatina fulica) died in 72 hours when exposed to doses ranging from 3.5 x 10-7 oz. to 1.1 x 10-6 oz. (0.01 to 0.04 mg) (Gerlach and Florens unpubl., Booth et al. 2003). In another lab study, brodifacoum equivalent to 8.0 x 10-5 to 1.6 x 10-4 oz./lb. (5 to 10 mg a.i./kg) was found to cause 100% mortality of earthworms (Apporectodea calignosa) (Booth et al. 2003). However, causing this level of mortality required grinding pellets and mixing them with the soil, a scenario that would be unlikely to occur in the field. In general, most invertebrate species are not known to be susceptible to toxic effects from the use of brodifacoum in the field (Booth et al. 2003, Hoare and Hare 2006). Preferred Alternative would be unlikely to directly affect any terrestrial invertebrates on the island and there are no mitigation measures planned to reduce impacts to invertebrates. Removal of mice from Sand Island would eliminate mice predation on invertebrates, and thus potentially could have a beneficial effect to these species. However, personnel are currently conducting studies to monitor the ecosystem (including the invertebrate community) prior to and post mouse eradication. Therefore, any negative affects to terrestrial invertebrates will be documented. We expect there to be some minor additional impacts to terrestrial invertebrates should the need arise to implement the rapid response portion of the refuge’s biosecurity plan and during implementation of the projects monitoring plan. A biosecurity plan has been in place on MANWR for many years and any additional bio-security measures would only occur if there was a new incursion of a non-native species. Biosecurity measures enacted would include rapid response to an incursion by non-native invasive species and implementing the monitoring plan would include follow-up observations and sample collecting (see Appendix A and B). Potential impacts such trampling of vegetation and soil could occur from the additional human disturbance along with the collection of some animals to test for brodifacoum residues. However, over 80 percent of terrestrial invertebrates on MANWR are non-native. In summary, we expect any additional impacts to invertebrates to be minor, localized, and short-term.

3.3.6.14 Marine Fish

3.3.6.14.1 Existing Conditions A total of 266 species of fish, including 7 pelagic species, have been recorded at Midway. Some of these species are rare or absent from the main Hawaiian Islands. Despite its low species diversity, Midway’s reef fish biomass is higher than in the main Hawaiian Islands, largely due to lower fishing pressures. Midway and its neighboring atolls have the highest rates of endemic reef within the archipelago, with up to 52% of all fish observed being endemic. Many Midway species grow to larger than average size. All trophic levels are well represented, including jacks

3-91 MIDWAY SEABIRD PROTECTION PROJECT FINAL ENVIRONMENTAL ASSESSMENT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

and 4 species of sharks. Several species of fish found elsewhere only in deep waters are found at shallow diving depths at Midway, including the endemic Hawaiian black grouper (hapu‘upu‘u). One alien algae, one alien fish, the blueline snapper (Lutjanus kasmira), and 4 alien marine invertebrate species are established at Midway as found in 2000- 2003 surveys. Incidental observations of 2 other introduced species, blacktail snapper (Lutjanus fulvus) and bluespotted grouper (Cephalopholis argus), have occurred at Midway in the last decade. A total of 111 species of reef fish were observed during the 2008 surveys at Midway. Ten families represent the majority of the species of reef fish at Midway: Acanthuridae, Kyphosidae, Scareidae, Mullidae, Pomacentridae, labridae, Carcharhinidae, lutjanidae, Holocentridae and Oplegnathidae. The Kyphosids, or chubs, and the Acanthurids, or surgeon fish and tangs, are the most abundant and represent the highest biomass. Parrotfish (Scaridae) represented the third largest contributor to total fish biomass. Reef fish biomass density was higher in the forereef environment compared to the backreef and lagoon and was slightly higher in the south and southeast quadrants of the atoll. Of the larger bodied reef fish greater than 50 cm in length, the fishes (family scaridae) were the most abundant, composing 44% of the individual reef fish counted, and had the highest biomass density at 33%. Acanthurids represented the highest biomass in the backreef and lagoon habitats, and was third to sharks and chub, in the forereef habitat. Large jacks (Caranx ignobilis) are also known to frequent the forereef and reef crest environments.

3.3.6.14.2 Probable Impacts: No Action Alternative Under the No Action Alternative, Sand Island’s mouse population would not be the subject of an island-wide eradication project, but limited mouse control efforts would continue using the less toxic AGRID3 to protect nesting seabirds and for human health and safety. No bait would be used in the vicinity of the marine environment; therefore, there would be no effects to marine fish or the marine ecosystem under the no-action alternative.

3.3.6.14.3 Probable Impacts: Preferred Alternative Primary Exposure The likely pathways for contamination of fish would be through primary and secondary exposure. A potential, but unlikely, pathway would be absorption through skin or gills (Empson and Miskelly 1999). Bait pellets dissolve quickly in the near shore environment, often within 15 to 30 minutes, and the concentration of rodenticide in sea water would be at undetectable levels and would pose very low, if any risk to fish. In a study on the Island of Lehua, Hawaiʻi, of the 48 species of fish documented to occur around the island, 29 were found to consume inert bait pellets; another 7 species made contact with the bait but did not consume it (Mazurek 2015). In this trial with inert baits, pellets were applied at a rate equivalent to 653 lb./ac. (733 kg/ha), which is more than 10 times higher than the maximum rate proposed for this eradication project. In a similar trial, in New Zealand, 3 species of fish were seen to eat non-toxic baits (Empson and Miskelly 1999). A recent laboratory study by the USGS showed that black triggerfish (Rhinecanthus aculeatus), smallmouth bass (Micropterus dolomieu) and fathead minnows (Pimephales promelas) refuse to eat bait pellets containing diphacinone (USGS Columbia Environmental Research Center). However, during the second 2017 Lehua rat eradication effort, fish were caught from shore by line

3-92 FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

and hook baited with shrimp on the day of each bait application; fish were euthanized, and gut contents examined for signs of bait material and the pyranine biomarker fluorescence. Bait material and/or the biomarker were observed in specimens of pinktail triggerfish (Melichthys vidua), black triggerfish, stocky hawkfish (Cirrhitus pinnulatus), and blue-lined snapper. No evidence of bait consumption was found in blacktail snapper or blotcheye soldierfish (Myripristis berndti) caught (S. Siers, personal communication, USDA NWRC 2016). When diphacinone (bait or a.i.) is administered by gavage, the fish rapidly regurgitate the material (R. Riegerix, personal communication, USGS/University of Missouri-Columbia 2016), which indicates that some fish species can detect and avoid bait containing diphacinone. Results from aquarium trials associated with the Kapiti Island, New Zealand rat eradication found similar results using brodifacoum (Empson and Miskelly 1999). Three species of fish, variable triplefin (Forsterygion varium, 24 individuals), spotty (Notolabrus celidotus, 30), and blue cod (Parapercis colias, 6) were presented with bait pellets containing 0.002% brodifacoum. Of these 60 fish, only 6 spotties were seen to eat the bait, with one dying from brodifacoum poisoning. Several fish died that did not eat the bait, and no brodifacoum residues were found in their livers. The authors speculated the fish may have absorbed the chemical through their skins or gills. When small quantities of brodifacoum bait entered the marine environment during aerial application operations on Anacapa Island and Isabel Island, Mexico, no marine organisms (fish or invertebrates) were observed to consume the bait (Howald et al. 2009, Samaniego-Herrera et al. 2014). However, in the Kapiti Island trials, 3 species of fish were seen to eat the non-toxic bait within 15 minutes of entering the marine environment (Empson and Miskelly 1999). In summary, some species of fish may feed on pellets on the day of application. Whether the consumption is abundant or prolonged enough to cause illness or mortality to some fish remains unknown. The rapid dissolution of unconsumed pellets in water would prevent the prolonged feeding on pellets. Given the relatively small amount of bait that would be expected to enter the marine environment, the rapid dissolution of pellets, and the fact that some fish appear to avoid eating bait pellets with active rodenticide, it appears consumption of rodenticide baits for some species of fish would be minor and unlikely to have adverse affects on their populations. Secondary Exposure Secondary exposure of fish to brodifacoum around Sand Island would be a potential risk. However, on the Anacapa Island rat eradication project, 6 species of intertidal organisms (fish and invertebrates) were sampled 15, 30, and 90 days post-application, and all samples tested negative for brodifacoum residues (Howald et al. 2009). In addition, 2 studies conducted in conjunction with the 2008 Mōkapu and 2009 Lehua rat eradication projects using diphacinone indicate that the risk would be extremely low (Orazio et al. 2009, Gale et al. 2008). Within days of the last aerial broadcast of diphacinone pellets, samples were collected of fish, invertebrates, and seawater, and none of the species sampled showed detectable levels of diphacinone in their tissues. However, brodifacoum residues have been found in fish tissues after rat eradications. A review of rat eradication projects using brodifacoum found marine residue monitoring and analysis had been conducted in 10 applications between 1997 and 2011 (Masuda et al. 2015). Of the 10 applications, 1 detected brodifacoum residues in fish. Two of 65 fish sampled had residues, with concentrations 0.026 and 0.092 ppm in the liver, which exceeded the 96-hour LC50 for rainbow trout (Oncorhynchus mykiss) (0.015 ppm) and bluegill sunfish (Lepomis macrochirus) (0.026 ppm)

3-93 MIDWAY SEABIRD PROTECTION PROJECT FINAL ENVIRONMENTAL ASSESSMENT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

(USEPA 1991). Following the rat eradication on Palmyra Atoll, rodenticide residues were detected in all fish samples collected from the lagoon which included mullet fishes (Moolgarda engeli and Liza vaigiensis) and one puffer fish. Fish were found dead and collected opportunistically for this study (Pitt et al. 2015). Mullet fish contamination ranged from 0.058–1.160 ppm (mean=0.337 ppm) and the single puffer fish (family Tetraodontidae) sample had 0.438 ppm of brodifacoum in homogenized tissue. However, a number of factors will reduce the likelihood of marine secondary exposure around Sand Island relative to Palmyra including: (i) the shape of Sand Island has less coastline per acre of treated area; (ii) the much broader beaches with a lack of vegetation on Sand Island and few coastal trees to block the view of the coastline by the helicopter pilot; (iii) the use of hand-broadcast of bait in some areas without beaches or with armoring on the shoreline and; (iv) greater lagoon flushing at MANWR because land doesn’t protect the lagoon as it does on Palmyra. The above examples demonstrate that marine fish can be contaminated during an aerial application of brodifacoum. However, it also demonstrates that this occurs only infrequently. It is also noteworthy that the very large amount of brodifacoum used on Palmyra Atoll at 75.6 lb./ac. (84.8 kg/ha) and 71.5 lb./ac. (80.1 kg/ha), for the first and second application respectively, was unprecedented, which likely influenced the available brodifacoum residues consumed by non- target species. The bait density used for Palmyra was 10-16% higher than that proposed for Sand Island. Following the accidental brodifacoum spill in New Zealand, fish samples were collected, and only one individual fish had detectable rodenticide residues (Primus et al. 2005). However, this site had a higher energy coastline with greater wave action than that on MANWR. In the above cases involving brodifacoum, it is unclear whether the fish exposure was primary or secondary. In summary, the main secondary exposure risk from anticoagulant rodenticides would be from the consumption of fish that had died or had recently ingested toxicants, and therefore contained a high level of toxin, or from contaminated invertebrates. If any fish do die from primary exposure, those carcasses may pose secondary hazard to other species that may consume them. However, animals that survive direct ingestion would rapidly metabolize (24-48 hrs) the vast majority of the toxin, and residues in liver (and far lower in muscle tissue) are likely to be too low to be biologically significant. Based on the above data, it would be unlikely that a sufficient quantity of brodifacoum would enter the water, or that the pellets would remain intact, in the environment, long enough to present an absorption or primary poisoning risk to any fish populations. In addition, the bulk of the reef around MANWR Island is not in close proximity to Sand Island. Therefore, although some reef fish near the island may be affected, the majority of the populations around the atoll would unlikely to be affected. Some individual fish may be at risk if they were to ingest pellets. The low levels of residues found are not likely to be biologically significant with respect to secondary exposure or bioaccumulation, especially in such a diffuse environment. It also would be unlikely that the rodenticide would contaminate sufficient prey of fish to pose a secondary poisoning risk to any fish populations. Therefore, it would be unlikely that Preferred Alternative would directly or indirectly impact any fish populations. Minor direct and indirect impacts may occur to some individual fish. Besides the multiple steps being taken to minimize bait pellets entering the marine environment (see below), there are no other mitigation measures proposed for marine fish.

3-94 FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

We expect there to be no additional impact to marine fish from the implementation of the projects monitoring plan. Implementing the monitoring plan would include follow-up observations and sample collecting. Some game fish may be collected for the sampling of brodifacoum residues. Additional impacts to marine fish are not expected from implementation of the bio-security plan or monitoring plan since these activities will mostly occur in the terrestrial environment (see Appendix A and B). Therefore, we expect no potential impacts to marine fish from the additional terrestrial human disturbance, and the number of fishes collected will be small. Mitigation To minimize bait from entering the marine environment, prior to the application of bait pellets with rodenticide, the bait delivery system (bait bucket, controller, GPS units, and helicopter) would be tested and calibrated to ensure an accurate application rate. An onboard computer linked to a GPS and light bar would guide the pilot along pre-programmed flight lines over the island at a prescribed airspeed, which would facilitate application of bait over the terrestrial environment only. Aerial application of bait pellets would not occur during wind speeds in excess of 35 mph (56 km/hr), or when heavy rains are forecast to occur within 72 hours. While 35 mph will be a maximum sustained wind for operations, we will be looking at wind in conjunction with other weather conditions and operational hazards before conducting flights. Health and human safety will be of the utmost importance in this project, and we will use best practices from previous projects and take all conditions and expert recommendations into consideration before flying. With those considerations in mind, we may use a much slower threshold for baiting activities. In addition, for areas near the shoreline, the bait hopper would be fitted with a deflector that spreads bait out to only one side (approx. 120° angle) to minimize bait application directly into the water. For narrow projecting structures such as the outer harbor breakwater, the area will be baited by hand. Every reasonable effort would be made to minimize the amount of bait drift into the water; however, it is expected that some bait will enter the ocean. The pilot and on-the-ground observers would visually monitor the application of bait, and if a malfunction were detected, operations would cease until the problem is corrected. Lastly, bait would be applied at the lowest rate possible to achieve eradication, and any bait spills above a defined threshold would be collected and disposed of according to label instructions.

3.3.6.15 Coral

3.3.6.15.1 Existing Conditions MANWR is one of the northernmost coral atolls in the world, presenting a unique opportunity to study the effect of colder waters on the growth, development, and ecology of coral reefs. Its neighbor, Kure Atoll, is the northernmost atoll in the world. MANWR drops off steeply outside the barrier reefs, making it possible to observe in a relatively small area the different organisms and communities associated with pelagic, reef crest, ocean facing reef slope, deep reef, and lagoon habitats. The lagoon is filled with dense networks of linear reticulated and circular reefs that trap sand washed over the northeastern reef rim. As in many atoll lagoons, sediments limit coral growth on Midway, except in the deeper central lagoon, where a modest amount of finger coral gardens still exists. Meadows of seagrass are common in the lagoon, as are rock-boring urchins, calcareous

3-95 MIDWAY SEABIRD PROTECTION PROJECT FINAL ENVIRONMENTAL ASSESSMENT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

green algae, and brown turban algae. The deep southern ship channel between the ocean and lagoon was dredged during the World War II era and has substantially modified circulation and lowered lagoon water levels. Together with lagoon reefs, these changes reduced or blocked water circulation in much of the lagoon and created higher levels of turbidity. In 2000 and 2004, coral bleaching episodes were reported at Midway, as well as at the neighboring atolls of Kure, Pearl, and Hermes. Lagoon lobe and finger corals have declined during the past decade, although blue encrusting coral continues to thrive. Massive spurs and grooves consisting mostly of coralline algae face the open ocean along the northwest to southwest perimeter reefs and protect the atoll from heavy wave action common during the winter months. These massive reefs offer evidence of the importance of coralline algae as a major reef builder in the far end of the Northwestern Hawaiian Islands. Corals on ocean-facing reef habitats are generally not as abundant compared to neighboring atolls to the southeast but are common in a few sheltered reefs and especially on shallow back-reefs and lagoon pinnacles. High concentrations of the rock-boring urchin are presently eroding much of the shallow perimeter reef crests dominated by coralline algae. Although not grazing corals directly, the sea urchins are hollowing out the dead interior skeletons of living lobe corals and undermining other attached corals. A total of 32 species of stony coral have been recorded at Midway, mostly Pocillopora, Porites, and Montipora, plus one zoanthid soft coral, Palythoa. Blue encrusting coral tentatively identified as Montipora cf. turgecens occurs in spectacular formations in the lagoon and back reef habitats and may be endemic to the Northwestern Hawaiian Islands. Coral cover around MANWR varies with both habitat and within region of the atoll. Hard coral cover was generally low island-wide, averaging around 2.45% and never exceeding 20%. The highest coral cover was generally located along several sections of the northern and eastern backreef, where coral assemblage was primarily composed of Montipora, Pocillopora, and Porites spp. Mean coral cover was low inside the lagoon (6.6 ± 3.0%) and on the forereef (4.0 ± 1.5%), while mean cover was moderately high in the backreef (32.3 ±14.2%) (NOAA CREP 2008). Coral community structure varied both between and within habitats. Two northern backreef sites consisted of large encrusting colonies of 3 Montipora species. The eastern backreef site was characterized by scattered Porites lobata and Pocillopora heads. On the forereef, the western area was scoured and corals were quite depauperate, most likely due to strong wave action. The southern forereef south of Sand Island is composed largely of Porites sp., while the forereef south of Eastern Island is composed of mostly Pocillopora sp. The coral communities at 2 lagoon patch reefs, one in the southeastern portion, and the other in the southwest (MID-03 and MID-H11, Figure 3.11), were quite different with one dominated by old Porites compressa mounds (MID-03, Figure 3.11) and the other (MID-H11, Figure 3.11) consisting of scarce Pocillopora heads.

3-96 FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

Figure 3.11 Spatial Distribution of Benthic Cover and Coral Composition at NOAA CREP Rapid Ecological Assessment Sites at Midway in 2008

Source: NOAA (2008)

3.3.6.15.2 Probable Impacts: No Action Alternative Under the No Action Alternative, Sand Island’s mouse population would not be the subject of an island-wide eradication project, but limited mouse control efforts would continue using the less toxic AGRID3 to protect nesting seabirds and for human health and safety. No bait would be used in the vicinity of the marine environment; therefore, there would be no effects to corals or the marine ecosystem under the no-action alternative.

3.3.6.15.3 Probable Impacts: Preferred Alternative Primary and Secondary Exposure As described above, a variety of corals inhabit the waters around MANWR, and there would be potential for these organisms to be exposed to rodenticide that inadvertently reached the marine environmental and dissolved in the water. However, the level of bait entering the nearshore environment would be very low, pellets would dissolve quickly in the nearshore environment, often within 15-30 minutes, and the rodenticide would be diluted rapidly, disperse and sink to the ocean bottom and start decomposing (see Section 3.3.2). In addition, as described below under essential fish habitat, the nearshore, directly adjacent to Sand Island and the project area, is depauperate of corals. The benthic substrate, adjacent to the mouse eradication project area is categorized as sand, hardbottom-uncolonized, pavement, pavement with sparse algae, and pavement-uncolonized (see Figure 3.11). The only documented toxic effect of brodifacoum is that it prevents the production of vitamin K coagulating factors in mammals. These are “poisons” only due to the anticoagulant property and invertebrates do not have the same blood clotting system as mammals. There are no data to indicate corals have been impacted by anticoagulant rodenticides from previous eradication projects. On Palmyra Atoll, no impact to corals was documented after the application of bait containing brodifacoum applied using 2 applications at the rate of 75.6 and 71.5 lbs./ac. (84.8 and 80.1 kg/ha), which is 10-16% more than the application rate proposed for MANWR (65 lbs./ac. or 73 kg/ha).

3-97 MIDWAY SEABIRD PROTECTION PROJECT FINAL ENVIRONMENTAL ASSESSMENT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

Based on the above data, it would be unlikely that a sufficient quantity of brodifacoum would enter the water, or that the pellets would remain intact, to present an absorption or primary poisoning risk to any corals. In addition, no short- or long-term discharge of brodifacoum-impacted groundwater into the marine environment is anticipated (see Section 3.3.2). Therefore, it would be unlikely that the Preferred Alternative would impact any corals. We expect there to be no additional impact to coral from the implementation of the projects monitoring plan. Implementing the monitoring plan would include follow-up observations and sample collecting. Additional impacts to coral are not expected from implementation of the bio- security plan or monitoring plan (see Appendix A and B) since these activities will mostly occur in the terrestrial environment. Mitigation The same minimization measures outlined under marine fish would also minimize the amount of rodenticide entering the marine environment.

3.3.6.16 Essential Fish Habitat

3.3.6.16.1 Existing Conditions The marine water column and seafloor of MANWR, adjacent to the proposed project area for the mouse eradication, is identified as Essential Fish Habitat (EFH). The EFH is said to support various life stages for Management Unit Species (MUS) identified under the Western Pacific Regional Fishery Management Council’s Pelagic and Hawaiʻi Archipelago Fishery Ecosystem Plans. Of key relevance to this project is the designation of EFH for Coral Reef Ecosystems, including the water column and all bottom down to 328 ft. (100 m) depth from the shoreline out to the exclusive economic zone (EEZ) boundary (NOAA 2013). The species composition, abundance and condition of the corals at MANWR vary depending on location within the atoll (see Figure 3.12). The nearshore, directly adjacent to Sand Island and the project area, is depauperate of corals. The benthic substrate, adjacent to the mouse eradication project area is categorized as sand, hardbottom-uncolonized, pavement, pavement with sparse algae, and pavement-uncolonized (see Figure 3.11). The benthic maps show only 2 small areas in the marine environment, close to the project area, that are categorized as pavement with >10% live coral (see Figure 3.12).

3-98 FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

Figure 3.12 NOAA Midway Island Detailed Habitat Cover Image at 13m Resolution

Source: National Centers for Coastal and Ocean Science (2017)

Sand Island is located in the southern portion of the atoll and is nearest to the southern forereef. The high level of wave energy on the southern shore of Sand Island makes the area adjacent to the project area poor coral habitat. The seawall area of the island was added by the Navy during the Cold War (circa 1957-58) to lengthen the runway, and the nearshore environment was damaged by dredging and filling activities, and additional acres were added to the island (Figure 3.13). The majority of the nearshore habitat is categorized as reef flat (semi-exposed area between the shoreline intertidal zone and the reef crest of a fringing zone), though the boundary between land and reef flat has become less obvious as the sheet pile has eroded and vegetation and seabird nesting areas have appeared. Debris is present in the majority of the benthic habitat adjacent to the sheet pile seawall (USFWS PIFWO 2016). The hard substrate of the seawall sheet pile walls and areas where riprap has been placed to control erosion are artificial habitat, and have more abundant coral colonization than the sand, and pavement environments (USFWS PIFWO 2016). A marine survey of the seawall portion of the adjacent marine environment was conducted by the Pacific Islands Fish and Wildlife Office and NOAA staff in February 2013, and again in April 2016, to determine presence and density of protected species and sensitive habitat for the seawall repair project (Figure 3.14) (Godwin 2013, Klavitter 2013, USFWS PIFWO 2016). The 2016 seawall survey covered nearshore areas, including the full length of seawall that may be included in future seawall repair efforts. Within the survey area, 97 reef fish, 9 coral, 32 non-coral macro- invertebrates, and 28 algae species were identified. The 9 coral species identified were from the families Acroporidae, Faviidae, Pocilloporidae, and Poritidae, with lobate, encrusting, and branching morphologies present. These species were found in low densities, with an average of

3-99 MIDWAY SEABIRD PROTECTION PROJECT FINAL ENVIRONMENTAL ASSESSMENT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

0.77 colonies per 10.8 ft.2 (0.77/m2) and had a maximum diameter of approximately 19.7 in. (50 cm). Corals were found on both benthic and artificial substrate. No ESA-listed coral species were found during this survey, and there are no reports of listed corals occurring at Midway or within the Hawaiian Archipelago. During a seawall site visit in June 2017, sites seaward of the repair areas were scouted and assessed for suitability as coral translocation sites (Figure 3.15). These sites are representative of the patch reefs in the area. Transplant site Y had a total of 93 corals from 3 species, 2 from the genus Pocillopora and one Porites, within a 2,152 ft.2 (200 m2) area. Transplant site Z had a total of 79 corals from 2 species in the genus Pocillopora within a 1,076 ft.2 (100 m2) survey area. These sites are small areas of raised structure that are suitable for corals to persist. There is very little structure suitable to support live coral in the area, and what is colonized has a low abundance of live coral, as is represented by sites Y and Z. Sites Y and Z are over 500 ft. (157 m) away from the seawall, and well outside of the mouse eradication project area. The area inshore from sites Y and Z, along the length of the seawall, were scouted for coral transplant suitability, and were categorized as unfit, due to the lack of appropriate structure and live coral.

3-100 FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

Figure 3.13 Augmentation of Sand Island during the Cold War

An airfield was added to Sand Island and the coastline along the southern and southeastern shore was hardened with a seawall. Source: National Centers for Coastal and Ocean Science (2017)

3-101 MIDWAY SEABIRD PROTECTION PROJECT FINAL ENVIRONMENTAL ASSESSMENT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

Figure 3.14 Marine Survey Area for Seawall Repair Project

Areas of surveyed (orange) and interpolated (green) project area. “Target Area” identifies anticipated area of direct impact for the Seawall Repair Project. (USFWS PIFWO 2016)

Figure 3.15 Possible Coral Translocation Site for the Seawall Repair Project

Possible coral translocation site for the seawall repair project, located on the southern shore of Sand Island over 500 ft. (157 m) from the seawall. Source: USFWS (2017)

3-102 FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

The back reef and reef crest habitat has a higher occurrence, abundance and species diversity of corals than the reef flat and lagoon habitats. On the southern shore of Sand Island, except for the area directly off of “Bulky Dump,” the closest back reef and reef crest habitat is over 600 ft. (180 m) from shore. The area adjacent to the shoreline is sand or un-colonized pavement with few scattered coral heads. This area is subject to tidal exposure, further reducing the number of live coral present. The “Bulky Dump” area is categorized as pavement with sparse algae and is the closest point to the southern reef crest, at approximately 90 ft. (27 m). Excluding the seawall that has been addressed above, very few (if any) corals occur within a 30 ft. (10 m) zone along the southern shoreline of Sand Island. The northern shoreline of Sand Island is predominately sand, and the environment immediately adjacent to the shore is sandy lagoon habitat devoid of coral. Very few (if any) corals grow within the 33 ft. (10 m) marine zone. The high coral cover areas of the lagoon occur out at the northern reef crest and back reef at site such as MID-01, MID-H21 and MID-156 (Figure 3.11). The high coral cover areas of the back reef and reef crest on the Northern and Northwestern sides of the island are over 5,000 ft. (1.5 km) from shore.

3.3.6.16.2 Probable Impacts: No Action Alternative Under the No Action Alternative, Sand Island’s mouse population would not be the subject of an island-wide eradication project, but limited mouse control efforts would continue using the less toxic AGRID3 to protect nesting seabirds and for human health and safety. No bait would be used in the vicinity of the marine environment; therefore, there would be no effects to essential fish habitat under the no-action alternative.

3.3.6.16.3 Probable Impacts: Preferred Alternative Primary Exposure Based on the above data, it would be unlikely that a sufficient quantity of brodifacoum would enter the water, or that the pellets would remain intact, and in the environment, long enough to present a poisoning risk to essential fish habitat. Some dissolved particles of brodifacoum would likely settle in the benthic marine environment and there is uncertainty how some organisms may be affected by trace amounts. However, since the toxic effect of brodifacoum is the prevention of the production of vitamin K coagulating factors in mammals, the effects would likely be negligible. Marine invertebrates do not have the same blood clotting systems as mammals and documented evidence show marine invertebrates such as crabs with no ill effects. As stated in the previous sections, the Preferred Alternative is unlikely to have any effects to coral, non-coral macro marine invertebrates (see Section 3.3.6.17), or algae (see Section 3.3.6.18) and only minor effects to individual reef fish. Since broadifacoum can stay active to ~6 months, there is a possibility that there could be sufficient time for the toxin to percolate into the groundwater aquifer of Sand Island and possibly discharge active grains of ingrediant that settle on the benthic marine environment. However, as discussed in Section 3.3.2, should brodifacoum dissolved at low concentrations (it is nearly insoluble) enter the groundwater aquifer, it would be degraded by bacteria in the subsurface, be diluted, and take years or decades for that water to discharge into the marine environment. Due to saturation with percolating rainwater, groundwater flows down and outward towards the shore, where it eventually

3-103 MIDWAY SEABIRD PROTECTION PROJECT FINAL ENVIRONMENTAL ASSESSMENT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES discharges into the ocean. However, the gradients and flux are small (GeoEngineers 2011). At Wake Island it was assessed that it would take 30 to 50 years for groundwater to migrate from the center of the island to either shore (a distance of approximately 1,000 ft. [305 m]) (AFCEE 2006). Residence times at Sand Island could be as much as 3 times longer because it is roughly 6,000 ft (1,828 m) across. Hydrogeological studies at Wake Island found that groundwater chemistry and temperature conducive to the biodegradation of organic compounds like brodifacoum, similar conditions are present at Sand Island. The Wake Island study also found that significant dilution, driven by vertical groundwater movement from tidal fluctuations, would occur (AFCEE 2006). Therefore, no short- or long-term discharge of brodifacoum-impacted groundwater into the marine environment is anticipated. Therefore, it would be unlikely that the Preferred Alternative would directly or indirectly impact essential fish habitat. Multiple steps are being taken to minimize the number of bait pellets entering the marine environment. Any impacts would likely be minor and localized. In summary, the Preferred Alternative would likely have some adverse effects on essential fish habitat, but those effects would be minor and temporary. We expect there to be no additional impact to essential fish habitat from the implementation of the projects monitoring plan. Implementing the monitoring plan would include follow-up observations and sample collecting. Additional impacts to essential fish habitat are not expected from implementation of the bio-security plan or monitoring plan since these activities will mostly occur in the terrestrial environment (see Appendix A and B). Mitigation The same minimization measures outlined under marine fish would also minimize the amount of rodenticide entering the marine environment.

3.3.6.17 Marine Invertebrates

3.3.6.17.1 Existing Conditions The first systematic marine invertebrate survey was conducted at Midway in 1997. It documented 316 invertebrate species, 250 of which had not been previously recorded at Midway. Crustaceans were the dominant macroinvertebrates, composing 46% of the total species. The , Echinostrephus aciculatus, Echinometra mathaei, and Heterocentrotus mammilatus, are the most abundant benthic invertebrate in all 3 habitats (forereef, backreef, and lagoon.) The boring sea urchin (Echinometra mathaei) is the most abundant non-coral invertebrate present on the forereef, with over 73,000 individuals observed during towed diver surveys, and has its highest densities along the southern shore. Free urchins were relatively uncommon, with all records noted within the lagoon and backreef environments. Trapezid crabs and holothuroids or sea cucumbers, Actinopyga obesa and Bohadaschia paradoxa, are abundant within the lagoon at Midway. Sea cucumbers the second most abundant group of mobile invertebrate present on the forereef, with approximately 1,000 individuals counted during towed diver surveys, with the highest abundance present on the northern and southern backreefs.

3-104 FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

3.3.6.17.2 Probable Impacts: No Action Alternative Under the No Action Alternative, Sand Island’s mouse population would not be the subject of an island-wide eradication project, but limited mouse control efforts would continue using the less toxic AGRID3 to protect nesting seabirds and for human health and safety. No bait would be used in the vicinity of the marine environment; therefore, there would be no effects to marine invertebrates under the No Action Alternative. Any negative effects of mice on marine invertebrates or the intertidal and nearshore ecosystems would continue.

3.3.6.17.3 Probable Impacts: Preferred Alternative Primary or Secondary Exposure Based on the above data, it would be unlikely that a sufficient quantity brodifacoum would enter the water, or that the pellets would remain intact long enough, to present an absorption or primary poisoning risk to any marine invertebrates. As noted above (see Section 3.3.6.14), samples of fish, invertebrates (crabs and limpets), and seawater collected days after aerial application of diphacinone on Mōkapu and Lehua showed no detectable levels of diphacinone in their tissues (Orazio et al. 2009, Gale et al. 2008). Brodifacoum residues have been found in invertebrate tissues after rat eradications. A review of rat eradication projects using brodifacoum found marine residue monitoring and analysis had been conducted in 10 applications between 1997 and 2011 (Masuda et al. 2015). Of the 10 applications, 3 detected brodifacoum residues in invertebrates. Of the 196 invertebrates sampled, 11 had residues detected in their tissue, with concentrations ranging from 0.001 to 0.022 ppm (mean =0.008 ppm). After the rat eradication on Palmyra, brodifacoum residues were found in fiddler crabs (Uca tetragonon) (Pitt et al. 2015). In general, invertebrates appear to be relatively insensitive to anticoagulant rodenticides (Pain et al. 2000, Hoare and Hare 2006). However, on Palmyra Atoll, some fiddler crabs may have died from brodifacoum poisoning in conjunction with the rat eradication (Pitt et al. 2015). In addition, the lagoon environment on Palmyra is not analogous to the conditions on Sand Island, which experiences a higher level of water exchange due to how the channel was dredged into the atoll lagoon by the Navy to facilitate boat access. Primus et al. (2005) suggest mortality of marine invertebrates may have occurred as a result of a large spill of rodenticide containing brodifacoum. However, this situation would not be representative of how bait might enter and interact with the nearshore environment in the Preferred Alternative. Therefore, it would be unlikely that the Preferred Alternative would impact any marine invertebrates. The minimization measures outlined for marine fish are not a factor in this determination. We expect there to be no additional impacts to marine invertebrates from the implementation of the projects monitoring plan. Implementing the monitoring plan would include follow-up observations and sample collecting. Some mollusks may be collected for the sampling of brodifacoum residues. Additional impacts to marine invertebrates are not expected from implementation of the bio-security plan or monitoring plan since these activities will mostly occur in the terrestrial environment (see Appendix A and B). In addition, the number of marine animals collected will be small.

3-105 MIDWAY SEABIRD PROTECTION PROJECT FINAL ENVIRONMENTAL ASSESSMENT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

Mitigation The same minimization measures outlined under marine fish would also minimize the amount of rodenticide entering the marine environment. In addition, no short- or long-term discharge of brodifacoum-impacted groundwater into the marine environment is anticipated (see Section 3.3.2).

3.3.6.18 Algae

3.3.6.18.1 Existing Conditions More than 100 species of algae are known to exist on Midway, including 35 previously unrecorded species at Midway and 1 seaweed species new to science, Dudresnaya babbittiana. Algal communities contribute to the health of coral reef ecosystems, providing trophic resources for grazers and herbivorous fish (Gattuso et al. 1998), as well as contributing to coral reef productivity (Rowan 1998). The transfer of nutrients from algae to coral occurs when fish or invertebrates forage on primary producers and excrete on the coral reef (Birkeland and Grosenbaugh 1985). The Algae Benthic communities around MANWR are composed primarily of turf and macroalgal functional groups with a combined total of 23 species of macroalgae observed (7 chlorophytes, 5 ochrophytes, 11 rhodophytes) from CREP sites surveyed (NOAA CREP 2008). Macroalgae cover averaged 21.1% island-wide and was generally highest along the eastern backreef areas (average 35.4%). In 2008, 2 towed-diver surveys approximately 5 mi (8 km) to the north of East Island encountered a boodlea algae bloom, with several segments recording 100% cover of the benthos in a layer up to 1.5 ft. (0.5 m) thick. Coralline algae were generally low island-wide, recording 5.09% cover (range 0–40%). The highest coralline algae cover was generally noted along the northern forereef, reaching up to 50.1– 62.5% cover. In 2008, 9 benthic sites were surveyed and showed that Laurencia galtsoffii is the most prevalent species encountered, with cover varying from 0% to 35.2% of the substrate across all survey sites in the lagoon, backreef and forereef (Table 3.19). Dictyota ceylanica, Lobophora variegata, Stypopodium flabelliforme and a species of Padina are present on the atoll and can often be among the most dominant algal species. Overall, turf algae is abundant and was the dominant algal functional group at 6 of the 9 sites surveyed in 2009, with a % cover range of 15.6% to 84.8% (NOAA CREP 2008). The lagoon habitat was characterized by high % cover of both turf and macroalgal communities, and very low presence and % cover (3.6 to 1.6%) of crustose coralline algae (CCA). Microdictyon setchellianum has been the most prevalent species documented at some lagoon sites, with 25.2% to 31.6% algal cover. The backreef habitat was characterized by high % cover of turf algae (34-40%) and a low presence of macroalgae (9%) at most sites, and little CCA cover. The forereef habitat was dominated by turf algae (50-84%) and also hade high cover of macro algae (36-44%). CCA cover on the forereef was less than 3%.

3-106 FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

Table 3.19 Additional species recorded at each site at MANWR during roving diver survey Site Chlorophyta MID-R20 Boodlea composita MID-H11 Codium edule MID-R7, MID-H10, MID-H11, MID-H21 Dictyosphaeria versluysii MID-H11 Halimeda discoidea MID-01, MID-02, MID-R3, MID-H21 Halimeda velasquezii MID-H21 Microdictyon setchellianum MID-01, MID-02, MID-H21 Neomeris sp. Ochrophyta MID-R20 Dictyota ceylanica MID-02 Distromium flabellatum MID-R20 Sargassum sp. MID-H21 Stypopodium flabelliforme MID-H21 Turbinaria ornate Rhodophyta MID-H11 Chondrophycus parvipapillatus MID-H10, MID-H11 Galaxaura filamentosa MID-02 Galaxaura sp. MID-01 Halichrysis coalescens MID-H11 Halymenia sp. MID-02, MID-H10, MID-H11 Jania sp. MID-02 Laurencia galtsoffii MID-H11 Liagora sp. MID-H10 Martensia sp. MID-01 Peyssonnelia sp. MID-R20 Portieria hornemannii

3.3.6.18.2 Probable Impacts: No Action Alternative Under the No Action Alternative, Sand Island’s mouse population would not be the subject of an island-wide eradication project, but limited mouse control efforts would continue using the less toxic AGRID3 to protect nesting seabirds and for human health and safety. No bait would be used in the vicinity of the marine environment; therefore, there would be no effects to marine algae or the marine ecosystem under the no-action alternative.

3.3.6.18.3 Probable Impacts: Preferred Alternative Primary or Secondary Exposure There are no known effects (or pathways) of anticoagulant rodenticides on marine algae and operations would not extend into the marine environment. Therefore, the Preferred Alternative would be unlikely to directly or indirectly impact marine algae. We expect there to be no additional impacts to algae from the implementation of this projects monitoring plan. Implementing the monitoring plan would include follow-up observations and sample collecting. Additional impacts to algae are not expected from implementation of the bio- security plan or monitoring plan since these activities will mostly occur in the terrestrial environment (see Appendix A and B).

3-107 MIDWAY SEABIRD PROTECTION PROJECT FINAL ENVIRONMENTAL ASSESSMENT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

Mitigation The same minimization measures outlined under marine fish would also minimize the amount of rodenticide entering the marine environment. In addition, no short- or long-term discharge of brodifacoum-impacted groundwater into the marine environment is anticipated (see Section 3.3.2).

3.3.6.19 Sea Turtles

3.3.6.19.1 Existing Conditions The ranges of 5 species of marine turtles (all listed under the ESA) encompass the waters of Midway, including the loggerhead turtle (Caretta caretta), the olive ridley (Lepidochelys olivacea), and the leatherback (Dermochelys coriacea). The loggerhead sea turtle was first listed as threatened under the ESA in 1978 (NMFS and USFWS 1998c). In 2011, the loggerhead was grouped into nine DPSs, of which the population surrounding Midway is within the endangered North Pacific Ocean DPS (NMFS 2011a). Turtles within this DPS forage in the central and eastern Pacific, returning to natal beaches in Japan to reproduce, and then remain in the western Pacific. There is no loggerhead nesting on the western seaboard of the United States or in Hawai‘i, nor is there any documentation to suggest that nesting occurs in any of the U.S. unincorporated island territories of the Pacific (Balazs 1982). Similarly, there are no nesting records from , Commonwealth of the Northern Mariana Islands, Palau, Republic of the Marshal Islands (Thomas 1989), Federated States of Micronesia (Pritchard 1982), or American (Tuato’o-Bartley et al. 1993). There are very few records of loggerheads nesting on any of the many islands of the Central Pacific; the species is considered rare or vagrant in this region (NMFS and USFWS 1998c). This loggerhead DPS is not known to cross the equator nor to mix with individuals from the Southern Pacific DPS and are genetically isolated from populations outside of the Pacific by an estimated one million years. The North Pacific Ocean DPS is threatened by elimination and degradation of nesting habitat, sea level rise, and incidental bycatch in fishing gear. Critical habitat has been designated only for the Northwest Atlantic DPS of loggerheads. Adult olive ridleys are federally listed as threatened and are easily distinguished from other sea turtle species. The olive ridley is the smallest living sea turtle with an adult carapace length usually between 24-28 in. (60-70 cm) and a weight rarely exceeding 110 lbs. (50 kg) (Schulz 1975). Their name comes from their round-shaped, pale olive carapace, and yellow-olive plastron. Globally, their population numbers are high but heavily exploited (NMFS and USFWS 1998d). The species is threatened by the impacts of poaching, incidental bycatch, and habitat degradation and development (NMFS 2014). No critical habitat has been designated for the olive ridley (NMFS and USFWS 1998d). This species commonly nests along continental margins using arribidas (mass, synchronized nesting). It is possible that young turtles move offshore and occupy areas of surface current convergences to find food and shelter among aggregated floating objects until they are large enough to recruit to the nearshore benthic feeding grounds of the adults (NMFS 2014). It is unknown whether the olive ridley has a truly pelagic habit; further research is needed in this regard. At sea in the eastern tropical Pacific, olive ridleys readily associate with objects floating in the

3-108 FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

water including anything from logs to plastic debris to dead whales (Pitman 1992; Arenas and Hall 1992) and appear strongly attracted to brightly colored objects (Arenas and Hall 1992). Olive ridleys often bask at the surface in the eastern Pacific where they are frequently accompanied by seabirds. The birds, mainly boobies, roost on the exposed carapaces of the turtles and feed on fish that aggregate beneath them (NMFS and USFWS 1998d, Pitman 1993). While rare in Hawaii, olive ridleys have occasionally been killed by commercial fishing vessels (NMFS and USFWS 1998d). The entanglement of juveniles and adults in marine debris around the Hawaiian Islands is reported from Kailua-Kona (Hawaii), Pukoo (Molokai), Hana (Maui), and Oahu (Balazs 1985). A single nesting event was recorded on Maui in 1985 (NMFS and USFWS 1998d). Genetic analysis of olive ridley sea turtles taken in the Hawaii-based longline fishery shows that about two-thirds of the animals came from the eastern Pacific, while the remaining one- third originated in the western Pacific or Indian Ocean, suggesting that Hawaii represents a point of convergence for these source areas (Mitchell et al. 2005). The leatherback is the largest of the sea turtle species, federally listed as endangered, and the only one that lacks a hard-shelled carapace and scutes. Instead their carapace is a tough rubbery black skin about 1.5 in. (3.8 cm) thick and consists of leathery, oil-saturated connective tissue overlaying loosely interlocking dermal bones (NMFS and USFWS 1998c). Their carapace has seven ridges along its length and tapers to a blunt point, which helps the leatherback move more effectively in water. Their front flippers are proportionally longer than in other sea turtles and their back flippers are paddle-shaped. A ridged carapace and large flippers improve energy efficiency and enable long distance foraging migrations (NOAA Fisheries 2018a). Leatherback turtles in the Pacific Ocean are experiencing a dramatic drop in nesting numbers, and recent reports estimate the number of breeding females at between 2,700 and 4,500, though this number is uncertain due to a lack of information on the typical number of nests per female (NMFS 2013). Leatherback turtles have a variety of adaptations that allow them the widest foraging range of any living reptile, however nesting is confined to tropical and subtropical latitudes. Western Pacific leatherback populations nest in , , Papua , and the Solomon Islands; they are not known to breed with Eastern Pacific leatherbacks and are genetically distinct though their ranges overlap. Significant threats to the leatherback result from poaching, development, marine debris, beach erosion, and low hatch rates (NMFS 2013). Western Pacific leatherback populations are estimated to have dropped by 80% in recent decades. Critical habitat for the leatherback turtle was established in the U.S. Virgin Islands in 1979 and along much of the west coast of the United States in 2012, although it does not extend to the Hawaiian Islands (NMFS 2012). The United States does not have any nesting of leatherbacks in its jurisdiction in the Pacific but has important foraging areas on the continental U.S. west coast and near the Hawaiian Islands (NMFS and USFWS 1998c). Leatherbacks encountered in Hawaii, including those caught incidental to fishing operations, may represent individuals in transit from one part of the Pacific to another (Balazs 1973). In August 1979 at least ten individuals, including juveniles, were sighted in pelagic waters northwest of Hawaii (40-42° N, 175-179° W) (Balazs 1982). A large (682 kg) female became entangled at night two miles offshore Kailua-Kona (Balazs 1985). One leatherback sea turtle washed up dead at Midway Atoll in the early 1990s (D. Williams, USFWS, pers. comm.). As only one leatherback has ever been observed at Midway Atoll, the species is most likely very

3-109 MIDWAY SEABIRD PROTECTION PROJECT FINAL ENVIRONMENTAL ASSESSMENT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

uncommon within the Refuge, probably only occasionally migrating through deep, pelagic waters. Satellite tracking studies show that leatherback turtles tagged on the California coast migrated in offshore waters at the southeastern end of the Hawaiian archipelago on their way to an area just north of Australasia (TOPP 2006). Worldwide populations of the green sea turtle (Chelonia mydas) have seriously declined as a direct result of overharvesting of turtles and eggs over the last centuries (Parsons 1962). In 1978, the Hawaiian subpopulation of the green turtle was listed as threatened under the Endangered Species Act of 1973. In 1978, the green sea turtle in Hawai‘i was ESA-listed as threatened, and in 2016, they were removed from the range-wide and breeding population listing of ‘green sea turtle’ and instead were listed as one of eight distinct population segments (DPSs) as threatened (USFWS 2016c). In the NWHI, it is the Central North Pacific DPS of green sea turtles that are present. Additional protective regulations are enforced throughout all areas within U.S. jurisdiction to conserve and restore marine turtle populations to their former levels of abundance. Inclusion of green turtles into the Convention on International Trade in Endangered Species of Wild Fauna and Flora (CITES), also known as the Washington Convention, made it illegal to trade any products made from this species in the U.S. and 130 other countries. Threatened Hawaiian green turtles are frequently seen inside the MANWR lagoon and basking on beaches. They are present year-round. No turtle nesting had been documented until successfully hatched eggs were discovered on Spit Islet in July 2006. High surf uncovered the eggs, which probably hatched in 2005. In 2007, a successful sea turtle nest was documented on Sand Island and in June 2017 several crawls and possible nests were observed but not confirmed on the beach to the west of Bulky Dump and North Beach. Adult green sea turtles are , feeding on seaweeds, seagrasses, and algae (NMFS and USFWS 1998a). Juveniles are omnivores, eating a range of insects, crustaceans, worms, and seagrasses. Sea turtles are also reported to feed on marine debris (Schuyler et al. 2014). Green turtles of nearshore habitats in the Hawaiian Islands feed on benthic algae of the following genera: Codium, Amansia, Pterocladia, Ulva, Gelidium, and Caulerpa (NMFS 1998). Green turtles often nest on wide sandy beaches, and nesting has been documented at least once at Midway. Endangered hawksbill sea turtles (Eretmochelys imbricata) are infrequently seen in the lagoon, and other species of turtles have been observed foraging at MANWR. Hawksbill turtles are specialist sponge carnivores, selecting just a few genera of sponges for their principal diet (Vicente 1994). This feeding strategy is unique, as few vertebrates are capable of digesting sponges without being injured by the sponges' silicate spicules (i.e., needles).

3.3.6.19.2 Probable Impacts: No Action Alternative Under the No Action Alternative, Sand Island’s mouse population would not be the subject of an island-wide eradication project, but limited mouse control efforts would continue using the less toxic AGRID3 to protect nesting seabirds and for human health and safety. Therefore, there would be no effects to sea turtles under the no-action alternative.

3-110 FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

3.3.6.19.3 Probable Impacts: Preferred Alternative Primary and Secondary Exposure Little is known about the effect that brodifacoum may have on turtles. Rodenticide toxicity experiments have not been conducted in many turtle species and therefore the LD50 values are unknown for the green sea turtles or other species of turtles present in the waters surrounding MANWR. However, an initial assessment from preliminary findings of a USDA National Wildlife Research Center (NWRC) turtle-anticoagulant hazards study indicates ornate wood turtles (Rhinoclemmys pulcherrima) were not negatively affected by brodifacoum consumption. Turtles that were fed high brodifacoum doses received 2.5 x 10-5 oz./lb. (1.6 mg/kg) of turtle body weight of brodifacoum, and none died or showed signs of ill health prior to being euthanized one week later. The turtle with the highest liver residue level (2.02 ppm) weighed 0.7 lbs. (319 g), which means that it received about 500 ppm (0.5 mg) of brodifacoum. Since a Brodifacoum-25D pellet contains 25 ppm, the turtle essentially received the equivalent of 20 pellets (USFWS 2011). Adult green sea turtles weigh on average 325 lbs. (147 kg) (NOAA Fisheries Service 2011a), thus, using similar metrics, one adult green turtle would have to consume approximately 9,200 pellets or 40.5 lbs. (18.4 kg) of pellets to receive a comparable exposure to that the ornate wood turtle received which showed no ill effects. Adult hawksbill turtles weigh on average 125 lbs. (57 kg) (NOAA Fisheries Service 2011b), thus one hawksbill turtle would have to consume approximately 3,500 pellets or 15.4 lbs. (7.0 kg) of pellets to receive a comparable exposure to that the ornate wood turtle received. Given that sea turtles feed in water, and that they would need to ingest a significant amount of brodifacoum to experience toxicosis, the risk of negative outcomes from baiting over areas where sea turtles are resting is low. Green sea turtles could potentially eat baits that drift into the water. However, it would be very unlikely turtles would consume this bait given the rapid decomposition of bait in water. Other sea turtles are absent or too rare around MANWR to be a concern. Therefore, there would likely be no effect on sea turtles as a result of implementation of the Preferred Alternative. Secondary exposure or other indirect effects to sea turtles also would be very unlikely due to their diet. There is a low probability of bait entering the water, and bait pellets decompose rapidly in water and are not taken up by marine plants. Because a sea turtle’s diet is primarily plant-based and given that turtles would need to ingest a significant amount of brodifacoum to experience toxicosis, secondary exposure is unlikely, and therefore, any impacts are unlikely. Operational Hazard (Noise/Human Disturbance/Ship Strike) Green sea turtles may be disturbed when loafing on the beach, but this occurs predominantly later in the day and thus disturbance should be minimal during the early morning and afternoon when the operation would take place. With a helicopter speed ranging from 29–58 mph (46-93 km/hr) during each bait drop, even at the slowest airspeed the helicopter would take only 16 seconds to travel 656 ft. (200 m). Thus, disturbance would only last a few seconds. The level of disturbance for individuals is likely to be low since the disturbance would only occur at each site once or twice during each baiting application. Therefore, these noise impacts would be short-lived and negligible. Sea turtles are more likely to experience significant disturbance from hand-broadcasting activities. It will not be possible to estimate the time a turtle will spend in one area, so waiting until the

3-111 MIDWAY SEABIRD PROTECTION PROJECT FINAL ENVIRONMENTAL ASSESSMENT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

turtle(s) have moved on before broadcasting bait would cause significant delays in obtaining full bait coverage. Given the low risk of exposure and higher risk of disturbance of hand-broadcasting, we currently plan to bait via helicopter over areas with turtles. We will also have specialists on site that can monitor monk seals and turtles throughout the bait drop for disturbance and exposure risk and provide recommendations to the project manager based on those observations. We will also follow all conservation recommendations from our Section 7 consultation with the NMFS which is currently in progress. We expect there could be some additional impacts to green sea turtles from the implementation of the projects monitoring plan (Appendix B). Implementing the monitoring plan would include follow-up observations and sample collecting. Potential impacts such as disturbing beach-basking turtles could occur from the additional human disturbances. Additional impacts to sea turtles are not expected from implementation of the bio-security plan since this will occur in the terrestrial environment. Given the minimization measures that will be in place to protect the Hawaiian green sea turtle, we expect any additional impacts from disturbance to be minor, localized, and short- term. Ships navigating from Honolulu and through Refuge waters and entering MANWR’s lagoon have the potential to accidentally strike sea turtles in the water and potentially causing injury or mortality. Only green sea turtles are observed around MANWR. Hawksbill sea turtles are rarely observed in Refuge waters and within MANWR’s lagoon and loggerhead, olive ridley, and leatherback turtles are also rare around MANWR and in the NWHI. Minimization measures that will be in place to protect sea turtles from ship strikes are described below. Mitigation The same minimization measures outlined under marine fish would also minimize the amount of rodenticide entering the marine environment. In addition, all project personnel on the ground would maintain a 100 ft. (30.5 m) buffer from sea turtles during operations. During aerial bait broadcast, helicopters would avoid hovering near turtle basking areas and would avoid distributing pellets over turtles on the beaches to the extent possible without sacrificing the chance of project success. Ship strike avoidance measures would be in place to avoid impacting sea turtles during the transport of materials and equipment to and from Honolulu. These include reducing the vessel’s speed to 5-10 knots in the presence of, and in known areas of, listed marine species, maintaining a 656 ft. (200 m) buffer around marine species encountered at sea, implementing shutdown protocol, if necessary, until species leave(s) the area, and keeping a watchful eye for sea turtles when navigating Refuge waters. These mitigation measures would make the potential for ship strikes low. There have been no known record of ship strikes with a marine protected species by any authorized vessel operator in the Monument. Therefore, we expect the likelihood of a marine protected species being the victim of a ship strike to be very unlikely. The USFWS will consult with the Protected Species Division of NOAA, under Section 7 of the Endangered Species Act, to identify the best course of action to ensure protection of sea turtles. This process will address monitoring protocols, bait application around turtles, and a response plan in the unlikely event of finding an individual sick or dead during or following the operation. We will monitor sea turtles, and other non-target organisms for unintentional hazardous exposure and will follow best management practices as recommended by the Section 7 consultation. We will also have specialists on site monitoring this species. These ground observers can quickly note any

3-112 FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

adverse impacts and relay them through the chain of command to the project manager. The project manager will maintain communication with the helicopter team and can course correct as needed. These mitigation measures would reduce any potential impacts on sea turtles.

3.3.6.20 Marine Mammals, Rays, and Sharks

3.3.6.20.1 Existing Conditions: Dolphins and Hawaiian Monk Seals:

Marine mammals of 2 orders, Cetacea and Pinnipedia, have been observed in Figure 3.16 Hawaiian monk seal the pelagic waters surrounding Midway. The most commonly sighted cetaceans are bottlenose dolphins (Tursiops truncatus) and spinner dolphins (Stenella longirostris). Although protected under the Marine Mammal Protection Act (MMPA) of 1972, none of the cetaceans mentioned here are listed under the ESA or otherwise considered threatened. Approximately 200-300 Hawaiian spinner dolphins rest within Midway’s lagoon and forage outside the atoll. Bottlenose, striped (Stenella coeruleoalba), spotted (Stenella Source: USFWS (2018) attenuate), and rough-toothed dolphins (Steno bredanensis) may occasionally be seen in the open ocean, as well as beaked (family; Ziphiidae), pilot (Globicephala macrorhynchus), and the endangered humpback whales (Megaptera novaeangliae). Spinner dolphins frequently use inshore island and atoll habitat for day-time rest and social interactions, and forage over deep waters at night. The groups of spinner dolphins associated with Kure and Midway Atolls have been the subject of long-term research since 1998. Depth contours are used as vectors of movement within the atoll’s lagoon, while areas of extreme local depth and shallows are generally avoided. Socially important behaviors, such as resting and socializing, occur over spatially restricted areas, contrary to behaviors such as travel and play that might occur over a considerably larger portion of the lagoon. Potential measurable impacts from mice eradication activities to cetaceans in the waters surrounding MANWR would be negligible, since all of the activities described in the proposed action are terrestrial and these marine mammals forage outside of the lagoon. However, there is a small risk injury or mortality of listed marine species from collision from a ship or barge transporting materials to and from MANWR from Honolulu for the eradication project. Therefore, part of the project area includes the waters between Honolulu, HI and the lagoon at MANWR. To account for the possibility of a ship strike, we have included these listed species in this analysis. These species include the Main Hawaiian Islands (MHI) false killer whale (Pseudorca crassidens), sperm whale (Physeter macrocephalus), fin whale (Balaenoptera physalus), blue whale

3-113 MIDWAY SEABIRD PROTECTION PROJECT FINAL ENVIRONMENTAL ASSESSMENT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

(Balaenoptera musculus), sei whales (Balaenoptera borealis), and North Pacific right whale (Eubalaena japonica), the giant manta ray (Manta birostris) and the oceanic whitetip shark (Carcharhinus longimanus). The Hawaiian monk seal (Neomonachus schauinslandi) is one of the most endangered marine mammals in the world and the rarest pinniped in US waters (see Figure 3.16). Following decades of decline, Hawaiian monk seal abundance increased 3% annually from 2013-2016 and is estimated at about 1,400 individuals. This positive trend is largely due to multiple years of increased juvenile survival in the Northwestern Hawaiian Islands (NWHI). Weighing between 400-600 lbs. (180-270 kg) and about 7-7.5 ft. (2.1-2.3 m) in length, females are slightly larger than males. Pups are approximately 24-33 pounds (11-15 kg) at birth and about 3 ft. (1 m) long. The Hawaiian monk seal’s entire range is within U.S. waters and the majority of seals reside in the NWHI within the Papahānaumokuākea Marine National Monument at 8 breeding subpopulations. These islands include; (i) Kure Atoll; (ii) Midway Islands; (iii) Pearl and Hermes Reef; (iv) Lisianski Island; (v) Laysan Island and; (vi) French Frigate Shoals, (vii) Mokumanamana (Necker) Island and (viii) Nihoa Island. A sub-population of about 300 seals is now also found on the main Hawaiian Islands where births have occurred on all major islands. Hawaiian monk seals live in warm subtropical waters and spend two-thirds of their time at sea. They use waters surrounding atolls, islands, and areas farther offshore on reefs and submerged banks. Hawaiian monk seals have also been found using deepwater coral beds as foraging habitat. Hawaiian monk seals sometimes spend days at sea before returning to the islands where they sleep and digest their food. Hawaiian monk seals are primarily benthic foragers, feeding on a variety of prey including fish, cephalopods, and crustaceans. Hawaiian monk seals generally hunt for food outside of the immediate shoreline areas in waters 60-300 ft. (18-90 m) deep and at depths of up to 1,600 ft. (500 m) where they prey on eels and other benthic organisms (NMFS- NOAA 2007). Hawaiian monk seals come ashore to rest, give birth and rear their pups, and during the molting period. They haul-out on sand, corals, volcanic rock, and other substrates. Sandy, protected beaches surrounded by shallow waters are preferred when pupping. Hawaiian monk seals are often seen resting on beaches on Sand Island (see Figure 3.17). Designated critical habitat for Hawaiian monk seals was last revised in 2015 (80 FR 50925). Critical habitat for monk seals is defined by three essential features: “preferred pupping and nursing areas, marine foraging areas, and/or significant haul-out areas that will support conservation for the species” (NOAA-NMFS 2015). Suitable areas on Sand, Eastern and Spit islands are included in this designation. However, “areas that are inaccessible to seals and/or have manmade structures that lack the essential features are not included in the designation for Hawaiian monk seal critical habitat” in the NWHI (NOAA-NMFS 2015). Females generally mature at age 7-10, the gestation period is believed to be about 10-11 months, and most births occur between February and July, with a peak in April to May. However, birthing has been recorded year-round. Nursing occurs for about 39 days, during which time the mother fasts and remains on land. After this period, the mother abandons her pup and returns to sea. The information below summarizes Hawaiian monk seal beach counts conducted on Sand Island, Midway from 2012 through 2016 (Johanos 2017). Data were collected on each hauled out seal during systematic whole-island beach surveys which began around 1300. In total, observers

3-114 FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES conducted 41 standardized beach surveys and recorded 486 Hawaiian monk seal sightings over this 5-year period. For the purpose of seal data collection, the perimeter of Sand Island is divided into 29 areas, or sectors, to describe the spatial properties of seal habitat use (Figure 3.17) (Table 3.20 and Table 3.21).

Figure 3.17 Sand Island Hawaiian Monk Seal Sightings: 2012-2016

Source: NOAA (2017)

Beach count data are summarized below by sector location (Table 3.20) and month (Table 3.21). In 2016, the total number of endangered Hawaiian monk seals present at Midway was estimated to be 74 seals (12 pups and 62 non-pups). Seals within the Midway population move freely between all islets. Therefore, these animals could be present on or around both Sand and Eastern Island during the bait drop. Of course, seals are always coming and going from MANWR, and only a portion of the subpopulation is expected to be hauled out on beaches at any one time. Pupping levels have increased significantly since 1996, with a record number of 17 in 2004. These seals are present at all times of the year, generally hauling ashore to rest. The pupping season on Midway is predominantly February to July with only 3-4 animals typically having pups on Sand Island. However, as is common throughout the Hawaiian Islands, survivorship of juveniles is low and contributes to the endangered status of the species. In an effort to increase survivorship, NOAA-Fisheries established a captive care program on Sand Island in 2006. Six females were released in March 2007.

3-115 MIDWAY SEABIRD PROTECTION PROJECT FINAL ENVIRONMENTAL ASSESSMENT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

Table 3.20 Hawaiian Monk Seal Beach Count Data Summarized by Sector Average number of seals hauled out on Sand Island, MANWR, by sector, per standardized beach count, conducted during 2012 through 2016. This is the number of seals you would expect to encounter, on average, on one beach count. Sector Non-Pup Size Classes Totals Adult Sub-Adult Juvenile Non-Pup Pup All Seals 1 0.3 0.1 0.1 0.5 0.0 0.5 2 0.4 0.0 0.0 0.5 0.0 0.5 3 0.2 0.2 0.1 0.4 0.0 0.4 4 0.2 0.0 0.0 0.2 0.2 0.4 5 0.1 0.0 0.0 0.4 0.2 0.6 6 0.2 0.1 0.0 0.3 0.0 0.3 7 0.5 0.2 0.0 0.7 0.0 0.7 8 0.7 0.3 0.0 1.0 0.0 1.0 9 0.7 0.3 0.2 1.1 0.0 1.1 10 0.9 0.2 0.1 1.2 0.1 1.3 11 0.5 0.7 0.3 1.5 0.1 1.7 12 0.0 0.2 0.1 0.3 0.0 0.3 13 0.0 0.0 0.0 0.1 0.0 0.1 14 0.0 0.1 0.0 0.2 0.0 0.2 15 0.1 0.0 0.0 0.1 0.0 0.1 16 0.0 0.0 0.0 0.0 0.1 0.1 17 0.0 0.0 0.0 0.0 0.0 0.1 18 0.1 0.0 0.0 0.1 0.0 0.1 19 0.6 0.1 0.0 0.7 0.0 0.7 20 0.0 0.0 0.0 0.0 0.0 0.0 21 0.0 0.0 0.0 0.0 0.0 0.0 22 0.0 0.0 0.0 0.0 0.0 0.0 23 0.0 0.0 0.0 0.0 0.0 0.0 24 0.1 0.0 0.0 0.2 0.0 0.2 25 0.0 0.0 0.0 0.0 0.0 0.0 26 0.0 0.0 0.0 0.0 0.0 0.0 27 0.0 0.0 0.0 0.0 0.0 0.0 28 0.1 0.1 0.0 0.1 0.0 0.1 29 0.6 0.1 0.0 0.6 0.0 0.7 Source: PIFSC 17-021 (Johanos2017)

3-116 FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

Table 3.21 Hawaiian Monk Seal Beach Count Data Summarized by Month Average number of seals hauled out on Sand Island, MANWR, by month, during standardized beach counts conducted during 2012 through 2016. Month No. of Surveys Non-Pup Size Classes Totals Adult Sub-Adult Juvenile Non-Pup Pup All Seals Std.

March 1 5.0 3.0 6.0 14.0 0.0 14.0 - April 2 5.0 2.5 3.0 10.5 0.5 11.0 1.41 May 2 5.5 2.0 3.5 11.0 1.0 12.0 4.24 June 11 7.1 4.4 1.1 12.6 2.6 15.1 5.82 July 9 6.6 3.6 0.1 10.2 0.8 11.0 3.12 August 6 6.2 1.7 0.8 8.7 0.5 9.2 2.64 September 7 7.9 1.9 0.9 10.6 0.6 11.1 1.68 October 2 6.0 2.0 1.5 9.5 0.0 9.5 3.54 November 1 5.0 1.0 3.0 9.0 0.0 9.0 - Source: PIFSC17-021 (Johanos 2017)

3.3.6.20.2 Probable Impacts: No Action Alternative Under the No Action Alternative, Sand Island’s mouse population would not be the subject of an island-wide eradication project, but limited mouse control efforts would continue using the less toxic AGRID3 to protect nesting seabirds and for human health and safety. No bait would be used in the vicinity of the marine environment; therefore, there would be no effects to marine mammals under the no-action alternative.

3.3.6.20.3 Probable Impacts: Preferred Alternative Primary Exposure As many as 74 Hawaiian monk seals could be present on or around Sand Island during the proposed operation. Because they do not forage on land, direct consumption of bait pellets from the ground would not be a risk. Bait pellets that might drift into the water would fall close to the shoreline far from the typical offshore foraging areas of monk seals, however weaned pups often mouth and may consume sea cucumbers and other prey items in the wave wash and nearshore environment (NMFS unpublished data). Older Hawaiian monk seals were found not to interact with bait pellets during a placebo trial on the island of Lehua in 2015 for a rat eradication study (Mazurek 2015). Given that monk seals feed in water, and that they would need to ingest a significant amount of brodifacoum to experience toxicosis, the risk of negative outcomes from baiting over areas where monk seals are resting is low. Moreover, bait pellets degrade quickly in water and fragments sink to the bottom (Howald et al. 2010, Siers et al. 2018, Pitt et al. 2015). In addition, the bait bucket will be equipped with a deflector to eliminate or reduce the amount of drift into the water. Given that bait pellet longevity is expected to be short-lived and that seals are unlikely to interact with bait pellets, impacts from primary exposure are unlikely. Secondary Exposure Hawaiian monk seals would only be at risk of secondary exposure to brodifacoum in the unlikely event that a very large quantity of bait was accidently dropped into the ocean, and fish or other

3-117 MIDWAY SEABIRD PROTECTION PROJECT FINAL ENVIRONMENTAL ASSESSMENT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

prey items were able to consume it before ocean currents dissipated the spilled bait. Baitpellets degrade quickly in water and fragments sink to the bottom, and so would only be available to monk seals for a very short period of time. However, fish have been demonstrated to be intoxicated with anticoagulant rodenticide that has entered the ocean directly or indirectly from spills and eradication projects (Primus et al. 2005, Pitt et al. 2015). In a 2015 placebo trial on the island of Lehua, 19 species of fish in the near shore environment consumed the baits (Mazurek 2015). However, post-application sampling of the near shore marine environment from 2 eradication projects in Hawaiʻi found no detectable levels of rodenticide in fish, invertebrates, or seawater (Orazio et al. 2009, Gale et al. 2008). Furthermore, because older Hawaiian monk seals typically forage in offshore areas, it would be unlikely they would prey on any fish or invertebrates that may have consumed rodenticide pellets. In the unlikely event a non-pup seal did forage in the near shore environment there would be a low probability that it would encounter a prey item that had consumed rodenticide. Pupping in MANWR is predominantly February to July, peaking in April to May, and nursing lasts 39 days; only 3-4 seals typically have pups on Sand Island (USFWS, unpublished data). Additionally, the monk seal pupping areas on Sand Island are known and can be surveyed prior to aerial baiting operations for the presence of pups. Using this information will allow staff to implement a more focused and effective mitigation strategy. Thus, impacts from secondary exposure to nursing and newly weaned pups would be unlikely. In addition, a 220 lbs. (100 kg) juvenile seal would have to eat 22-88 lbs. (10-40 kg) of intoxicated fish to receive the calculated lethal doses. Hawaiian monk seals normally consume 5.8 to 12.9% of their body weight per day. For a 220 lbs. (100 kg) juvenile, that would correspond to around 13-28 lbs./day (6 to 13 kg/day). It is not likely that a seal would consume more than their daily food intake of contaminated fish. Because nursing and newly weaned pups remain near shore for extended periods, there is some risk of exposure for this age class. Based on the above data, it would be unlikely that the Preferred Alternative would primarily or secondarily expose non-pup Hawaiian monk seals to a sufficient quantity of rodenticide to have any negative effects. There have been no documented cases of impacts to seals or sea lions after aerial bait application, including the 2009 rodenticide application on Lehua Island. Operational Hazard (Noise/Human Disturbance/Ship Strike) Air and ground-based operations (such as hand-broadcasting bait) have the potential to cause Hawaiian monk seals to move or flush into the ocean. However, with a helicopter speed ranging from 29–58 mph (46-93 km/hr.) during each bait drop, even at the slowest airspeed the helicopter would take only 16 seconds to travel 656 ft. (200 m). Thus, disturbance would only last a few seconds. Therefore, these noise impacts would be short-lived and negligible, and have been determined as not likely to adversely affect. The level of disturbance for individuals is likely to be low since the disturbance would only occur at each site once or twice during each baiting application, which is believed not to pose a significant risk. The exception is during the pupping season. Although pups can be born at any month of the year, most pups are born between February and July, which is predominantly earlier than the proposed bait drop. There is little that can be done to mitigate this disturbance, but as only 3-4 seals pup on Sand Island each year, and most pupping would be completed before the baiting period, the disturbance would be short lived and minor.

3-118 FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

Monk seals may experience impacts from air and ground-based operations. Injury may occur if seals are disturbed and are caused to move or flush into the ocean. Aerial distribution of bait has been determined to be the least disturbing method to seals and ground crews will maintain a 100 ft. (30.5 m) distance from resting seals. Since we will not be able to estimate any given time a monk seal will spend in one area, waiting until animals have moved on before broadcasting bait would cause significant delays in obtaining full bait coverage. We currently plan to bait via helicopter over areas with monk seals and turtles. Helicopter broadcast of inert bait over monk seals was conducted at Lehua Island, and observations showed that seals were not disturbed by helicopters. We will also have specialists on site that can monitor seals throughout the bait drop for disturbance and exposure risk and provide recommendations to the project manager based on those observations. There may be some additional disturbances to monk seals near the boat dock on Eastern Island due to additional human activity from the construction of Laysan duck aviaries and additional boat traffic ferrying personnel and equipment from Sand Island. However, these animals are used to some level of disturbance since personnel visit Eastern Island frequently to conduct bird surveys, vegetation restoration activities, and to maintain guzzlers for Laysan ducks. We will follow all conservation recommendations from our Section 7 consultation with the NMFS which is currently in progress. We expect there will be no additional impacts to bottlenose and spinner dolphins but some possible additional impacts to Hawaiian monk seals should the need arise to implement the rapid response portion of the refuge’s biosecurity plan and during implementation of this projects monitoring plan (see Appendix A and B). Biosecurity measures enacted would include rapid response to an incursion by non-native invasive terrestrial species and implementing the monitoring plan would include follow-up observations and sample collecting. Additional impacts to dolphins are not expected from implementation of the bio-security plan or from the additional human disturbance, since these will almost all occur in the terrestrial environment. However, some additional impacts such as disturbing basking seals and pupping activities could occur from these activities. Given the minimization measures that will be in place to protect the Hawaiian monk seal, we expect any additional impacts to be minor, localized, and short-term. There is a small risk injury or mortality of Hawaiian monk seals from collision from a ship or barge transporting materials to and from MANWR from Honolulu for the eradication project. Given the minimization measures that will be in place to protect the Hawaiian monk seal from ship strikes (see below), and the fact that there have been no known record of ship strikes with a marine protected species by any authorized vessel operator in the Monument, we expect the likelihood of a marine protected species being the victim of a ship strike to be very unlikely. Mitigation The same minimization measures outlined under marine fish would also minimize the amount of rodenticide entering the marine environment. In addition, all project personnel on the ground would maintain a 100 ft. (30.5 m) buffer from seals during operations. During aerial bait broadcast, helicopters would avoid hovering near Hawaiian monk seals and would avoid distributing pellets over seals on the shore to the extent possible without sacrificing the chance of project success. These mitigation measures would reduce impacts on Hawaiian monk seals. Indirect effects from the Preferred Alternative, particularly for non-pups, would be unlikely.

3-119 MIDWAY SEABIRD PROTECTION PROJECT FINAL ENVIRONMENTAL ASSESSMENT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

Ship strike avoidance measures would be in place to avoid impacting Hawaiian monk seals during the transport of materials and equipment to and from Honolulu. These include reducing the vessel’s speed to 5-10 knots in the presence of, and in known areas of, listed marine species, maintaining a 656 ft. (200 m) buffer around marine species encountered at sea, implementing shutdown protocol, if necessary, until species leave(s) the area, and keeping a watchful eye for sea Hawaiian monk seals when navigating Refuge waters. These mitigation measures would make the potential for ship strikes low. The USFWS will consult with the Protected Species Division of NOAA, under Section 7 of the Endangered Species Act, to identify the best course of action to ensure protection of monk seals. This process will address monitoring protocols, bait application around monk seals, and a response plan in the unlikely event of finding an individual sick or dead during or following the operation. We will monitor seals, turtles, and other non-target organisms for unintentional hazardous exposure and will follow best management practices as recommended by the Section 7 consultation. We will also have specialists on site monitoring these species. These ground observers can quickly note any adverse impacts and relay them through the chain of command to the project manager. The project manager will maintain communication with the helicopter team and can course correct as needed. A biologist from the Hawaiian Monk Seal Research Program will also be on MANWR during the seabird protection project to monitor project activities.

3.3.6.20.4 Existing Conditions: False Killer Whale The Main Hawaiian Islands (MHI) insular DPS of false killer whales was listed as endangered under the ESA in 2012 while the NWHI population was determined to be separate and distinct but is not listed under the ESA. The MHI DPS is most threatened by interactions with local fisheries. Genetic differentiation between the two populations and offshore pelagic populations of false killer whales sufficiently shows that the populations do not interbreed and that the MHI population is unlikely to be readily replaced by other populations should it become extinct. The MHI DPS is known to preferentially use habitat on “the northern coast of Moloka‘i and Maui, the north end of Hawaii Island, and a small region southwest of Lāna‘i” (NMFS 2012b).

3.3.6.20.5 Existing Conditions: Blue Whale Estimates of commercial whaling takes of blue whales are at least 9,500 between 1910 and 1965. Since protection by the International Whaling Commission in 1966 and listing under the ESA in 1970 as endangered, insufficient data has been available to estimate population trends. Distribution of blue whale populations is not well understood, however recent studies suggest that blue whales from the central Pacific appear to summer southwest of Kamchatka, the Aleutian Islands, and the Gulf of Alaska and to spend winters in the western and central Pacific, including Hawai‘i. Presence of blue whales within the Hawaiian Islands is registered through very infrequent sightings and through recordings of whale song. Blue whales are generally found further offshore than other whale species. The most recent estimates of individuals within the Hawai‘i exclusive economic zone (EEZ) are between 38 and 81 whales. Threats from fishery-related mortality or serious injury are considered insignificant, but the effects of sonar and other noise sources are still of concern (NMFS-NOAA 2014a).

3-120 FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

3.3.6.20.6 Existing Conditions: Fin Whale The fin whale was listed as endangered under the ESA in 1970. Hawai‘i populations of fin whales are believed unlikely to mix with two other North Pacific populations of fin whales in Alaska and California/Oregon/Washington. The eastern North Pacific population was estimated at 25,000 – 27,000 prior to whaling and 8,000 – 11,000 in 1973 (following ESA listing and the cessation of fin whale hunting in the Pacific in 1972) (NMFS-NOAA 2011b). The 2010 stock estimate reports similar numbers, with only 101 – 174 in the Hawai‘i stock (NMFS-NOAA 2014b). There is no critical habitat designated for the fin whale in the Pacific.

3.3.6.20.7 Existing Conditions: Sei Whale Listed as endangered, in a 2011 stock estimate, the population of sei whales remaining in U.S. Pacific waters was estimated at a minimum of 120 whales. Whaling catch data suggests a decline to about 8,600 individuals in 1974 from 42,000 in 1963 following decades of heavy harvests in Japanese waters that suggest the population was already below carrying capacity at that time. Sei whales were listed as endangered under the ESA in 1970. Relatively little data exists to provide information about species recovery since that time (NMFS 2012c). Sei whales in Hawaiian waters was estimated to number 77 individuals (Smultea et al. 2010). The limited available data on this species suggest that it is rare near the MHI and that a small number use deep offshore waters during the fall (Smultea et al. 2010).

3.3.6.20.8 Existing Conditions: Sperm Whale Sperm whales typically occur in deep pelagic waters and are uncommon in waters less than 984 ft. (300 m). They were listed as endangered under the ESA in 1970 and no change to the listing has been made since that time. Estimated population size of the Hawai‘i stock of sperm whales is between 2,500 and 3,400 (NMFS-NOAA 2015). There is no critical habitat designated for the sperm whale. Only one sighting has been recorded at Midway Atoll; a sperm whale washed up dead on the surrounding coral reef in the late 1990s (N. Hoffman, USFWS, pers. comm.). The skeleton is currently on display outside the USFWS National Wildlife R visitor center at Midway Atoll.

3.3.6.20.9 Existing Conditions: North Pacific Right Whales Right whales have large heads, no dorsal fins, and are mostly black in color with callosities around the head region and a broad, deeply notched tail (NMFS 2006). The North Pacific and North Atlantic populations of right whales were first considered as one species of ‘northern right whale’ before they were reclassified into two separate species in 2008 into the North Pacific right whale (Eubalaena japonica) and North Atlantic right whale (E. glacialis) (NOAA Fisheries 2018b). North Pacific right whales (listed as endangered) encompass two geographically distinct populations, eastern and western. Globally, this species is among the rarest of all the large whale species (NOAA Fisheries 2018b). The range of eastern North Pacific right whales is believed to encompass the Gulf of Alaska and the Bering Sea, while the western population ranges from near the Commander Islands, the coast of Kamchatka, along the Kuril Islands, and in the Sea of Okhotsk (NMFS 2017).

3-121 MIDWAY SEABIRD PROTECTION PROJECT FINAL ENVIRONMENTAL ASSESSMENT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

Both North Pacific right whale populations are at serious risk of extirpation, especially the eastern population (Reilly et al. 2008, LeDuc et al. 2012, Sekiguchi et al. 2014). The most recent stock assessment report for the eastern stock provides a best estimate of abundance of 31, with a minimum population estimate of 26 right whales (Muto et al. 2017). LeDuc et al. (2012) provided data on the genetic characteristics of a sample of 24 North Pacific right whales (eastern and western), which showed some loss in genetic diversity, and indicated the eastern stock remains at severe risk of extinction with an effective population size of only 11.6 animals (the number of individuals in a population who contribute offspring to the next generation). Critical habitat was designated for the North Pacific right whale on April 8, 2008 in the Bering Sea and the Gulf of Alaska (NOAA Fisheries 2018b). Eastern North Pacific right whales have been documented in Hawaii. A sighting occurred off Maui on 2 April 1996 (Salden, and Mickelsen 1999), and two encounters were recorded in 1979, one between the islands of Maui and Kaho'olawe on 25 March 1979 and the other southwest of Moloka'i on the Bank on 11 April 1979 (Herman et al. 1980). Rowntree et al. (1980) reported a sighting in the 'Au'au Channel off west Maui on 25 March 1979. There is speculation that the Eastern North Pacific right whale may migrate with humpback whales (Megaptera novaeangliae) into Hawaiian waters (Herman et al. 1980). One right whale observed in both Hawaii and the Bering Sea in 1996 represents the only confirmed evidence of an annual migration (Kennedy et al. 2010).

3.3.6.20.10 Existing Conditions: Giant Manta Ray There are two species of manta rays: giant manta rays (Manta birostris) and reef manta rays (Manta alfredi). Both species are recognized by their large diamond-shaped body with elongated wing- like pectoral fins, ventrally placed gill slits, laterally placed eyes, and a wide terminal mouth with two structures called cephalic lobes that bring water into the mouth for feeding activities. Manta rays also have distinct spot patterns on their bellies that can be used to identify individuals (Miller and Klimovich 2016). The giant manta ray was recently listed as Threatened throughout its range in January 2018 (NMFS-NOAA 2018a). It is the largest of the two species, reaching weights of up to 5,291 lb. (2,400 kg) and having a wingspan of up to 23 ft. (7 m) (the width is about 2.2 times the length of the body) (Marshall et al. 2018). Mantas are filter feeders that mainly eat , and unlike other rays, their tail lacks a stinger. It takes a female 8-10 years to mature; gestation is approximately one year and only one pup is birthed every 2-3 years (MarineBio 2017). The giant manta is considered migratory (Marshall et al. 2018). In the Pacific Ocean, the largest subpopulations and records of individuals of the species come from the Indo-Pacific and the eastern Pacific portion; at present, there is insufficient data to determine and designate critical habitat for this species (NMFS-NOAA 2018a). This species is circum-global in range, but within this broad distribution, individual populations are scattered and highly fragmented (CITES 2013). In Hawai‘i, the species has a high site fidelity for reefs around French Frigate Shoals (Miller and Klimovich 2016). Under threat by illegal trade practices, such as fisheries that harvest the animal’s gill rakes for use in Chinese medicine, giant mantas also face threats from the fishery industries (killed as by-catch), from a decline of offshore reefs used for feeding, cleaning, etc., from marine debris and pollution, and from a lack of prey availability (NMFS-NOAA 2018a). In 2009, Hawaii became the first of the United States to

3-122 FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

introduce a ban on the killing or capturing of manta rays. Previously, no fishery for mantas existed in the state, but migratory fish that pass the islands are now protected (Marshall et al. 2018). Because the ruling for ESA protection is so recent, recovery strategies and ongoing conservation measures include basic fact-gathering, such as further solicitation for relevant information and drafting a species recovery plan.

3.3.6.20.11 Existing Conditions: Oceanic Whitetip Shark At one time oceanic whitetip sharks were thought to have been among the most numerous pelagic sharks on the planet. Bycatch in commercial fishing practices and harvesting for the shark fin trade have all but extirpated this species (Young et al. 2016, NOAA Fisheries 2018c). In 2018, NOAA Fisheries listed the species as threatened under the ESA (NMFS-NOAA 2018b). A stocky-built shark with a bluntly rounded snout and incredibly powerful jaws, the oceanic whitetip shark is so named because of the white tips on the pectoral, first dorsal, pelvic, and caudal fins. Solitary and slow moving (Baum et al. 2015), this shark is distributed worldwide. The species is epipelagic, found mainly in offshore, tropical, and warm-temperate waters (Young et al. 2016). On occasion individuals are seen in shallower waters near land, especially around oceanic islands but are found at depths of up to 492 ft. (150 m) (NMFS-NOAA 2018b). This species has a clear preference for open ocean waters between 10˚N and 10˚S and a trend of decreasing numbers out to latitudes of 30˚N and 35˚S, with abundance decreasing with greater proximity to continental shelves (Backus et al. 1956, Strasburg 1958, Compagno 1984, Bonfil et al. 2008). Little is known about the movement or possible migration paths for oceanic whitetips. In the Pacific, Musyl et al. (2011) used pop-up satellite tags to observe this shark’s behavior and found a complex movement pattern that was generally restricted to the central region of Pacific tropical waters north of the North Equatorial Countercurrent and near to the tagging location. Oceanic whitetips are often accompanied by remoras (Echeneidae species), dolphinfish (Coryphaena species) and pilot fish (Naucrates doctor), and reportedly demonstrates an unusual association with the shortfin pilot whale (Globicephala macrorhynchus) in Hawaiian waters (Baum et al. 2015). Although the exact reason for this shark swimming along with pods of pilot whales is unknown, it is thought that oceanic whitetip sharks are following them to sources of squid, which the pilot whales are extremely efficient at locating (Young et al. 20016). Drastic declines have occurred in geographic-specific areas of the oceanic whitetip’s global population (Young et al. 2016). However, these declines are not indicative of total extirpation of the species. There appears to be some indication for relative stability in population sizes 5-10 years at the post-decline depressed state, as evidenced by the potential stabilization of two extremely small populations (e.g., NW Atlantic and Hawaii) (Young et al. 2016), which suggests that this species is potentially capable of persisting at a reduced population size (NMFS-NOAA 2018b). Genetic bottlenecks may be a cause for concern in the near future because a species already relatively low on genetic diversity that undergoes significant levels of exploitation may experience increased risk in terms of reduced fitness, evolutionary adaptability, and potential extirpations (Camargo et al. 2016). Besides bycatch and shark finning, threats to this species include loss of habitat, lack of regulatory protection, decreased genetic diversity, and climate change.

3-123 MIDWAY SEABIRD PROTECTION PROJECT FINAL ENVIRONMENTAL ASSESSMENT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

3.3.6.20.12 Probable Impacts: No Action Alternative Under the No Action Alternative, Sand Island’s mouse population would not be the subject of an island-wide eradication project, but limited mouse control efforts would continue using the less toxic AGRID3 to protect nesting seabirds and for human health and safety. No bait would be used in the vicinity of the marine environment and no additional barges or ships would be delivering equipment or supplies to Midway from Honolulu. Therefore, there would be no effects to the false killer whale, blue whale, fin whale, sei whale, sperm whale, or North Pacific right whale and no effects to the giant manta ray and oceanic white tip shark under the no-action alternative.

3.3.6.20.13 Probable Impacts: Preferred Alternative Operational Hazard (Ship Strike Only) There is a small risk injury or mortality to the false killer whale, blue whale, fin whale, sei whale, sperm whale, North Pacific right whale, giant manta ray and oceanic white tip shark from collision from a ship or barge transporting materials to and from MANWR from Honolulu for the eradication project. However, except for the false killer whale MHI Insular DPS, these other marine species are rare to infrequent visitors to Hawaiian waters and would be at a very low risk. Given the minimization measures that will be in place to protect these other marine species from ship strikes (see below), and the fact that there have been no known record of ship strikes with a marine protected species by any authorized vessel operator in the Monument, we expect the likelihood of any of these marine protected species being the victim of a ship strike to be very low. Mitigation Strike avoidance measures would be in place to avoid the risk of negative affects to endangered marine species. During transportation of supplies and materials, vessels would monitor and adjust speeds to no more than 10 knots in the presence of these protected marine species and ship captains would be mindful of maintaining a 656 ft. (200 m) buffer around species encountered at sea. Shutdown protocols would also be implemented if necessary, until species leave(s) the area. A watchful eye for these marine species would also occur when navigating Refuge waters.

3.3.6.21 Human Health and Safety

3.3.6.21.1 Existing Environment A safe environment is one in which there is little or no potential for death or serious bodily injury or illness, or property damage. Human health and safety addresses: (i) workers’ health and safety during project operations, and (ii) public safety during project operations. Project safety is managed via adherence to regulatory requirements imposed for the benefit of worker and public safety, and implementation of operational practices that reduce the potential for illness, injury, death, and property damage. The health and safety of MANWR and PMNM workers are safeguarded by the numerous regulations designed to comply with standards promulgated by OSHA and the EPA. These standards specify the amount and type of training required for industrial workers, the use of protective equipment and clothing, engineering controls, and maximum exposure limits for

3-124 FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

workplace stressors. Compliance with OSHA and other applicable laws and regulations for the protection of employees is exclusively the obligation of the commercial contractor. On Sand Island, MANWR, the various agencies and contractors have health and safety managers, and all personnel are reminded to focus on workers safety on an ongoing basis. This focus on health and safety is particularly important given the Atoll’s remote and inaccessible location. There is a doctor and/or nurse practitioner on the Atoll at all times and that individual is on call 24-hours a day and 7 days a week in case of injury or illness. All fuel deliveries operate in compliance with USFWS regulations and the Midway Atoll Spill Prevention Control and Countermeasure Plan. Safety and accident hazards can often be anticipated, identified, and either reduced or eliminated. Necessary elements for an accident-prone situation or environment include: • The presence of one or more hazards; • The presence of an exposed population; and • The absence of appropriately-observed safety procedures. The degree of hazard is largely dependent on the proximity of the hazard to the exposed population. Activities that are innately hazardous include: (i) transportation; (ii) construction; (iii) maintenance and repair activities; (iv) training; and (v) noisy operations or environments. The proper operation, maintenance, and repair of vehicles and equipment are vital to creating a safe work environment. Extremely noisy environments can damage hearing and can also mask verbal or mechanical warning signals such as sirens, bells, or horns.

3.3.6.21.2 Aerial Operations Safety Aircraft safety is based on the physical risks associated with aircraft flight, as well as operational procedures related to aircraft and aircraft maintenance safety. Weather conditions at MANWR are generally clear, with light trade winds for most of the year, interrupted by periodic rainstorms. Adverse weather conditions of concern during the time of year proposed for the Midway Seabird Protection Project include: (i) thunderstorms, (ii) hail, (iii) severe turbulence, (iv) wind shear, and (v) hurricanes. There are no obstructions, such as towers or electrical transmission lines and poles, which would cause a safety concern for helicopter operations on Sand Island, MANWR. The extreme concentration of birds is also a concern for aircraft operations and would be taken into account in planning the exact timing of the eradication operation. In order to address this, MANWR has adopted a BASH Program that includes methods to reduce the risk of BASH using harassment techniques (such as pyrotechnics) under the terms of a USFWS permit. Per the terms of that permit, the airfield is inspected and the taxiway/runway area is cleared of birds prior to aircraft operations, which occur approximately twice per month. When needed based on the result of these inspections, a sweeper vehicle is used to prevent the accumulation of Foreign Object Debris (FOD) on the runway and taxiway.

3-125 MIDWAY SEABIRD PROTECTION PROJECT FINAL ENVIRONMENTAL ASSESSMENT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

3.3.6.21.3 Ground Operations Safety Ground safety conditions on Sand Island, MANWR are subject to existing safety plans which incorporate and adhere to the requirements of OSHA and USFWS standards. All personnel working on Sand Island have attended a Basic Safety Course as part of their orientation when arriving on island, which includes lecture, discussion, and video materials intended to familiarize new arrivals with basic safety procedures and protocols. Periodically, a USFWS-designated Safety Officer will electronically disseminate messages, lessons, or other safety topics to all supervisors to be posted in work areas. Safety signs, instructions, and safety bulletin boards with required OSHA material are posted throughout the work spaces on the Atoll. Many accidents occur as a result of improper tool usage, foot injuries from rough coral present on the Atoll, and the typical lacerations and contusions resulting from construction and demolition work. Safety information and materials are available onsite for all hazardous materials used on Sand Island, MANWR, and there is an active industrial hygiene program. Emergency services such as police, fire, and emergency medical services are available on Sand Island, MANWR to allow for prompt response to emergencies, 24-hours per day. The Sand Island Fire Station is located at Henderson Field Airport and is staffed by refuge workers qualified in firefighting. All firefighters are certified with 40-hour Hazardous Waste Operations and Emergency Response Standard training and serve as the MANWR Hazardous Materials Team. Available firefighting and emergency equipment include 2 crash trucks, a structure fire truck, a water tanker, a mini-pumper, a service truck, and an ambulance. One fire truck equipped with foam is on standby at the airfield. Fire hydrants are distributed across Sand Island to provide a firefighting water supply. Sand Island, MANWR has a Physician’s Assistant (PA) and a small clinic. The PA provides medical care to all personnel on the Atoll. In addition, the PA serves as the food inspector. In the event of a serious emergency, the patient would have to be airlifted off the island to Honolulu. Emergency medical evacuation services are provided by the U.S. Coast Guard.

3.3.6.21.4 Probable Impacts If the implementation of the proposed action or alternatives were to substantially increase risks associated with the safety of personnel, contractors, or the contractor residences, or would substantially hinder the ability to respond to an emergency, it would represent a significant impact to safety on Sand Island, MANWR. Under the proposed action, short-term minor safety risks are anticipated as a result of: (i) conducting aerial broadcast operations; (ii) conducting hand broadcast operations in compromised structures; and (iii) broadcasting a toxicant, in this case brodifacoum. Specific risks could include handling the brodifacoum, loading the bait hopper, and flying the helicopter. To address this, a USFWS contractor would be required to establish and maintain safety programs and comply with all guidelines for handling bait pellets and operating or maintaining aircraft. In addition, the Contractor would comply with all existing aviation, airfield, and installation safety procedures and standards.

3-126 FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

Because the project involves the broadcast of a toxin, a potential short-term impact on safety on Sand Island, MANWR is the potential for accidental poisoning of staff and contractors. For the proposed action only USFWS and its Contractor would come in contact with brodifacoum. This contact with brodifacoum may occur in several ways, including: • Direct consumption of bait by personnel unfamiliar with its appearance. • Incidental consumption of bait through inadvertent contamination of food stocks with the toxin, whether through direct contact with bait or secondary transmission via rodent feces or urine. • Secondary ingestion of the toxin through consumption of animals that were primary consumers of the bait. While inadvertent consumption of brodifacoum by humans is unlikely, a small risk remains. The ingestion of 3.5 x 10-5 oz. (1 mg) of brodifacoum was reported to result in bleeding that persisted for more than 2 months. The brodifacoum concentration on the bait being proposed for use under the conservation label 0.0025 %, which would require that an adult ingest 1.3 lbs. (600 g) of bait to achieve an average fatal dose; this would be equivalent to ingesting 600 bait pellets. Symptoms of acute intoxication may be an increased tendency to bleed in less severe poisonings, to massive hemorrhages in more severe cases. The signs of poisoning develop with a delay of 1 to several days after ingestion. Both intentional and unintentional poisonings have been reported. To reach the toxic or lethal does, the nontarget animals must consume comparatively large amounts of bait with a concentration of 0.0025% of the active ingredient (i.e., brodifacoum). Treatment of brodifacoum poisoning includes the use of Vitamin K1 to counter the effects. Physician-controlled Vitamin K1 supplements would be available for all Sand Island, MANWR residents during and after the eradication operation. Installation staff would be educated on the entire program and how to deal with operation mishaps, accidental release or poisoning, and transport of brodifacoum into non-target areas such as food stocks or drinking water. In addition, certain restrictions would be placed on residents during the proposed action, such as limiting the number of individuals authorized to come into contact with the bait to further mitigate the potential for inadvertent safety risks. Personnel would be advised to limit their outdoor activities on those days where bait is aerially applied along recommendations to wear long sleeves and long pants. Procedures would also be implemented to protect potable water supplies (see Section 3.3.2). Over the longer-term, the proposed action would have a beneficial effect on safety by eliminating an invasive pest species that can act as a disease vector, contaminate food supplies, and damage MANWR infrastructure.

3-127 MIDWAY SEABIRD PROTECTION PROJECT FINAL ENVIRONMENTAL ASSESSMENT AFFECTED ENVIRONMENT, ENVIRONMENTAL IMPACTS, AND MITIGATION MEASURES

THIS PAGE INTENTIONALLY LEFT BLANK

3-128 FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT CUMULATIVE IMPACTS

CUMULATIVE IMPACTS

4.1 CUMULATIVE IMPACTS Federal agencies like the USFWS are required by NEPA to consider direct and indirect impacts of any actions taken, but also must consider the cumulative impacts of their actions. Cumulative impacts on environmental resources result from incremental effects of a proposed action, when combined with other past, present, and reasonably foreseeable future actions. Such impacts result from individually minor, but collectively substantial, actions undertaken over a period of time. Informed decision-making is elicited by consideration of cumulative impacts resulting from projects that are proposed, under construction, recently completed, or anticipated to be implemented. There are no future actions on Sand Island or Eastern Island that are likely to negatively impact the environment because the atoll is managed in perpetuity as a National Wildlife Refuge. Additionally, it should be noted that many species of MANWR may be experiencing unrelated impacts elsewhere in their extensive ranges, or they may still be recovering from severe past impacts. In addition, the potential numbers of individuals of a species that could be killed from either rodenticide poisoning or aircraft collisions are not cumulative, since if an individual is killed by a collision it is taken out of the population and would not be exposed to the rodenticide. The following is a breakdown of the past, present, and foreseeable future actions that would likely contribute to the cumulative impacts associated with the 3 identified alternatives. Direct and indirect impacts from each alternative would be analyzed with the following list of activities to determine the cumulative impacts for a given alternative.

4.2 PAST ACTIONS Past actions are actions that occurred in the past but have lasting impacts that could contribute to the impacts associated with the proposed action. In the mid-1990s, the USFWS successfully eradicated black rats (Rattus rattus) from MANWR using bait stations with brodifacoum and snap traps. In a study on brodifacoum bait residues before a rat eradication project on the island of Palmyra in 2011, residue concentrations decreased with time, and the toxicant was not detected in most of the 28, 36, and 50-day samples tested; and only trace amounts (≤0.2 ppm for brodifacoum) of the toxicant were detected in a few samples from these groupings. The results from the study show that following a broadcast of rodenticide across Palmyra’s emergent land area, only small amounts of brodifacoum remained in the islands topsoil for a short period of time. Given the short half-life of brodifacoum in soil, no cumulative effects from the previous attempt on MANWR are expected. Therefore, cumulative effects from this previous eradication are not expected and would not negatively contribute to the impacts from the proposed action. In addition, other small mammal eradications that have already been conducted on islands throughout the world to conserve and protect nesting birds will likely have a cumulative beneficial

4-1 MIDWAY SEABIRD PROTECTION PROJECT FINAL ENVIRONMENTAL ASSESSMENT CUMULATIVE IMPACTS

effect on global seabird populations (Brooke at al. 2017), including some of the seabird species present at MANWR.

4.3 CURRENT ACTIONS Current actions are actions that are occurring within the same timeframe as the proposed action, or within the planning and compliance phase of the proposed action and could contribute to the impacts from the proposed action.

4.3.1 PLANT RESTORATION AND INVASIVE WEED CONTROL The USFWS is conducting a project to restore native plants to MANWR, along with eradicating invasive plants, including the ironwood tree and invasive plant verbesina spp. These 2 actions could positively contribute to the cumulative impacts of the proposed action by improving habitat for native species and contributing to the overall restoration of the atoll.

4.3.2 LEAD ABATEMENT Studies conducted by USFWS and others at Sand Island between the late 1980s and 2009 have shown that Laysan albatross chicks exhibited symptoms of lead toxicity, and that their exposure is likely related to ingestion of lead-based paint (LBP) chips and soil contaminated with LBP chips. Buildings and structures on Sand Island, some of which date back to the early 1900s, are the primary source of the paint chips that contaminate the soil where the albatross nest. From 2005 to present, the USFWS has remediated all of the 95 buildings with LBP, with the total removal of all LBP from the interior, and the encapsulation of all lead-based paint on the exterior. This action contributes positively to the cumulative impacts of the proposed action by improving habitat for native species, decreasing mortality to seabirds, and contributing to the overall restoration of the atoll.

4.3.3 MARINE DEBRIS REMOVAL On a continuing basis, NOAA Coral Reef Ecosystem Program has sent their marine debris team to remove tens of thousands of pounds of debris from MANWR. Funded by the NOAA Marine Debris Program and the NOAA Papahānaumokuākea Marine National Monument, the mission of the NOAA Marine Debris Program is to investigate and prevent the adverse impacts of marine debris. In 2016, crews collected 15,206 lbs. (6,897 kg) of debris from Midway beaches. This action would positively contribute to the cumulative impacts of the proposed action by improving habitat for native species, decreasing mortality to seabirds, and contributing to the overall restoration of the atoll.

4.3.4 MOUSE CONTROL/MANAGEMENT EFFORTS MANWR currently implements several ongoing measures on Sand Island to control/manage the spread and impact of invasive house mice. These efforts are in commensal areas to maintain human health standards and are around the nests of breeding albatross to prevent predation and mortality of adults and chicks. Control measures in commensal areas include bait stations and multi-catch live-traps deployed around habitations, food stores, and at the dining hall, and in areas near

4-2 FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT CUMULATIVE IMPACTS

buildings, man-made structures, at the airport, and on shipping docks receiving conveyances. To protect albatross, control measures are used in the winter months during the nesting season (November to February); these measures include bait stations and multi-catch live traps, as well as hand applications of rodenticide in areas where mouse attacks to albatross are documented. The Preferred Alternative of targeted mouse eradication on Sand Island would have a positive effect in that it would eliminate the need for these ongoing mouse control efforts.

4.4 FUTURE ACTIONS Future actions include actions that are reasonably foreseeable in the future, and that could contribute to the cumulative impacts from the proposed action. The U.S. Fish and Wildlife Service and the Federal Aviation Administration (FAA) propose to conduct repairs as needed over a ten year period (2018-2027) along a 5,720-foot-long seawall located on Midway Atoll’s Sand Island in order to protect Henderson Field and to control erosion of wildlife habitat along the southeast side of Sand Island. The proposed seawall repair project which is currently being planned would not contribute to additional, cumulative impacts when considered in combination with the proposed Midway Seabird Protection Project. There is not expected to be any persisting effects of the Midway Seabird Protection Project at the time that the seawall project is implemented and whatever impacts result from the seawall project will not “accumulate” with any impacts of the Midway Seabird Protection Project. Areas of impact linked to future global climate change, which may have the potential to affect MANWR, include warmer air temperatures and declines in rainfall (University of Hawaiʻi 2014). In addition, other small mammal eradications being planned on islands throughout the world to conserve and protect nesting birds will likely have a cumulative beneficial effect on global seabird populations as has been shown by Brooke at al. (2017).

4.5 PROBABLE CUMULATIVE IMPACTS

4.5.1 NO ACTION ALTERNATIVE Under the no-action alternative, the current negative impacts of mice on MANWR’s seabird populations and terrestrial ecosystem would continue in perpetuity. These impacts could be additive to other, unrelated future impacts on the resources of MANWR. The minor impacts that ongoing projects (primarily conservation-related) would have on the biological, physical, and cultural resources of MANWR are not likely to contribute to mice-related impacts. If mice persist on Sand Island, the biological resources of the island would continue to be negatively affected. Under the No Action Alternative, Sand Island’s mouse population would not be the subject of a targeted eradication project, but mouse control efforts to protect breeding albatross, which were started in 2016, would continue and expand as deemed appropriate. Mouse management on Sand Island currently consists of (i) trapping and rodenticide bait stations containing AGRID3 (active ingredient cholecalciferol) for human health and safety in commensal areas such as food storage and preparation areas, as well as in areas near buildings, man-made structures, at the airport, and on shipping docks receiving conveyances; and (ii) multi-catch live trapping, and rodenticide bait

4-3 MIDWAY SEABIRD PROTECTION PROJECT FINAL ENVIRONMENTAL ASSESSMENT CUMULATIVE IMPACTS

stations and hand broadcasting of AGRID3 for seabird and listed candidate plant protection in areas where mouse predation of seabirds and mouse damage to listed or candidate plants are detected. Trapping (i) includes multi-catch live traps and mechanical traps (snap-traps and glue boards) used in the commensal environment as deemed appropriate per the terms of each structure’s SMP (see Section 2.3.1). Hand broadcasting (ii), which is extremely labor and time intensive, involves 2 separate hand-broadcast applications of AGRID3 pellets approved for restricted use at MANWR under a supplemental label for mouse control. Per label instructions, a 17 lbs./ac. (20 kg/ha) application rate is used over 27.2 ac. (11 ha) where evidence of mouse attacks occurred in 2016. The second application (same rate, same area) occurs in the month of November prior to the albatross’s main egg laying season, and then only as needed if and where there were mouse attacks throughout the incubation period, up until the following February. There were no observations of any non-target organism, such as shorebirds or Laysan ducks, interacting with bait pellets in the field, or being found sick or dead in the colony as a result of the baiting process (USFWS, Unpublished b). Both control measures, (i) and (ii), are ongoing. The proposed action of targeted mouse eradication on Sand Island would have a positive effect in that it would eliminate the need for these ongoing control efforts.

4.5.2 PREFERRED ALTERNATIVE There would be no major adverse impacts to the biological, physical, or cultural resources of Sand Island under the Preferred Alternative. The minor negative impacts to biological, physical, and cultural resources as a result of the Preferred Alternative would not contribute to the impacts related to any separate, current, or future projects. Similarly, the expected positive impacts of the Preferred Alternative to Sand Island’s biological resources could contribute to the cumulative, positive impacts from separate, current, or future projects.

4-4 FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT LIST OF PREPARERS AND CONTRIBUTORS

LIST OF PREPARERS AND CONTRIBUTORS

5.1 PREPARERS Table 5.1 below identifies the principal agencies, organizations, and individuals that participated in the preparation of this report.

Table 5.1 Principal Prepares of this Environmental Assessment Agency or Organization Preparers U.S. Fish and Wildlife Service, Pacific Islands Michael Marxen Fish and Wildlife Office Beth Flint Thomas Hamer Hamer Environmental Kristin Murray Cary Deringer Jim Hayes Planning Solutions, Inc. Mākena White Source: Compiled by Hamer Environmental (2017)

5.2 CONTRIBUTORS Table 5.2 below identifies the agencies, organizations, and individuals that contributed critical information for the development of this report.

Table 5.2 Contributors to this Environmental Assessment Agency or Organization Preparers Susan Hunter Kelly Goodale Kauaoa Fraiola Amanda Pollock U.S. Fish and Wildlife Service Leila Nagatani Michael Green Jennifer Miller Roberta Swift Patty Baiao Island Conservation Gregg Howald Wes Jolley Source: Compiled by Hamer Environmental (2017)

5-1

THIS PAGE INTENTIONALLY LEFT BLANK

5-2 FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT REFERENCES

REFERENCES ACAP. 2012a. Black-footed Albatross (Phoebastria nigripes). Agreement on the Conservation of Albatrosses and Petrels [cited September 2017]. Available from http://www.acap.aq/en/acap-species. ---. 2012b. Laysan Albatross (Phoebastria immutabilis). Agreement on the Conservation of Albatrosses and Petrels [cited September 2017]. Available from http://www.acap.aq/en/acap- species. Alifano, A., A. Wegmann, B. Puschner, and G. Howald. 2012. “Migration of Brodifacoum and Diphacinone from Bait Pellets into Topsoil at Palymyra Atoll National Wildlife Refuge. In Proc. 25th Vertebrate Pest Conference, edited by R.M. Timm: University of California, Davis. Pp. 139-143. Amarasekare, P. 1994. “Ecology of Introduced Small Mammals on Western Mauna Kea, Hawai'i.” J. Mammal. 75:24-38. Amerson, A.B., Jr., R.B. Clapp, and Wirtz. W.O. 1974. "The natural history of Pearl and Hermes Reef, Northwestern Hawaiian Islands." Atoll Res Bull 174:1-306. Anderson, B., S. Borges, and K. Garber. 2011. “Risks of Non-Compliant Rodenticides to Nontarget Wildlife.” USEPA. Andrews, C. 1909. “On the fauna of Christmas Island.” Proceedings of the Zoological Society of London:101-103. Angel, A., R.M. Wanless, and J. Cooper. 2009. “Review of impacts of the introduced house mouse on islands in the Southern Ocean: are mice equivalent to rats.” Biological Invasions 11:1743-1754. Apfelbaum, S.I., J.P. Ludwig, and C.E. Ludwig. 1983. Ecological problems associated with disruption of dune vegetation dynamics by Casuarina equisetifolia L. at Sand Island, Midway Atoll. Washington DC: Smithsonian Institution. Arata, J. A., Sievert P.R., and M.B. Naughton. 2009. Status assessment of Laysan and black- footed albatrosses, North Pacific Ocean, 1923-2005. http://www.iucnredlist.org/details/biblio/22698365/0. Arenas, P., and M. Hall. 1992. The association of sea turtles and other pelagic fauna with floating objects in the eastern tropical Pacific Ocean. Pages 7-10 in M. Salmon and J. Wyneken (compilers), Proc. Eleventh Annual Workshop on Sea Turtle Biology and Conservation. U.S. Dep. Comm., NOAA Tech. Memo. NMFS-SEFSC-302. 195 pp. Arthur, K.E., and G.H. Balazs. 2008. "A comparison of immature green turtle (Chelonia mydas) diets among seven sites in the Main Hawaiian Islands." Pac. Sci. 62:205-217. Aruch, S. 2006, in litt. "Email to C. Russell regarding candidate plant species on Moloka‘i, Pseudognaphalium sandwicensium var. molokaiense and Solanum nelsonii."1 p. Ashmole, N.P., and M.J. Ashmole. 1967. Comparative feeding ecology of seabirds of a tropical oceanic island. Peabody Museum Bulletin 24: 1-131.

6-1 MIDWAY SEABIRD PROTECTION PROJECT FINAL ENVIRONMENTAL ASSESSMENT REFERENCES

Atkinson, I. 1985. “The spread of commensal species of Rattus to oceanic islands and their effects on island avifaunas.” In Conservation of Island Birds, edited by P. J. Moors, pp 35-81. International Council for Bird Preservation: Cambridge, UK. Austin, O.L. 1949. "The Status of Steller’s Albatross." Pacific Science 3:283-295. Avenant, N.L., and V.R. Smith. 2003. "The microenvironment of house mice on Marion Island (sub-Antarctic)." Polar Biology 26:129-141. Backus, R.H., S. Springer, and E.L. Arnold Jr. 1956. A contribution to the natural history of the white-tip shark, Pterolamiops longimanus (Poey). Deep-sea Research, 3, 178-188. Baker, H., P.E. Baker, and N. Seto. 1999. "Common Myna on Midway Atoll NWR, 1996." 'Elepaio 59:17-19. Baker, J.D., and T.C. Johanos. 2004. "Abundance of the Hawaiian monk seal in the main Hawaiian Islands." Biological Conservation 116:103-110. Balazs, G.H. 1973. Status of marine turtles in the Hawaiian Islands. 'Elepaio [J. Hawaii Audubon Soc.] 33(12):1-5. ---. 1980. Synopsis of biological data on the green turtle in the Hawaiian Islands. U.S. Dep. Commer., NOAA Tech. Memo. NMFS, NOAA-TM-NMFS-SWFC-7. 141 pp. ---. 1982. Growth rates of immature green turtles in the Hawaiian Archipelago. In Biology and conservation of sea turtles (K. A. Bjorndal, ed.), p. 117–125. Smithsonian Inst. Press, Washington, D.C. ---. 1985. Impact of ocean debris on marine turtles: entanglement and ingestion. Pages 387- 429 in R. S. Shomura and H. O. Yoshido, editors. Proceedings of the workshop on the fate and impact of marine debris. U.S. Department of Commerce, National Oceanic and Atmospheric Administration (NOAA) technical memorandum 54. National Marine Fisheries Service, Honolulu ---. 1994. Homeward bound: satellite tracking of Hawaiian green turtles from nesting beaches to foraging pastures. In Proceedings of the Thirteenth Annual Symposium on Sea Turtle Biology and Conservation. Pages 205-208 in B. A. Schroeder and B. E. Witherington (compilers): NOAA Tech. Memo. NMFS-SEFSC-341. 281 pp. Balazs, G.H., R.G. Forsyth, and A.K.H. Kam. 1987. Preliminary assessment of habitat utilization by Hawaiian green turtles in their resident foraging pastures. U.S. Dep. Commer., NOAA Tech. Memo. NMFS, NOAA-TM-NMFS-SWFC-71. 107 pp. Balazs, G.H., H. Hirth, P. Kawamoto, E. Nitta, L. Ogren, R. Wass, and J. Wetherall. 1992. Interim Recovery Plan for Hawaiian Sea Turtles. Honolulu Lab., Southwest Fish. Sci. Cent., Natl. Mar. Fish. Serv., NOAA, Honolulu, HI; Southwest Fish. Sci. Cent. Admin. Rep. H-92- 01. 76 pp. Balazs, G.H., W.C. Dudley, L.E, Hallacher, J.P. Coney, and S.K. Koga. 1994. Ecology and cultural significance of sea turtles at Punalu’u, Hawaii. U.S. Dep. Commer., NOAA Tech. Memo. NMFS-SEFSC-351:10–13.

6-2 FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT REFERENCES

Balazs, G.H., K.S. Van Houtan, S.A. Hargrove, S.M. Brunson, and S.K.K. Murakawa. 2015. "A review of the demographic features of Hawaiian green turtles (Chelonia mydas) " Chelonian Conserv. Biol. (14):119−129. Bamford, M., D. Watkins, W. Bancroft, G. Tischler, and J. Wahl. 2008. Migratory Shorebirds of the East Asian - Australasian Flyway; Population Estimates and Internationally Important Sites. Canberra, Australia: Wetlands International - Oceania. Baum, J., E. Medina, J.A. Musick, and M. Smale. 2015. Carcharhinus longimanus. The IUCN Red List of Threatened Species 2015: e.T39374A85699641. http://dx.doi.org/10.2305/IUCN.UK.2015.RLTS.T39374A85699641.en. BirdLife International. 2015. Bubulcus ibis. The IUCN Red List of Threatened Species [cited 06 October 2017]. ---. 2016a. Anas laysanensis. The IUCN Red List of Threatened Species: e.T22680203A92849158 [cited Feb 22 2018]. Available from http://dx.doi.org/10.2305/IUCN.UK.2016-3.RLTS.T22680203A92849158.en. ---. 2016b. Numenius tahitiensis. The IUCN Red List of Threatened Species [cited 07 Oct 2017]. Available from http://dx.doi.org/10.2305/IUCN.UK.2016-3.RLTS.T22693182A95218288.en. ---. 2016c. Serinus canaria. The IUCN Red List of Threatened Species [cited 08 October 2017]. Available from http://dx.doi.org/10.2305/IUCN.UK.2016-3.RLTS.T22720056A88429138.en. Birkeland, C., and D. Grosenbaugh. 1985. Ecological interactions between tropical coastal ecosystems. United Nations Environmental Program. Blackburn, T.M., P. Cassey, R.P. Duncan, K.L. Evans, and Gaston. K.J. 2004. “Avian extinction and mammalian introductions on oceanic islands.” Science 305:1955-1958. Bonfil, R., S. Clarke, and H. Nakano. 2008. The biology and ecology of the oceanic whitetip shark, Carcharhinus longimanus. In: Sharks of the Open Ocean: Biology, Fisheries, and Conservation. M.D. Camhi, E.K. Pikitch and E.A. Babcock (eds): Blackwell Publishing. pp. 128-139. Booth, L.H., F. Fisher, V. Heppelthwaite, and C.T. Eason. 2003. Toxicity and residues of brodifacoum in snails and earthworms. In Doc Science Internal Series Wellington, New Zealand: Department of Conservation. Bremner, A., C. Butcher, and G. Patterson. 1984. “The density of indigenous invertebrates on three islands in Breaksea Sound, Fiordland, in relation to the distribution of introduced mammals.” Journal of the Royal Society of New Zealand 14:379-386. Brooke, M. 2004. Albatrosses and petrels across the world. New York, N.Y.: University Press. 499 pp. Brooke, M., R. Cuthbert, G. Harrison, C. Gordon and M. Taggart. 2013. Persistence of brodifacoum in cockroach and woodlice: Implications for secondary poisoning during rodent eradications. Ecotoxicology and Environmental Safety: Volume 97:183-188. https://doi.org/10.1016/j.ecoenv.2013.08.007

6-3 MIDWAY SEABIRD PROTECTION PROJECT FINAL ENVIRONMENTAL ASSESSMENT REFERENCES

Brooke, M., E. Bonnaud, B.J. Dilley, E.N. Flint, N.D. Holmes, H.P. Jones, P. Provost, G. Rocamora, P.G. Ryan, C. Surman, and R.T. Buxton. 2017. "Seabird population changes following mammal eradications on islands." Animal Conservation. doi: 10.1111/acv.12344. Broome, K.G., A. Cox, C. Golding, P. Cromarty, P. Bell, and P. McClelland. 2014. Rat eradication using aerial baiting: Current agreed best practice used in New Zealand (Version 3.0). New Zealand Department of Conservation internal document: Wellington, New Zealand. Brown, K. 1997. Impact of brodifacoum poisoning operations on South Island Robins (Petroica australis australis) in a New Zealand Nothofagus forest. International 7(04):399 – 407. DOI:10.1017/S0959270900001726 Brown, P.R., and G.R. Singleton. 1999. "Rate of increase as a function of rainfall for house mouse Mus domesticus populations in a cereal-growing region in southern Australia." Journal of Applied Ecology 36 (4):484-493. doi: 10.1046/j.1365-2664.1999.00422.x. Brownell, R.L., P.J. Clapham, T. Miyashita, and T. Kasuya. 2001. Conservation status of North Pacific right whales. Journal of Cetacean Research and Management (Special Issue 2):269- 286. Bruegmann, M. 1998. Report on a botanical survey of Midway Atoll, April 1-7, 1995. Honolulu, HI: U.S. Fish and Wildlife Service Report. Bryan, W.A. 1905. "Notes on the American birds collected in the Hawaiian Islands by Mr. Gerrit Wilder." Occas. Pap. B. P. Bishop Mus. 2:245-248. Buckle, A.P., and R. Smith. 2015. Rodent pests and their control. 2nd ed. CAB International: Oxfordshire, UK. Buckelew, S., Howald G.R., Wegmann A., Sheppard J., Curl J., McClelland P., Tershy B., Swift K., Campbell E., and Flint B. 2005. Progress in Palmyra Atoll restoration: rat eradication trial, 2005. Island Conservation unpublished report: Santa Cruz, CA. 50 pp. Bullock, D. 1986. “The ecology and conservation of reptiles on Round Island and Gunner's Quoin, Mauritius.” Biological Conservation 37:135-156. Byrd, G.V. 1979. “Common Myna predation on Wedge-tailed Shearwater eggs.” `Elepaio 39:69- 70. Campbell, D., and I. Atkinson. 2002. “Depression of tree recruitment by the Pacific rat (Rattus exulans peale) on New Zealand's northern offshore islands.” Biological Conservation 107 (19-35). Campbell, D., H. Moller, G.W. Ramsay, and J. Watt. 1984. “Observations on food of kiore (Rattus exulans) found in husking stations on northern offshore islands of New Zealand.” New Zealand Journal of Ecology 7:131-138. Campbell, E. 1991. “The effects of introduced roof rats on bird diversity of Antillean Cays.” Journal of Field 62:343-348. Chardine, J.W., and R.D. Morris. 1996. Brown Noddy (Anous stolidus), version 2.0. In The Birds of North America, edited by P. G. Rodewald. Cornell Lab of Ornithology: Ithaca, New York, USA. https://doi.org/10.2173/bna.220.

6-4 FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT REFERENCES

Christophersen, E., and E.L. Caum. 1931. "Vascular plants of the Leeward Islands, Hawai'i." Bernice P. Bishop Museum Bul 81:41. CITES. 2013. Consideration of proposals for amendment of Appendices I and II: Manta Rays. Sixteenth meeting of the Conference of the Parties. Bangkok, , March 3-14. Clapp, R.B., E. Kridler, and R.R. Fleet. 1977. The natural history of Nihoa Island, Northwestern Hawaiian Islands. In Atoll Research Bulletin 207. The Smithsonian Institution, Washington, DC. Clapp, R.B., and W.O. Wirtz, II. 1975. "The natural history of Lisianski Island, Northwestern Hawaiian Islands." Atoll Res. Bull. 186:1-196. Collar, N.J., L.P. Gonzaga, N. Krabbe, A.M. Nieto, L.G. Naranjo, T. A. P. III, and D.C. Wege. 1992. Threatened Birds of the Americas, The ICBP/IUCN Red Data Book, Part 2. Third edition. Smithsonian Institution Press: Washington, DC. Compagno, L.J.V. 1984. FAO species catalogue. Vol. 4. Sharks of the world. An annotated and illustrated catalogue of shark species known to date. FAO Fisheries Synopsis No. 125, Volume 4, Part 1. Conant, S., C.C. Christiansen, P. Conant, W.C. Gagne, and M.L. Goff. 1983. The Unique Terrestrial Biota of the Northwestern Hawaiian Islands. Proceedings of the Second Symposium on Resource Investigations in the Northwestern Hawaiian Islands, at Honolulu, HI. pp 491. Conry, P. 2012, in litt. "State of Hawai‘i comments on USFWS species assessment and listing priority assignment forms."CNOR 2012, Apr 9 2012. Cree, A., C. Daughtery, and J.M. Hay. 1995. “Reproduction of a rare New Zealand reptile, the tuatara Sphenodon punctatus on rat-free and rat-inhabited islands.” Conservation Biology 9:373-383. Cromarty, P. L., K.G. Broome, A. Cox, R.A. Empson, W.M. Hutchinson, and I. McFadden. 2002. “Eradication planning for invasive alien animal species on islands - the approach developed by the New Zealand Department of Conservation.” In Turning the Tide: The Eradication of Invasive Species, edited by C. R. Veitch and M. N. Clout, pp 85-91. IUCN SSC Invasive Species Specialist Group, Gland: Switzerland and Cambridge, UK. Daniel, M., and G. Williams. 1984. “A survey of the distribution, seasonal activity and roost sites of New Zealand bats.” New Zealand Journal of Ecology 7:9-25. Day, T.R., and L.R. Matthews. 1999. "Do colours that deter birds affect cereal bait acceptance by possums (Trichosurus vulpecula)." New Zealand Journal of Ecology 23. Deaton, J.W., F. N. Reece, E.H. Mcnally and W.J. Tarver. 1969. Liver, heart and adrenal weights of broilers reared under constant temperatures. Science 48(1): 283–288. de Bruyn, P.J.N., A.D.S. Bastos, C. Eadie, C.A. Tosh, and M.N. Bester. 2008. "Mass Mortality of Adult Male Subantarctic Fur Seals: Are Alien Mice the Culprits." PLoS ONE 3 (11):e3757. doi: 10.1371/journal.pone.0003757. Delany, S., and D. Scott. 2006. Waterbird population estimates. Wetlands International. Wageningen, The Netherlands.

6-5 MIDWAY SEABIRD PROTECTION PROJECT FINAL ENVIRONMENTAL ASSESSMENT REFERENCES

DIISE. 2016. Database of Island Invasive Species Eradications. Available from http://diise.islandconservation.org. Donlan, C.J., G. Howald, B. Tershy, and D. Croll. 2003. “Evaluating alternative rodenticides for island conservation: roof rat eradication from the San Jorge Islands, Mexico.” Biological Conservation 114:29-34. Drost, C.A., and G.M. Fellers. 1991. "Density Cycles in an Island Population of Deer Mice, Peromyscus maniculatus." Oikos 60 (3):351-364. doi: 10.2307/3545078. DuhrSchultz, M., B. Flanders, G. Howald, E.N. Flint, D. Norwood, R. Taylor, S.A. Hunter. In Press. Interim Control of House Mice Preying on Adult Albatrosses at Midway Atoll Prior to an Eradication Attempt. Duncan, R., and T. Blackburn. 2007. “Causes of extinction in island birds.” Animal Conservation Biology 10:149-150. Eason, C.T., E.C. Murphy, G.R. Wright, E.B. Spurr. 2002. Assessment of Risks of Brodifacoum to Non-target Birds and Mammals in New Zealand. Ecotoxicology. 2002 Feb;11(1):35-48. Eason, C., and E. Spurr. 1995. “Review of the toxicity and impacts of Brodifacoum on nontarget wildlife in New Zealand.” New Zealand Journal of Zoology 22:371-379. Eason, C., and M. Wickstrom. 2001. Vertebrate Pesticide Toxicology Manual (Poisons), Technical Series 23. Wellington, New Zealand. 120pp. Eason, C.T., M. Wickstrom, R. Henederson, L. Milne, and D. Arthur. 2000. "Non-target and secondary poisoning risks associated with cholecalciferol." New Zealand Plant Protection 53:299-304. Ebert, D.A. 2003. Sharks, Rays, and Chimaeras of California. Issue 71, California Natural History Guides. University of California Press. ISBN 978-0-520-23484-0. 248 pp. Eijzenga, H. 2011. Vegetation change following rabbit eradication on Lehua Island, Hawaiian Islands. Island invasives: eradication and management, at IUCN, Gland, Switzerland. pp 290- 294. Eisemann, J.D., and C.E. Swift. 2006. "Ecological and human health hazards from broadcast application of 0.005% diphacinone rodenticide baits in native Hawaiian ecosytems." Proceedings of the Vertebrate Pest Conference 22:413-433. Ely, C.A., and R.B. Clapp. 1973. "The natural history of Laysan Island, Northwestern Hawaiian Islands.". Atoll Res. Bull. 171:1-361. Empson, R.A., and C.M. Miskelly. 1999. “The risks, costs and benefits of using Brodifacoum to eradicate rats from Kapiti Island, New Zealand.” New Zealand Journal of Ecology 23:241- 254. Engeman, R., W.C. Pitt, A.R. Berentsen, and J.D. Eisemann. 2013. Assessing spatial variation and overall density of aerially broadcast toxic bait during a rat eradication on Palmyra Atoll. USDA National Wildlife Research Center Staff Publications 2013. Available from http://digitalcommons.unl.edu/icwdm_usdanwrc/1131.

6-6 FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT REFERENCES

Engilis, Jr., A., and M.B. Naughton. 2004. U.S Pacific Islands Regional Shorebird Conservation Plan. In U.S. Shorebird Conservation Plan. Portland, OR: U.S. Department of the Interior, Fish and Wildlife Service. Erickson, W., and D. Urban. 2004. Potential risks of nine rodenticides to birds and nontarget mammals: a comparative approach. Edited by Pesticides Programs U.S. Environmental Protection Agency. Environmental Fate and Effects Division: Washington, DC. European Parliament and Council. 2010. Brodifacoum (PT 14) Assessment report. Standing Committee on Biocidal Products. Directive 98/8/EC concerning the placing of biocidal products on the market. 365 pp. Fisher, H.I. 1949. "Populations of birds on Midway and the man-made factors affecting them." Pacific Science 3:103-110. Fisher, H.I., and M.L. Fisher. 1969. “The visits of Laysan Albatrosses to the breeding colony.” Micronesia 5 (1):173-221. Fisher, P., C. O’Connor, G. Wright, and C. Eason. 2003. Persistence of four anticoagulant rodenticides in the livers of laboratory rats. DOC Science Internal Series 139. New Zealand Department of Conservation: Wellington, NZ. Fleet, R.R. 1974. "The Red-Tailed Tropicbird on Kure Atoll." Ornithological Monographs. 16:1- 64. Fortini, L., J. Price, J. Jacobi, A. Vorsino, J. Burgett, K. Brinck, F. Amidon, S. Miller, S.`Ohukani`ohi`a Gon III, G. Koob, and E. Paxton. 2013. A landscape-based assessment of climate change vulnerability for all native Hawaiian plants. Technical Report HCSU-044. Hawaiʽi Cooperative Studies Unit: University of Hawaiʽi at Hilo. 142 pp. Fukami, T., D. Wardle, P. Bellingham, C. Mulder, D. Towns, G. Yeates, K. Bonner, M. Durrett, M. Grant-Hoffman, and W. Williamson. 2006. “Above- and below-ground impacts of introduced predators in seabird-dominated island ecosystems.” Ecology Letters 9 (1299- 1307). Gale, R.W., M. Tanner, and C.E. Orazio. 2008. Determination of Diphacinone in Sea Water, Vertebrates, Invertebrates, and Bait Pellet Formulations Following Aerial Broadcast on Mokapu Island, Molokai, Hawai’i. edited by U.S. Geological Survey Open-File Report 2008- 1285. Gattuso, J.P., M. Frankignoulle, I. Bourge, S. Romaine, and R.W. Buddemeier. 1998. “Effect of calcium carbonate saturation of seawater on coral calcification.” Global and Planetary Change 18:37-46. Gauger Metz, V.H., and E.A. Schreiber. 2002. Great Frigatebird (Fregata minor). In The Birds of North America, edited by P. G. Rodewald. Cornell Lab of Ornithology: Ithaca, New York, USA. https://doi.org/10.2173/bna.681. Gauger, V.H. 1999. Black Noddy (Anous minutus). Version 2.0. Available from https://doi.org/10.2173/bna.412. GeoEngineers, Inc. 2011. Engineering Evaluation/Cost Analysis, Midway Atoll National Wildlife Refuge, Midway Island. In RG for U.S. Fish and Wildlife Service. Prepared by Joey F. Hickey, Neil F. Morton, and John H. Biggane: January 18, 2011.

6-7 MIDWAY SEABIRD PROTECTION PROJECT FINAL ENVIRONMENTAL ASSESSMENT REFERENCES

Gill, R., R. Lanctot, J. Mason, and C. Handel. 1990. “Observations on habitat use, breeding chronology, and parental care in bristle-thighed curlews on the Seward Peninsula, Alaska.” Wader Study Group Bulletin 61:28-36. Gill, R.E., B.J. McCaffery, and P.S. Tomkovich. 2002. Wandering Tattler (Tringa incana). version 2.0. Cornell Lab of Ornithology. Available from https://doi.org/10.2173/bna.642. Godwin, S. 2013. Inventory of shoreline and shallow subtidal marine invertebrates at proposed seawall repair site - Midway Atoll runway. Prepared for: Midway Atoll National Wildlife Refuge: U.S. Fish and Wildlife Service: NOAA-NOS-ONMS. Resource Protection Program, Papahānaumokuākea Marine National Monument. 12 pp. Goodale, K. 2017. Shorebird survey methodology, Midway Atoll. USFWS. Honolulu, HI. Unpublished data. Gross, A.O., J.M. Moulton, and C.E. Huntington. 1963. Notes on the Wedge-tailed Shearwater at Heron Island, Great Barrier Reef, Australia. In Atoll Research Bulletin, No. 99. Washington, D.C.: The Pacific Science Board, National Academy of Sciences - Natural Research Council. HBMP. 2010. "Hawaiian species database, GIS shapefiles and database, unpublished." Hawai‘i Biodiversity and Mapping Program. HDLNR. 2015. Hawai‘i’s State Wildlife Action Plan. Prepared by H. T. Harvey and Associates, Honolulu, HI: Hawai‘i Department of Land and Natural Resources. Herman, L.M., C.S. Baker, P.H. Forestell, and R.C. Antinoja. 1980. Right whale Balaena glacialis sightings near Hawaii: A clue to the wintering grounds? Mar. Ecol. Prog. Ser. 2: 271-275. https://www.int-res.com/articles/meps/2/m002p271.pdf. Hoare, J., and K. Hare. 2006. “The impact of Brodifacoum on non-target wildlife: gaps in knowledge.” New Zealand Journal of Ecology 30:157-167. Holway, D.A., L. Lach, A.V. Suarez, N.D. Tsutsui, and T.J. Case. 2002. “The Causes and Consequences of Invasions.” Annu. Rev. Ecol. Syst. 33:181-233. Horak, K., P. Fisher, and B. Hopkins. 2018. Pharmacokinetics of Anticoagulant Rodenticides in Target and Nontarget Organisms. USDA National Wildlife Research Center - Staff Publications. 2091. https://digitalcommons.unl.edu/icwdm_usdanwrc/2091. Howald, G., A. Samaniego, S. Buckelew, P. McClelland, B. Keitt, A. Wegmann, W. Pitt, D. Vice, E. Campbell, K. Swift, and S. Barclay. 2004. Palmyra Atoll rat eradication assessment trip report August 2004. Island Conservation: Santa Cruz, CA. Howald, G., C.J. Donlan, J.P. Galván, J. Russell, J. Parkes, A. Samaniego, Y. Wang, D. Veitch, P. Genovesi, M. Pascal, A. Saunders, and B. Tershy. 2007. “Invasive rodent eradication on islands.” Conservation Biology 21:1258-1268. Howald, G., C.J. Donlan, K.R. Faukner, S. Ortega, H. Gellerman, D.A. Croll, and B.R. Tershy. 2010. "Eradication of black rats Rattus rattus from Anacapa Island." Oryx 44:30-40. Howald, G.R., K. Faulkner, B.R. Tershy, B.S. Keitt, H. Gellerman, E.M. Creel, M. Grinnell, S. Ortega, and D.A. Croll. 2005. Eradication of Black Rats from Anacapa Island: Biological and Social Considerations. Proceedings of the Sixth California Islands Symposium, at Ventura, CA, December 1-3, 2003. pp 299-312.

6-8 FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT REFERENCES

Howald, G.R., P. Mineau, J.E. Elliot, and K.M Cheng. 1999. “Brodifacoum Poisoning of Avian Scavengers During Rat Control on a Seabird Colony.” Ecotoxicology 8:431-447. Howell, T.R. 1978. “Ecology and reproductive behavior of the White, or Fairy, Tern.” In Ecology and reproductive behaviour of the Gray Gull and of the Red-tailed Tropicbird and the White Tern on Midway Island, pp 274-284. Natl. Geogr. Soc. Res. Rep. Hutton, J.C., M. Pershavaria, and K.W. Brocklehurst. 1984. Biochem. J. 224:483-490. Island Conservation. 2017. Feasibility Report: House Mouse Eradication on Midway Atoll National Wildlife Refuge. Report prepared for the U.S. Fish and Wildlife Service. 69 pp. IUCN. 2017a. The IUCN Red List of Threatened Species (Version 2017-1) [cited 27 Sept 2017]. ---. 2017b. Phoebastria immutabilis (amended version published in 2016) [cited 05 Oct 2017]. Johanos, T. 2017. Pacific Islands Fisheries Science Center. PIFSC Internal Report, IR-17-021. 3 pp. Johnsgard, P.A. 1981. , Hemipodes, and ; birds of dry places. Oxford University Press: London. Johnson, O.W., and P.G. Connors. 2010. Pacific Golden-Plover (Pluvialis fulva), version 2.0. Cornell Lab of Ornithology. Available from https://doi.org/10.2173/bna.202. Jouventin, P., J. Bried, and T. Micol. 2003. “Insular bird populations can be saved from rats: a long-term experimental study of White-chinned Petrels aequinoctialis on Ile de la Possession (Crozet archipelago).” Polar Biology 26:371-378. Kalinin, S., N. Marangoni, K. Kowal, A. Dey, K. Lis., S. Brodsky, R. Van Breemen, Z. Hauck, R. Ripper, I. Rubinstein, G. Weinberg, and D. Feinstein. 2017. The Long-Lasting Rodenticide Brodifacoum Induces Neuropathology in Adult Male Rats. Toxicol Sci.;159(1):224-237. doi: 10.1093/toxsci/kfx134. Kannan, R., and D.A. James. 2001. "Common Myna." Birds N. Am. 58:1-20. Kaukeinen, D.E. 1993. Nontarget Organism Evaluations for Rodenticides. In Pesticides in Urban Environments. American Chemical Society. Pp 352-363. Keitt, B., R. Griffiths, S. Boudjelas, K. Broome, S. Cranwell, J. Millett, W. Pitt, and A. Samaniego-Herrera. 2015. “Best practice guidelines for rat eradication on tropical islands.” Biological Conservation 185:17-26. doi:10.1016/j.biocon.2014.10.014. Kennedy, A.S., B.K. Rone, A. Zerbini, and P.J. Clapham. 2010. Eastern North Pacific right whales (Eubalaena japonica): They really do exist! Pages 148 in Alaska Marine Science Symposium, Anchorage, Alaska. Kepler, A., and C. Kepler. 1994. “Natural history of Caroline Island. Part I: History, physiography, botany, and isle description.” Atoll Research Bulletin 397:1-225. Kessel, B. 1989. Birds of the Seward Peninsula, Alaska. University of Alaska: Fairbanks, AK. Klavitter, J.L. 2004. Midway Atoll National Wildlife Refuge Wildlife Assessment for Henderson Field Airport Operations. Honolulu, HI: USFWS Midway Atoll National Wildlife Refuge.

6-9 MIDWAY SEABIRD PROTECTION PROJECT FINAL ENVIRONMENTAL ASSESSMENT REFERENCES

---. 2103. Supplemental marine shallow subtidal observations at the proposed seawall repair site - Midway Atoll National Wildlife Refuge Henderson Field Runway. Midway Atoll National Wildlife Refuge, U.S. Fish and Wildlife Service. Kurle, C.M., D.A. Croll, and B.R. Tershy. 2008. “Introduced rats indirectly change marine rocky intertidal communities from algae- to invertebrate-dominated.” PNAS 105:3800-3804. LeDuc, R.G., B.L. Taylor, K.K. Martien, K.M. Robertson, R.L. Pitman, J.C. Salinas, A.M. Burdin, A.S. Kennedy, P.R. Wade, P.J. Clapham, R.L. Brownell Jr. 2012. Genetic analysis of right whales in the eastern North Pacific confirms severe extirpation risk. Endangered Species Research 18:163-167. https://www.int-res.com/articles/esr2012/18/n018p163.pdf. Lindsey, G.D., and S.M. Mosher. 1994. "Tests indicate minimal hazard to 'io from diphacinone baiting." Hawaii's Forests and Wildlife 9 (4). Mackay, J.W.B. 2011. Improving the success of mouse eradication attempts on islands, PhD diss., The University of Auckland, Auckland, New Zealand. Macwhirter, R.B., P. Austin-Smith Jr., and D.E. Kroodsma. 2002. Sanderling (Calidris alba), version 2.0. Cornell Lab of Ornithology. Available from https://doi.org/10.2173/bna.653. Marine Mammal Center. 2018. Hawaiian Monk Seal [cited Mar 8 2018]. Available from http://www.marinemammalcenter.org/education/marine-mammal- information/pinnipeds/hawaiian-monk-seal/. MarineBio. 2017. "Giant Mantas, Manta birostris ~ MarineBio.org." MarineBio Conservation Society. Web. Accessed Monday, October 29, 2018. . Last update: 1/14/2013 2:22:00 PM ~ Contributor(s): MarineBio. Marks, J.S. 1993. “Molt of Bristle-thighed Curlews in the northwestern Hawaiian Islands.” Auk 110:573-587. Marks, J.S., and R. Redmond. 1994. "Conservation problems and research needs for Bristle- thighed Curlews Numenious tahitiensis on their wintering grounds." Bird Conserv. Int. 4:329- 341. ---. 1996. “Demography of bristle-thighed curlews, Numensius tahitiensis, wintering on Laysan Island.” Ibis 138:438-447. Marks, J.S., R.L. Redmond, P. Hendricks, R.B. Clapp, and R.E. Jr. Gill. 1990. “Notes on longevity and flightlessness in Bristle-thighed Curlews.” Auk 107:779-781. Marks, J.S., R. Tibbitts, R.E. Gill Jr., and B.J. McCaffrey. 2002. Bristle-thighed curlew (Numensius tahitiensis). Cornell Lab of Ornithology [cited September 2017]. Available from http://bna.birds.cornell.edu/bna/species/705. Marshall, E.F. 1984. Cholecalciferol: A Unique Toxicant for Rodent Control. Vertebrate Pest Conference Proceedings collection. Marshall, A., M.B. Bennett, G. Kodja, S. Hinojosa-Alvarez, F. Galvan-Magana, M. Harding, G. Stevens, T. Kashiwagi. 2018. Mobula birostris (amended version of 2011 assessment). The IUCN Red List of Threatened Species 2018: e.T198921A126669349.

6-10 FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT REFERENCES

http://dx.doi.org/10.2305/IUCN.UK.2018-1.RLTS.T198921A126669349.en. Downloaded on 29 October 2018. Masuda, B.M., P. Fisher, and B. Beaven. 2015. "Residue profiles of brodifacoum in coastal marine species following an island rodent eradication." Ecotoxicol. Environ. Saf. 113:1-8. doi: 10.1016/j.ecoenv.2014.11.013. Mazurek, M. 2015. Lehua Island Inert Bait Study September 2015 Final Report. Honolulu, HI: Island Conservation. McCaffery, B., and G. Pelota Jr. 1986. “The status of the bristle-thighed curlew on the Yukon Delta National Wildlife Refuge, Alaska.” Wader Study Group Bulletin 47:22-25. McClelland, P. 2002. Eradication of Pacific rats (Rattus exulans) from Whenua Hou Nature Reserve (Codfish Island), Putauhinu and Rarotoka Islands, New Zealand. Turning the tide: the eradication of invasive species. IUCN SSC Invasive Species Specialist Group, at Gland, Switzerland and Cambridge, UK, IUCN. Pp 173-181. Megyesi, J.L., and C.R Griffin. 1996. “Brown Noddy Chick Predation by Great in the Northwestern Hawaiian Islands.” Condor 98 (2):322-327. doi: 10.2307/1369150. Merton, D. 1987. "Eradication of rabbits from Round Island, Mauritius, a conservation success story." . Journal of Jersey Wildlife Preservation Trust 24. Merton, D., G. Climo, V. Laboudallon, S. Robert, and C. Mander. 2002 of Conference. Alien mammal eradication and quarantine on inhabited islands in the Seychelles. Pgs. 182-198 in Turning the tide: The eradication of invasive species. Proceedings of the International Conference on Eradication of Island Invasives. Occ. Paper of the IUCN Species Survival Commission, No. 27, Report, at Gland, Switzerland and Cambridge, UK. Miller, M.H. and C. Klimovich. 2016. Endangered Species Act Status Review Report: Giant Manta Ray (Manta birostris) and Reef Manta Ray (Manta alfredi). Draft Report to National Marine Fisheries Service, Office of Protected Resources, Silver Spring, MD. December 2016. 127 pp. https://www.fisheries.noaa.gov/species/giant-manta-ray and https://www.cio.noaa.gov/services_programs/prplans/pdfs/ID353_FinalWorkProduct_MantaR ay.pdf. Mineau, P. 2014. A review and scientific opinion of the ecotoxicological data from Pinzon and Rabida Island lava lizards and the risk this represents to Galapagos hawks. Report prepared for Island Conservation. 16 pp. Mineau, P. 2018. A review of Midway Island’s mouse eradication mitigation plans. Report prepared for Island Conservation. 20 pp. Mitchell, C., C. Ogura, D. W. Meadows, A. Kane, L. Strommer, S. Fretz, D. Leonard, and A. McClung. 2005. Hawai‘i's comprehensive wildlife conservation strategy. Honolulu, HI: Department of Land and Natural Resources. 722 pp. Morrell, T.E., and S.M. Aquilani. 2000. “Nest-site characteristics of Red-tailed Tropicbirds on Rose Atoll, .” Journal of Field Ornithology 71 (3):455-459. Moulton, D.W., and A.P. Marshall. 1996. Laysan Duck (Anas laysanensis), version 2.0. In The Birds of North America, edited by P. G. Rodewald. Cornell Lab of Ornithology: Ithaca, New York, USA. https://doi.org/10.2173/bna.705.

6-11 MIDWAY SEABIRD PROTECTION PROJECT FINAL ENVIRONMENTAL ASSESSMENT REFERENCES

Moulton, D.W., and M.W. Weller. 1984. "Biology and conservation of the Laysan Duck (Anas laysanensis)." Condor 86:105-117. Musyl, M.K., R.W. Brill, D.S. Curran, N.M. Fragoso, L.M. McNaughton, A. Nielsen, B.S. Kikkawa, C.D. Moyes. 2011. Post-release survival, vertical and horizontal movements, and thermal habitats of five species of pelagic sharks in the central Pacific Ocean. Fishery Bulletin, 109, 341-368. Muto, M.M., V.T. Helker, R.P. Angliss, B.A. Allen, P.L. Boveng, J.M. Breiwick, M.F. Cameron, P.J. Clapham, S.P. Dahle, M.E. Dahlheim, B.S. Fadely, M.C. Ferguson, L.W. Fritz, R., Hobbs, Y.V. Ivashchenko, A.S. Kennedy, J.M. London, S.A. Mizroch, R.R. Ream, E.L. Richmond, K.E.W. Shelden, R.G. Towell, P.R. Wade, J.M. Waite, and A.N. Zerbini. 2017. Alaska Marine Mammal Stock Assessments, 2016. Naughton, M.B, M.D. Romano, and T.S. Zimmerman. 2007. A Conservation Action Plan for Black-footed Albatross (Phoebastria nigripes) and Laysan Albatross (P. immutabilis). http://www.iucnredlist.org/details/biblio/22698350/0. Navarrete, S.A., and J.C. Castilla. 1993. "Predation by rats in the intertidal zone of Central ." Mar. Ecol. Prog. Ser. 92:187-199. doi: 10.3354/meps092187. Neff, J.A., and P.A. DuMont. 1995. "A partial list of the plants of the Midway Islands." Atoll Research Bulletin. The Smithsonian Institution, Washington D.C. 45 (12):12. Nettleship, D.N. 2000. Ruddy Turnstone (Arenaria interpres), version 2.0. Cornell Lab of Ornithology. Available from https://doi.org/10.2173/bna.537. Niering, W.A. 1963. "Atoll Geography and Ecology." Ecology 44 (2). doi: 10.2307/1932200. Niethammer, K.R., and L.B. Patrick. 1998. White Tern (Gygis alba), version 2.0. Cornell Lab of Ornithology. Available from https://doi.org/10.2173/bna.371. Nishida, G.M., and J.W. Beardsley. 2002. “A Review of the Insects and Related Arthropods of Midway Atoll.” Bish. Mus. Occas. Pap. 68:25-69. NMFS. 1998. Recovery plan for the US Pacific populations of green turtles (Chelonia mydas). Edited by National Marine Fisheries Service. U.S. Fish and Wildlife Service: Silver Spring, MD. ---. 2006. Review of the status of the right whales in the North Atlantic and North Pacific Oceans. National Marine Fisheries Service, Office of Protected Resources, Silver Spring, MD. 68pp. ---. 2017. North Pacific Right Whale (Eubalaena japonica) Five-Year Review: Summary and Evaluation. National Marine Fisheries Service, Office of Protected Resources, Silver Spring, MD. 39 pp. http://www.nmfs.noaa.gov/pr/pdfs/species/northpacificrightwhale_5yearre view.pdf. NMFS-NOAA. 2007. Recovery Plan for the Hawaiian Monk Seal (Monachus schauinslandi). National Marine Fisheries Service, National Oceanic and Atmospheric Administration: Honolulu, HI. 165 pp.

6-12 FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT REFERENCES

---. 2014a. Blue Whale (Balaenoptera musculus musculus): Central North Pacific Stock. National Marine Fisheries Service, Silver Spring, MD. Revised June 4, 2014. http://www.nmfs.noaa.gov/pr/sars/2013/po2013_bluewhale-cnp.pdf. ---. 2014b. Fin Whale (Balaenoptera physalus physalus): Hawaii Stock. National Marine Fisheries Service, Silver Spring, MD. Revised June 4, 2014. http://www.nmfs.noaa.gov/pr/sars/2013/po2013_finwhale-hi.pdf. ---. 2015. Sperm Whale (Physeter macrocephalus) 5-Year Review: Summary and Evaluation. National Marine Fisheries Service, Silver Spring, MD. June 2015. http://www.fisheries.noaa.gov/pr/species/Status%20Reviews/sperm_whale_5year_review_fin al_june_2015.pdf. ---. 2018a. Final Rule to List the Giant Manta Ray as Threatened Under the Endangered Species Act. Rules and Regulations, January 22, 2018. Federal Register 83(14):2916-2931. ---. 2018b. Final Rule to List the Oceanic Whitetip Shark as Threatened Under the Endangered Species Act. Rules and Regulations, January 30, 2018. Federal Register 83(20):4153-4165. NMFS and USFWS. 1998a. Recovery plan for the US Pacific populations of green turtles (Chelonia mydas). National Marine Fisheries Service and U.S. Fish and Wildlife Service: Silver Spring, MD and Portland, OR. 84 pp. ---. 1998b. Recovery Plan for U.S. Pacific Populations of the Hawksbill Turtle (Eretmochelys imbricata). National Marine Fisheries Service: Silver Spring, MD. ---. 1998c. Recovery Plan for U.S. Pacific Populations of the Loggerhead Turtle (Caretta caretta). National Marine Fisheries Service, Silver Spring, MD. 71 pp. Accessed October 29, 2018. https://repository.library.noaa.gov/view/noaa/15967. ---. 1998d. Recovery Plan for U.S. Pacific Populations of the Olive Ridley Turtle (Lepidochelys olivacea). National Marine Fisheries Service, Silver Spring, MD. https://www.fpir.noaa.gov/Library/PRD/Sea%20Turtles/turtle_oliveridley_recoveryplan.pdf. ---. 1998e. Recovery Plan for U.S. Pacific Populations of the Leatherback Turtle (Dermochelys coriacea). National Marine Fisheries Service, Silver Spring, MD. https://www.fpir.noaa.gov/Library/PRD/Sea%20Turtles/turtle_leatherback_pacific_recoveryp lan.pdf. ---. 2013. Leatherback Sea Turtle (Dermochelys coriacea) 5-Year Review: Summary and Evaluation. National Marine Fisheries Service, Silver Spring, MD. November 2013. http://www.nmfs.noaa.gov/pr/listing/5yearreview_leatherbackturtle.pdf. ---.. 2014. Olive Ridley Sea Turtle (Lepidochelys olivacea) 5-Year Review: Summary and Evaluation. National Marine Fisheries Service, Silver Spring, MD. June 2014. http://www.nmfs.noaa.gov/pr/pdfs/species/oliveridleyturtle_5yearreview2014.pdf. NOAA CREP. 2008. Coral Reef Ecosystem Program. NOAA Fisheries, Pacific Islands Fisheries Science Center [cited October 2017]. Available from https://www.pifsc.noaa.gov/cred/. NOAA Fisheries Service. 2011a. Green Turtle (Chelonia mydas). Office of Protected Fisheries: Silver Spring, MD.

6-13 MIDWAY SEABIRD PROTECTION PROJECT FINAL ENVIRONMENTAL ASSESSMENT REFERENCES

---. 2011b. Hawksbill Turtle (Eretmochelys imbricata). Office of Protected Fisheries: Silver Spring, MD. ---. 2018a. “Leatherback Turtle.” Species Overview. Accessed October 29, 2018. https://www.fisheries.noaa.gov/species/leatherback-turtle. ---. 2018b. “North Pacific Right Whale.” Species Overview. Accessed October 30, 2018. https://www.fisheries.noaa.gov/species/north-pacific-right-whale. ---. 2018c. “Oceanic Whitetip Shark.” Species Overview. Accessed October 30, 2018. https://www.fisheries.noaa.gov/species/oceanic-whitetip-shark. NOAA-NMFS. 1986. Final Rule on Designated Critical Habitat; Hawaiian Monk Seal. Wednesday, April 30, 1986. Rules and Regulations. Federal Register 51(83): 16047-16053. http://cdn.loc.gov/service/ll/fedreg/fr051/fr051083/fr051083.pdf#page=53. ---. 2015. Endangered and Threatened Species: Final Rulemaking to Revise Critical Habitat for Hawaiian Monk Seals. Federal Register 80 (162), August 21, 2015. Rules and Regulations. 63 pp. https://www.federalregister.gov/documents/2015/08/21/2015-20617/endangered-and- threatened-species-final-rulemaking-to-revise-critical-habitat-for-hawaiian-monk. Ogden. 1994. Report on lead-based paint (LBP) and asbestos survey of all structures on Eastern and Sand Islands. Ogden Environmental and Energy Service. ---. 1996. NAF Midway Island Abbreviated RI Report (Draft) for the Active and Abandoned Housing Areas. Ogden Environmental and Energy Service. ---. 1997. Base Realignment and Closure (BRAC) Cleanup Plan: Environmental status of NAF Midway Atoll as of 1996. Ogden Environmental and Energy Service. Oppel, S., J.L. Lavers, A.L. Bond, and G. Harrison. 2016. "Reducing the primary exposure risk of Henderson crakes (Zapornia atra) during aerial broadcast eradication by selecting appropriate bait colour." Wildlife Research. Orazio, C.E., M.J. Tanner, C. Swenson, J. Herod, P. Dunlevy, and Gale G.W. 2009. Results of Laboratory Testing for Diphacinone in Seawater, Fish, Invertebrates, and Soil Following Aerial Application of Rodenticide on Lehua Island, Kauai County, Hawai'i. U.S. Geological Survey Open-File Report 2009-1142. Pain, D.J., D.de L. Brooke, J.K. Finnie, and A. Jackson. 2000. "Effects of brodifacoum on the land of Ascension Island." Journal of Wildlife Management 64 (2). Parkes, J., P. Fisher, and G. Forrester. 2011. "Diagnosing the cause of failure to eradicate introduced rodents on islands: brodifacoum versus diphacinone and method of bait delivery." Cons. Evidence 8:100-106. Parkes, J. 2014. “Eradication of house mice Mus musculus from Marion Island: a review of feasibility, constraints and risks.” In BirdLife South Africa Ocassional Report Series No. 1. Edited by R.M. Wanles, pp 27. Birdlife South Africa: Johannesburg, South Africa. Parsons, J. 1962. The Green Turtle and Man. University of Florida Press: Gainesville, FL. PEPP. 2013. "Plant Extinction Prevention Program annual report fiscal year 2013 (July 1, 2012- June 30, 2013)."207 pp.

6-14 FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT REFERENCES

Pierce, R., N. Anterea, U. Anterea, K. Broome, D. Brown, L. Cooper, H. Edmonds, F. Muckle, B. Nagle, G. Oakes, M. Thorsen, and G. Wragg. 2008. Operational work undertaken to eradicate rats and rabbits in the Phoenix Islands, Republic of Kiribati, May-June 2008. Pacific Expeditions Ltd. Pitman, R.L. 1992. Sea turtle associations with flotsam in the eastern tropical Pacific. Page 94 in M. Salmon and J. Wyneken (compilers), Proc. of the Eleventh Annual Workshop on Sea Turtle Biology and Conservation. U.S. Dep. of Comm., NOAA Tech. Memo. NMFS-SEFSC- 302. 195 pp. Pitman, R.L. 1993. Seabird associations with marine turtles in the eastern Pacific Ocean. Colonial Waterbirds 16:194-201. Pitt, W. C., A. R. Berentsen, S. F. Volker, and J. D. Eisemann. 2012. Palmyra Atoll Rainforest Restoration Project: Monitoring Results for the Application of Broadcast of Brodifacoum 25W: Conservation to Eradicate Rats. QA-1875 Final Report. USDA, APHIS, W, NWRC. Hilo, HI. 85 pp. Pitt, W. C., A.R. Berentsen, A. B. Shiels, S. F. Volker, J.D. Eisemann, A.S. Wegmann, and G.R. Howald. 2015. “Non-Target Species Mortality and the Measurement of Brodifacoum Rodenticide Residues after a Rat (Rattus rattus) Eradication on Palmyra Atoll, Tropical Pacific.” Biological Conservation 185:36-46. Pitt, W., A. Wegmann, J. Eisemann, and G. Howald. 2011. Palmyra Atoll Rainforest Restoration Project: rat eradication monitoring plan for all aerial broadcast alternatives. Island Conservation: Santa Cruz, CA. Polovina, J. J., I. Uchida, G. Balazs, E. A. Howell, D. Parker and P. Dutton. 2006. The Kuroshio extension bifurcation region: A pelagic hotspot for juvenile loggerhead sea turtles. Deep-Sea Research Part II 53:326–339. Pratt, H.D., P.L. Bruner, and D.G. Berrett. 1987. A Field Guide to the Birds of Hawai'i and the Tropical Pacific. Princeton University Press: Princeton, NJ. Primus, T.M., G. Wright, and P. Fisher. 2005. "Accidental discharge of brodifacoum baits in a tidal marine environment: a case study." Bull. Environ. Contam. Toxicol. 74:913-919. Pritchard, P.C.H. 1982. Sea turtles of Micronesia. Pages 263-274 in K.A. Bjorndal (ed.), Biology and Conservation of Sea Turtles. Smithsonian Inst. Press, Washington, D.C. 583 pp. Pye, T., R. Swain, and R. Seppelt. 1999. “Distribution and habitat use of the feral black rat (Rattus rattus) on subantarctic Macquarie Island.” Journal of Zoology 247:429-438. Pyle, R.L, and P. Pyle. 2017. The Birds of the Hawaiian Islands: Occurrence, History, Distribution, and Status (Version 2 (1 January 2017)). Honolulu, HI, U.S.A. Available from http://hbs.bishopmuseum.org/birds/rlp-monograph/. (downloaded 14 November 2017). Rehkemper, C. 2006, in litt. "Email regarding current status of Solanum nelsonii in the Northwest Hawaiian Islands."May 30, 2006. Reilly S.B., Bannister J.L., Best P.B., Brown M., Brownell Jr. R.L., Butterworth D.S., Clapham P.J., Cooke J., Donovan G.P., Urbán J. and Zerbini A.N. 2008. Eubalaena japonica. The IUCN Red List of Threatened Species. Version 2014.3. . Accessed October 29, 2018.

6-15 MIDWAY SEABIRD PROTECTION PROJECT FINAL ENVIRONMENTAL ASSESSMENT REFERENCES

Reynolds, M.H. and J. Citta. 2007. Post-fledgling survival of Laysan Ducks. Journal of Wildlife Management. 71: 383–388. Reynolds, M. and J. Klavitter. 2006. “Translocation of wild Laysan duck Anas laysanensis to establish a population at Midway Atoll National Wildlife Refuge, United States and US Pacific Possession.” Conservation Evidence 3:6-8. Reynolds, M., and K. Kozar. 2000. Laysan duck (Anan laysanensis) population status and translocation feasibility report. USFWS unpublished report: Honolulu, HI. 88 pp. Reynolds, M., K. Courtot, and J. Hatfield. In press. How many Laysan Teal (Anas laysanensis) are on Midway Atoll: Methods for monitoring abundance after reintroduction. U.S. Geologic Survey. Wildfowl & Wetlands Trust. Pacific Island Ecosystems Research Center. 60-71 pp. Reynolds, M.H., N.E. Seavy, M.S. Vekasy, J.L. Klavitter, and L.P. Laniawe. 2008. Translocation and early post-release demography of the endangered Laysan teal. Animal Conservation 11:160-168. Reynolds, M.H., C. Vanderlip, J. Klavitter, and K. Courtot. 2015 Translocation of wild Laysan teal to Kure Atoll: species conservation overview and update. In 23rd Annual Hawai‘i Conservation Conference. August 5, 2015., University of Hawai‘i at Hilo. Hilo, Hawai‘i. Reynolds, M., P. Berkowitz, K.N. Coutrot, and C.M. Krause. 2012. Predicting sea-level rise vulnerbility of terrestrial habitat and wildlife of the Northwestern Hawaiian Islands. U.S. Geological Survey Open-File Report 2012-1182, 139 pp. [cited Mar 2018]. Available from http://pubs.usgs.gov/of/2012/1182/. U.S. Geological Survey, Pacific Island Ecosystems Research Center, Kīlauea Field Station: Hawaiʻi National Park, Hawaiʻi 96718, USA. Reynolds, M., P. Berkowitz, J.L. Klavitter, and K.N. Courtot. 2017. “Lessons from the Tōhoku tsunami: A model for island avifauna conservation prioritization.” Wiley Ecology and 7:5873–5890. doi: 10.1002/ece3.3092. Reynolds, M. H.; Crampton, L. H.; Vekasy, M. S. 2007. Laysan Teal Anas laysanensis nesting phenology and site characteristics on Laysan Island. Wildfowl 57: 54-67 Reynolds, M.H., M.S. Veksay, and J. Klavitter. 2004. Laysan Teal Translocation Plan for Midway Atoll NWR. U.S. Geological Survey Report, Pacific Island Ecosystem Research Center, Kilauea Field Station; Hawai`i National Park, HI. Ross, D.B., N.L. Roberts, and C. Fairley. 1980. Brodifacoum: The acute oral toxicity (LD50) of Brodifacoum to the Mallard duck. Huntingdon Research Centre, Huntingdon, Cambridgeshire, UK. Submitted by ICI Americas, Inc. Agricultural Products, Wilmington, Delaware 19897. ICI’s Study No. ICI308WL/791275. EPA MRID 415633-03. Rowan, R. 1998. “Diversity and ecology of zooxanthellae on coral reefs.” Journal of Phychology 34:407-417. Rowntree, V., J. Darling, G. Silber, and M. Ferrari. 1980. Rare sightings of a right whale (Eubalaena glacialis) in Hawaii. Can. J. Zool. 58:308-312. Rueda, D.K., K.J. Campbell, P. Fisher, F. Cunninghame, and J.B. Ponder. 2016. "Biologically significant residual persistence of brodifacoum in reptiles following invasive rodent eradication, Galapagos Islands, ." Conservation Evidence 13:38.

6-16 FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT REFERENCES

Russell, J. C., and N.D. Holmes. 2015. “Tropical island conservation: rat eradication for species recovery.” Biological Conservation 185:1-7. Salden, D.R. and J. Mickelsen. 1999. Rare Sighting of a North Pacific Right Whale (Eubalaena glacialis) in Hawaii. Pacific Science 53(4):341-345. https://scholarspace.manoa.hawaii.edu/bitstream/10125/701/v53n4-341-345.pdf. Samaniego, A. 2016. Mouse eradication on Midway Atoll: a review of the record. In Pacific Invasives Initiative (PII). Samaniego-Herrera, A., J. Russell, D. Choquenot, A. Aguirre-Munoz, and M. Clout. 2014. Invasive Rodents on Tropical Islands: Eradication Recommendations from Mexico. 26th Vertebrate Pest Conference, at Davis University, CA. Samaniego-Herrera, A. 2016 Mouse eradication on Midway Atoll: a review of the record. In Pacific Invasives Initiative (PII). Schreiber, E.A. and R.W. Schreiber. 1993. Red-tailed tropicbird (Phaethon rubricauda). The Birds of North America, Number 43. The American Ornithologists' Union, Washington, D.C., USA, and The Academy of Natural Sciences, Philadelphia, Pennsylvania, USA. Schreiber, B.A., and R.W. Schreiber. 2009. Red-tailed Tropicbird (Phaethon rubricauda), version 2.0. Cornell Lab of Ornithology. Available from https://doi.org/10.2173/bna.43. Schreiber, E.A., P.F. Doherty, Jr., and G.A. Schenk. 2001. Effects of a Chemical Weapons Incineration Plant on Red-Tailed Tropicbirds. The Journal of Wildlife Management, (65)4: 685-695. Schreiber, E.A., and R.L. Norton. 2002. Brown Booby (Sula leucogaster), version 2.0. Cornell Lab of Ornithology. Available from https://doi.org/10.2173/bna.649. Schuyler, Q., B.D. Hardesty, C. Wilcox, and K. Townsend. 2014. "Global Analysis of Anthropogenic Debris Ingestion by Sea Turtles." Conserv Biol 28 (1):129-139. doi: 10.1111/cobi.12126. Seavy, N.E., M.H. Reynolds, W.A. Link, and J.S. Hatfield. 2009. Post-catastrophe population dynamics and density dependence of an endemic island duck. J Wildl Manage 73: 414−418. Sekiguchi, K., H. Onishi, H. Sasaki, S. Haba, Y. Iwahara, D. Mizuguchi, M. Otsuki, D. Saijo, B. Nishizawa, H. Mizuno, N. Hoshi and T. Kamito. 2014. Sightings of the western stock of North Pacific right whales (Eubalaena japonica) in the far southeast of the Kamchatka Peninsula. Marine Mammal Science DOI: 10.1111/mms.12105. Seto, N. 1994. The effects of rat (Rattus rattus) predation on the reproductive success of the Bonin Petrel (Pterodroma hypoleuca), Master's Thesis, University of Hawai‘i, Honolulu, HI. Seto, N.W., and D.L. O'Daniel. 1999. Bonin Petrel (Pterodroma hypoleuca), version 2.0. Shiels, A.B., C.A. Flores, A. Khamsing, P.D. Krushelnycky, S.M. Mosher, and D.R. Drake. 2013. “Dietary niche differentiation among three species of invasive rodents (Rattus rattus, R. exulans, Mus musculus).” Biolological Invasions 15:1037-1048. Shiels, A.B., G.W. Witmer, C. Samra, R.S. Moulton, E.W. Ruell, J.R. O’Hare, J.D. Eisemann, S.F. Volker, and D.A. Goldade. 2017. Assessment of bait density, bait availability, and non- target impacts during an aerial application of rodenticide to eliminate invasive rats on

6-17 MIDWAY SEABIRD PROTECTION PROJECT FINAL ENVIRONMENTAL ASSESSMENT REFERENCES

Desecheo Island, Puerto Rico. Final Report QA 2588. USDA, APHIS, WS, NWRC, Ft Collins, CO. 87 pp. Smultea, M.A., T.A. Jefferson, A.M. Zoidi. 2010. Rare Sightings of a Bryde’s Whale (Balaenoptera edeni) and Sei Whales (B. borealis) (Cetacea: Balaenopteridae) Northeast of O‘ahu, Hawai‘i. Pacific Science. 64(3):449–457. Snow, D.W., and C.M. Perrins. 1998. The Birds of the Western Palearctic Concise Edition: Volume 1 Non-Passerines, Volume 2 Passerines Oxford University Press, Oxford. Starr, F., and K. Martz. 1999. Botanical survey of Midway Atoll 1999 update. In U.S. Fish and Wildlife Service Report. Honolulu, HI. Starr, F., and K. Starr. 2008. Botanical Survey of Midway Atoll. In Report prepared for the United States Fish and Wildlife Service. Honolulu, HI. ---. 2015. Botanical survey of Midway Atoll. Report prepared for the U.S. Fish and Wildlife Service. Honolulu, HI. Strasburg, D. 1958. Distribution, abundance, and habits of pelagic sharks in the Central Pacific Ocean. Fishery Bulletin 138 Washington, U.S. Govt. Print. Off., 58, 335-361. SWCA. 2017. Laysan Teal and Bristle-Thighed Curlew Assessment Report for the Mouse Eradication Project, Midway Atoll National Wildlife Refuge. Prepared for Island Conservation. Prepared by SWCA Environmental Consultants September. 43 pp. Symon, D.E. 1999. "Solanaceae." In Manual of the Flowering Plants of Hawaii, edited by W.L. Wagner, D.R. Herbst, and S.H. Sohmer, Pages 1273-1274. University of Hawaii Press and Bishop Museum Press: Honolulu. Bishop Museum Special Publications. Tangalin, N. 2006, in litt. "Email regarding status of Solanum nelsonii on Nihoa." May 25, 2006.The Ornithological Council. 2010. The Rat Island Rat Eradication Project: A Critical Evaluation of Non-target Mortality. Prepared for Island Conservation, The Nature Conservancy, U.S. Fish and Wildlife Service, Alaska Maritime National Wildlife Refuge. Reviewed by T.P. Salmon, Sheffield, S.R. and Paul, E. 85 pp. Thibault, J.C. 1995. “Effect of predation by the black rat Rattus rattus on the breeding success of Cory's Shearwater Calonectris diomedea in Corsica.” Marine Ornithology 23:1-10. Thomas, P.E.J. (compiler). 1989. Report of the Northern Marshall Islands Natural Diversity and Protected Areas Survey, 7-24 September 1988. South Pacific Regional Environment Programme, Noumea, New Caledonia and East-West Center, Honolulu, Hawaii. 133 pp. Thomas, P.J., P. Mineau, R.F. Shore, L. Champoux, P. Martin, L. Wilson, G. Fitzgerald, J. Elliott. 2011. Second generation anticoagulant rodenticides in predatory birds: probabilistic characterization of toxic liver concentrations and implications for predatory bird populations in Canada, Environment International 37(2011): 914-920. (Erratum: 40(2012): 256). TNCH. 2007. "Geographic information system (GIS) shapefiles showing ecosystem regions for the main Hawaiian Islands, as developed by The Nature Conservancy Hawai‘i’s ecoregional planning team." Unpublished data.

6-18 FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT REFERENCES

Tobin, M., and M. Fall. 2005. Pest control: Rodents Agricultural Sciences. In Encyclopedia of Life Support Systems (EOLSS). Oxford, UK: EOLSS Publishers under the auspices of the United Nations Educational, Scientific, and Cultural Organization. Tomich, P. 1986. Mammals in Hawaiʻi. 2nd edition. Bishop Museum Press: Honolulu, HI. TOPP. 2006. Results of satellite tagging of leatherback turtles. Tagging of Pacific Pelagics. U.S. Fish and Wildlife Service. 2010. Midway Atoll National Wildlife Refuge Historic Preservation Plan 2010. Prepared by L.A. Speulda-Drews. December 2010. U.S. Fish and Wildlife Service, Region 1, Honolulu, HI. http://www.fws.gov/uploadedFiles/Region_1/NWRS/Zone_1/Midway_Atoll/Documents/Mid way%20Atoll%20HPP%202010.pdf. Towns, D. 1991. “Response of assemblages in the Mercury Islands, New Zealand, to removal of an introduced rodent, the kiore (Rattus exulans).” Journal of The Royal Society of New Zealand 21:119-136. Tuato'o-Bartley, N., T.E. Morrell, and P. Craig. 1993. Status of sea turtles in American Samoa in 1991. Pacific Science 47(3):215-221. UK University of Hertfordshire Pesticide Properties Database, Available from http://sitem.herts.ac.uk/aeru/ppdb/en/Reports/87.htm. University of Hawaii. 2009. Native Plants Hawai'i: Pritchardia remota, Dec 2009 [cited Mar 13 2018]. Available from http://nativeplants.hawaii.edu/plant/view/Pritchardia_hillebrandii. U.S. Army Space and Missile Defense Command. 2003. Ground-Based Midcourse Defense (GMD) Extended Test Range (ETR): Environmental Impact Statement. Volume 1. Huntsville, AL. USEPA. 1991. Reregistration Eligibility Decision (RED) Rodenticide Cluster. U.S. Environmental Protection Agency, Pesticides Programs: Washington, D.C. USEPA. 1995. “Great Lakes Water Quality Initiative. Methodology for Development of Wildlife Criteria.” 40 CFR Ch. App D. US EPA. U.S. National Park Service. 2000. Anacapa Projects. Final Environmental Impact Statement. Ventura, CA. 139 pp. USFWS. 1982. The Laysan Duck recovery plan (B. Giezentanner et al, eds.). Portland, OR: USFWS. ---. 1983. Atlas of Hawaiian Seabird Colonies. In Catalog of Hawaiian Seabird Colonies. Edited by S.I. Fefer, et al. Refuges and Wildlife, Pacific Islands Office: Honolulu, HI. 236 pp. ---. 1996. "Endangered and threatened wildlife and plants; determination of endangered or threatened status for fourteen plant taxa from the Hawaiian Islands; final rule." Federal Register 61 (198):53108-53124. ---. 1998. Final Recovery Plan for Three Plant Species on Nihoa Island. U.S. Fish and Wildlife Service: Portland, OR. 83 pp. ---. 2004. Draft Revised Recovery Plan for the Laysan Duck (Anas laysanensis). USFWS Region 1. Portland, OR, 94 pp.

6-19 MIDWAY SEABIRD PROTECTION PROJECT FINAL ENVIRONMENTAL ASSESSMENT REFERENCES

---. 2005. Regional Seabird Conservation Plan, Pacific Region. U.S. Fish and Wildlife Service, Migratory Bird and Habitat Programs, Pacific Region. Portland, Oregon. ---. 2008a. Papahānaumokuākea Monument Management Plan. Midway Atoll National Wildlife Refuge and Battle of Midway National Memorial, Jan 27 2015 [cited Mar 21 2015]. Available from https://www.fws.gov/refuge/Midway_Atoll/what_we_do/planning.html. ---. 2008b. Birds of conservation concern 2008. Edited by U.S. Fish and Wildlife Service. Division of Migratory Bird Management: Arlington, VA. ---. 2008c. Short-tailed Albatross Recovery Plan. Anchorage, AK. ---. 2009a. Revised Recovery Plan for the Laysan Duck (Anas laysanensis). U.S. Fish and Wildlife Service, Portland, Oregon. ix + 114 pp. ---. 2009b. Cenchrus agrimonioides (Kāmanomano): 5-Year Review Summary and Evaluation. Pacific Islands Fish and Wildlife Office Honolulu, Hawai‘i. 14 pp. ---. 2011. Final Environmental Impact Statement: Palmyra Atoll National Wildlife Refuge Rat Eradication Project. Pacific Region: Portland, OR. 659 pp. ---. 2014. “Make Way for Ducks.” U.S. Fish and Wildlife Service, National Wildlife Refuge System. News release on the translocation of Laysan ducks from Midway Atoll National Wildlife Refuge to Kure Atoll Wildlife Sanctuary, September 12 [cited Oct 3 2018]. Available from https://www.fws.gov/refuges/news/MakeWayForDucks.html. ---. 2015. Fact Sheet: Wedge-tailed Shearwater (Puffinus pacificus) [cited 10-5-17]. ---. 2016a. "Endangered and Threatened Wildlife and Plants; Endangered Status for 49 Species from the Hawaiian Islands." Federal Register 81 (190):67786-67859. ---. 2016b. Fact Sheet: Red-tailed Tropicbird (Phaethon rubricauda rothschildi). ---. 2016c. "Final rule to list eleven distinct population segments of the green sea turtle (Chelonia mydas)." Federal Register 81 (66):20058-20090. ---. 2017. Biological assessment for the seawall long-term maintenance project. Henderson Field Airport, Midway Atoll National Wildlife Refuge/Battle of Midway National Memorial and the Federal Aviation Administration. Honolulu, Hawaii. 32 pp. ---. 2018. Midway Seabird Protection Project. Draft Environmental Assessment. Sand Island, Midway Atoll, Papahānaumokuākea Marine National Monument. Prepared by Hamer Environmental L.P. and Planning Solutions, Inc. March 2018. 198 pp. ---. Unpublished a. "Midway Atoll National Wildlife Refuge 2002-2006." ---. Unpublished b. “Regarding seabirds and shorebirds of Midway Atoll National Wildlife Refuge.” USFWS and HDLNR. 2017. Final Environmental Assessment Lehua Island Ecosystem Restoration Project. Co-Lead Agencies: U.S. Fish and Wildlife Service and Hawaiʻi Department of Land and Natural Resources, Division of Forestry and Wildlife. Van Gils, J., and P. Wiersma. 1996. Family Scolopacidae (sandpipers, snipes and phalaropes). Edited by A. Elliott and J. Sargatal J. del Hoyo, Handbook of the birds of the world. Barcelona: Lynx Ediciones.

6-20 FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT REFERENCES

Vanderlip, C. 2011, in litt. "Tsunami changes, Kure Atoll Conservancy."Mar 11 2011. VanderWerf, E.A. 2012. Hawaiian Bird Conservation Action Plan. Pacific Rim Conservation, Honolulu, HI. 7 pp. Veitch, C., and M. Clout. 2002. Turning the Tide: the Eradication of Invasive Species. SSC Invasive Species Specialist Group: Gland, Switzerland and Cambridge, UK. Veitch, C.R., and M.N. Clout. 2011. “Island invasives: eradication and management.” In International Union for the Conservation of Nature, edited by D.R. Towns. Gland, Switzerland. Vicente, V.P. 1994 of Conference. Spongivory in Caribbean hawksbill turtles, Eretmochelys imbricata: Data from stranded specimens. Pgs. 185-188 in Thirteenth Annual Symposium on Sea Turtle Biology and Conservation, Report, at NOAA Tech Memo. NMFS-SEFSC-341, Jekyll Island, GA. WHO. 1995. Brodifacoum Health and Safety Guide No. 93, Section 2.3 Environmental Transport, Distribution, and Transformation. International Programme on Chemical Safety, World Health Organization. Geneva, Switzerland. [Downloaded 16 November 2017]. Available at http://www.inchem.org/documents/hsg/hsg/hsg093.htm. Walters, J.R., and M.H. Reynolds. 2013. "Experimental reintroduction reveals novel life-history variation in Laysan Ducks (Anas laysanensis)." The Auk 130 (4):573-579. Wanless, R.M, A. Angel, R.J. Cuthbert, G.M. Hilton, and P.G. Ryan. 2007. “Can predation by invasive mice drive seabird extinctions?” Biology Letters 3 (3):241-244. doi: 10.1098/rsbl.2007.0120. Weatherbase.com. 2017. Average temperature and rainfall at Henderson Field Airport, Sand Island, Midway Atoll, NW Hawaiian Islands. Available from https://www.weatherbase.com. Wegmann, A.S. 2009. Limitations to tree seedling recruitment at Palmyra Atoll. University of Hawai‘i: Honolulu, HI. Weir, S.M., S. Yu, A. Knox, L.G. Talent, J.M. Monk, and C.J. Salice. 2016. "Acute toxicity and risk to lizards of rodenticides and herbicides commonly used in New Zealand." New Zealand Journal of Ecology 40 (3):342-350. doi: 10.20417/nzjecol.40.43. Whitaker, A.H. 1973. “Lizard populations on islands with and without Polynesian rats Rattus exulans (Peale).” Proceedings of the New Zealand Ecological Society 20:121-130. Whittow, G.C. 1993a. “Black-footed Albatross (Diomedea nigripes).” In The Birds of North America, edited by A. Poole and F. Gill. Washington, D.C: The American Ornithologists' Union: Philadelphia: The Academy of Natural Sciences. ---. 1993b. “Laysan Albatross (Phoebastria immutabilis).” In The Birds of North America, edited by A. Poole and F. Gill. Washington, D.C.: The American Ornithologists' Union: Philadelphia: The Academy of Natural Sciences. Whittow, G.C., and G.H. Balazs. 1982. "Basking behavior of the Hawaiian green turtle (Chelonia mydas)." Pac. Sci. 36:129-139. Woodward, P.W. 1972. "The natural history of Kure Atoll, Northwestern Hawaiian Islands." Atoll Res. Bull. 164:1-318.

6-21 MIDWAY SEABIRD PROTECTION PROJECT FINAL ENVIRONMENTAL ASSESSMENT REFERENCES

Work, T.M., and M.R. Smith. 1996. “Lead exposure in Laysan albatross adults and chicks in Hawaii: prevalence, risk factors, and biochemical effects.” Archives of Environmental contamination and Toxicology 31:115-119. Work, T. M., Klavitter, J. L., Reynolds, M. H. & Blehert, D. 2010. “Avian botulism: a case study in translocated endangered Laysan Ducks (Anas laysanensis) on midway atoll.” J. Wildl. Dis. 46: 499–506. Young, H., D. McCauley, R. Dunbar, and R. Dirzo. 2010. “Plants cause ecosystem nutrient depletion via the interruption of bird derived spatial subsidies.” Proceedings of the National Academy of Science. 107:2072-2077. Young, C.N., J. Carlson, M. Hutchinson, C. Hutt, D. Kobayashi, C.T. McCandless, J. Wraith. 2016. Status review report: oceanic whitetip shark (Carcharhinius longimanus). Final Report to the National Marine Fisheries Service, Office of Protected Resources. November 2016. 162 pp.

6-22 FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT CONSULTATION & DISTRIBUTION

CONSULTATION & DISTRIBUTION

7.1 PRE-DRAFT SCOPING PROCESS The proposed mouse eradication action involves the aerial broadcast of bait pellets containing rodenticide into all potential mouse territories on Sand Island, MANWR. Mouse eradication would occur in the summer dry season (July to August) to maximize the probability of success, targeting the mice when their food resources are lowest and their abundance is declining. Conducting the operation during this period would also minimize the risk of exposure to both seabirds and shorebirds, as well as the risk of rain washing rodenticide pellets into the ocean. A Public Scoping effort by the USFWS was conducted to solicit input from the public, state and Federal regulatory agencies, and non-governmental organizations regarding the proposed Midway Seabird Protection Project on Sand Island, MANWR, and the Northwestern Hawaiian Islands. Prior to the preparation of the Draft Environmental Assessment for the proposed action, on August 7, 2017, a scoping letter describing the proposed action and interview questions were emailed to the individuals listed in Table 7.1:

Table 7.1 Scoping Letter Recipients No. Recipient Agency or Role 1 Karen Vitulano Environmental Protection Agency 2 Thierry Work United States Geological Service, National Wildlife Health Center 3 Richard Hall National Oceanic and Atmospheric Administration, National Marine Fisheries Service 4 Thea Johanos-Kam National Oceanic and Atmospheric Administration, National Marine Fisheries Service 5 Gabrielle Fenix Grange Hawaii Department of Health, Hazard Evaluation and Emergency Response Office 6 Maggie Sergio Island Watch Conservation Science 7 Sydney Ross Singer Good Shepherd Foundation 8 Joshua Atwood Hawai‘i Department of Land and Natural Resources, Division of Forestry and Wildlife 9 Mark Rauzon Pacific Seabird Group/Laney College 10 Richard and Jenny Johnson former Midway resident and Midway volunteer 11 Jill McIntire Midway volunteer Source: Hamer Environmental (2017)

These individuals, organizations, and agencies were contacted because they had previously expressed interest in similar island eradication projects or may have oversight or regulatory concerns about the project. All eleven responded to the interview questions either via email or by phone.

7-1 MIDWAY SEABIRD PROTECTION PROJECT FINAL ENVIRONMENTAL ASSESSMENT CONSULTATION & DISTRIBUTION

7.2 INTERVIEW PROCESS A series of interview questions were prepared by Hamer Environmental, Inc., and Planning Solutions. Scoping participants were informed of the proposed project’s purpose, need, method, schedule, background and alternatives. The 16 questions were categorized into 4 parts as shown in Table 7.2:

Table 7.2 Summary of Scoping Interview Part I: Contact Information No. Question 1 Please state your full name, preferred contact number, and email. 2 What is your occupation and/or affiliation? 3 How many people are in your organization? 4 How many days, week, months, or years have you spent on Midway? 5 When was your most recent trip to Midway? 6 What do you consider your particular areas of expertise related to the project purpose or location? Part II: Survey 7 Do you have input on the purpose and need? 8 Do you believe that conservation of the native environment (seabirds, wildlife, and botanical resources) adequately justify the eradication of invasive mice? Part III: Historical and Cultural Information 9 Is there anything you would like to say about the general history of the area, or past and present land use? Questions only for people who have visited or lived on Midway: 10 Have you participated in or observed any cultural events or practices on Midway? 11 Do you know anything about the canaries and their potential historic or cultural significance? Part IV: General Considerations 12 Have you encountered or are you aware of any facilities, areas, or resources on Midway that you believe require special consideration, either in terms of project implementation or potential project impacts? 13 What concerns would you have regarding the use of following rodenticide distribution methods on Midway: (a) aerial broadcast; (b) band broadcast; (c) bait stations? 14 Are you aware of any human activities on Midway that may be impacted by or affect the efficacy of an eradication effort? 15 What factors, such as structures, topography, or human habits, do you feel represent the biggest challenges to the eradication effort? How would you approach them? 16 What impacts do you foresee (water or marine environment or others) from the action and how would you (a) avoid, (b) minimize, or (c) mitigation for them? Source: Hamer Environmental (2017)

The USFWS compiled responses from the public scoping process and prepared a list of environmental issues that warranted specific consideration in the EA analysis. The concerns that were identified are: (i) damage to seabird populations; (ii) secondary poisoning; (iii) poisoning of rodents and non-target species; (iv) disturbance; (v) cultural issues; (vi) project effectiveness; and (vii) the EA process and Alternatives.

7-2 FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT STATEMENT OF COMPLIANCE

STATEMENT OF COMPLIANCE FOR IMPLEMENTATION OF THE PROPOSED ACTION AT MANWR

8.1 MIDWAY SEABIRD PROTECTION PROJECT NEPA AND COMPLIANCE PROCESS The USFWS is the lead agency on the Seabird Protection Project at MANWR. The USDA APHIS Wildlife Services National Research Center is a Cooperating Agency for the project. The following Regulatory and Policy Compliance Requirements and Permit processes are either completed or currently underway. All required parmits and approvals would be completed prior to implementing the proposed action. All compliance documents and permits can be found at: https://www.fws.gov/refuge/midway_atoll/

8.2 NATIONAL ENVIRONMENTAL POLICY ACT (1969) A Final Environmental Assessment has been completed in accordance with NEPA Implementing Procedures, Department of the Interior and Service procedures and has been performed in coordination with the affected public. The USFWS released the Midway Seabird Protection Project Draft Environmental Assessment for a 30-day public review in March 2018. Public comments received and USFWS responses are located in Appendix F of this Final EA. Minor updates were also made to the Environmental Assessment following public comment and compliance processes.

8.3 ENDANGERED SPECIES ACT OF 1973 The Service completed two consultations related to Section 7 of the Endangered Species Act for all listed species at MANWR. Informal consultation was completed with NOAA/National Marine Fishery Service in November 2019 and a Formal Consultation was completed in January 2019 when the Service received a Biological Opinion (01EPIF00-2019-F-0049) from the Pacific Islands Fish and Wildlife Office. A Biological Assessment was prepared for both of the consultation processes. The Biological Assessment and the Biological Opinion are available for public review along with all consultation letters on the Midway Atoll website.

8.4 MAGNUSON-STEVENS ACT ESSENTIAL FISH HABITAT The Service completed consultation with NOAA/National Marine Fishery Service for effects to Essential Fish Habitat (EFH) from the Proposed Action in September 2018. The Draft EA served as the technical document to satisfy EFH provisions.

8.5 PAPAHĀNAUMOKUĀKEA MARINE NATIONAL MONUMENT (PMNM) PERMIT In accordance with Presidential Proclamation 8031 and codifying regulations in 50 CFR Part 404, all activities in the Monument, with limited exceptions, require a permit. The National Marine Monument permit would be obtained prior to implementation of the proposed action at Midway.

8-1 MIDWAY SEABIRD PROTECTION PROJECT FINAL ENVIRONMENTAL ASSESSMENT STATEMENT OF COMPLIANCE 8.6 NATIONAL HISTORIC PRESERVATION ACT, SECTION 106 (1966) The USFWS completed a Section 106 review in May 2018 based on information contained in the Draft Environmental Assessment. The Review was completed with receipt of a letter from the USFWS Historic Preservation Officer, with the finding, that: “while the mouse eradication program is an NHPA undertaking, it is not of the type that has potential to cause effects to historic properties (36CFR800.3.a.1). Therefore, the USFWS has no further obligation to consider or consult on historic properties under NHPA Section 106 for the mouse eradication program.”

8.7 EPA- APPROVED RODENTICIDE LABEL The proposed action would be implemented using the product Brodifacoum 25ppm. A Supplemental label has been obtained from US EPA that is specific to the mice eradication operation on Midway Atoll NWR and is based on Brodificoum-25D: Conservation (EPA Reg. No. 56228-37). Field applications of the rodenticide would be made in full compliance with the EPA supplemental label. An alternate bait formulation of Brodifacoum 25ppm may be considered. Given that the concentration of the active ingredient, brodifacoum, would be the same, this would not alter the environmental risk profile of the operation and all potential consequences discussed for Brodifacoum-25D: Conservation in this FEA and the Biological Assessment would apply.

8.8 EO 12372 INTERGOVERNMENTAL REVIEW Coordination and consultation with affected State governments, and other Federal agencies has been completed through project consultation procesesses and personal contact by Service planners, refuge managers, and supervisors.

8.9 MIGRATORY BIRD TREATY ACT AND EO 13186 FEDERAL PROTECTION OF MIGRATORY BIRDS The Final NEPA analysis evaluates the effects of proposed action on migratory birds and outlines avoidance and minimization measures that will be implemented to reduce or eliminate take of migratory birds. Appendix C of the Final EA identifies mitigation measures to protect migratory shorebirds.

8.10 EO 13112 RESPONSIBILITIES OF FEDERAL AGENCIES PERTAINING TO INVASIVE SPECIES EO 13112 requires Federal agencies to prevent the introduction of invasive species and provide for their control and to minimize the economic, ecological, and human health impacts that invasive species cause. This project is in compliance with this Executive Order.

8-2 FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT STATEMENT OF COMPLIANCE

8.11 INTEGRATED PEST MANAGEMENT, 517 DM 1 AND 569 FW 1 In accordance with 517 DM 1, and 569 FW1, an integrated pest management (IPM) approach has been adopted to eradicate, control, or contain pest and invasive species on the Refuge. In accordance with 517 DM 1, only pesticides registered with the Environmental Protection Agency (EPA) in full compliance with the Federal Insecticide, Fungicide, and Rodenticide Act (FIFRA), and as provided in regulations, orders, or permits issued by EPA, may be applied on lands and waters under Refuge jurisdiction. Consistent with agency policy, a Pesticide Use Plan (PUP) would be prepared prior to implementing the proposed action.

8.12 EPA – NATIONAL POLLUTANT DISCHARGE ELIMINATION SYSTEM (NPDES) GENERAL PERMIT The proposed action requires a Pesticide General Permit under the NPDES Program. Once the application is submitted, there is a 30-day period for EPA to respond. Permit application will be completed online at: epa.gov/NPDES/pesticide- permitting prior to implementing the proposed action.

8-3

FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT GLOSSARY OF TERMS

GLOSSARY OF TERMS Anticoagulant - a class of drugs that work to prevent blood clotting. Atoll - a ring-shaped reef, island, or chain of islands formed of coral. Bait Bucket - a piece of equipment used in many types of industry to discharge products at a steady rate.

Behaviorally plastic - change in an organism’s behaviors or habits that results from change in the environmental conditions, such as a shift to a new primary food source due to changes in food abundance.

Biological Control - the control of a pest by the introduction of a natural enemy or predator. Broadcast – (of pesticide) the uniform treatment to a broad area using various procedures of application, such as hand and aerial broadcast methods.

Brodifacoum - a second-generation rodenticide that requires only one feeding for a rodent to receive a lethal dose.

Climate - how the atmosphere “behaves” over relatively long periods of time. Coagulopathy - a condition in which the blood's ability to coagulate (form clots) is impaired. This condition can cause a tendency toward prolonged or excessive bleeding.

Colony (of seabirds) - a large group of birds from one or more species that nest or roost (sleep) close to each other at a particular location. Most seabirds are social nesters and display extraordinary site fidelity.

Colonization - the process in biology by which a species successfully spreads to a new area. Commensal - (pertaining to humans) eating together at the same table; (in ecology, of an animal, plant, fungus, etc.) living with, on, or in another, without injury to either.

Coumarin-based - (pertaining to anticoagulants) a class of vitamin K antagonist (VKA) anticoagulant drug molecules derived from coumarin (4-hydroxycoumarin).

Diphacinone - a first-generation rodenticide which requires multiple feedings over several days for a rodent to receive a lethal dose.

EC50 - Half Maximal Effective Concentration; the concentration of a toxicant that gives half- maximal response; measured in ppm.

ED50 -dose at which half of the test population will show an effect from an ingested substance (toxicant), but not necessarily death.

ED5 - 5% of the exposed population will be affected by coagulopathy from a rodenticide – but not necessarily death.

9-1 MIDWAY SEABIRD PROTECTION PROJECT FINAL ENVIRONMENTAL ASSESSMENT GLOSSARY OF TERMS

Endemic - a species that is native to just one place. Ephemeral (plants) - those which sprout, reproduce, and die back very quickly as an evolutionary adaptation to take advantage of brief wet periods in an otherwise dry climate.

Epipelagic - the oceanic zone extending from the surface to about 656 ft. (200 m) where enough light penetrates to allow photosynthesis Eradication - the complete removal of a damaging species from a specific location to enable ecosystem recovery.

Executive Order - presidential directives issued by United States presidents; generally directed towards officers and agencies of the Federal government.

Extinction - when the last of a species dies and that species ceases to exist anywhere in the world. Extirpation - the complete removal of an organism from a specific location but which continues to exist in other places; also known as local extinction.

Federal Mandate - orders that induce “responsibility, action, procedure or anything else that is imposed by constitutional, administrative, executive, or judicial action” for state and local governments and/or the private sector.

Federal Acts - laws, also known as statutes, passed by a legislature. Granivore - seed predators that feed on the seeds of plants as a main or exclusive food source, leaving seeds damaged and not viable.

Groundwater - water held underground in the soil or in pores and crevices in rock. Half-life - the time it takes for a certain amount of a pesticide to be reduced by half as it dissipates or breaks down in the environment; a pesticide will break down to 50% of the original amount after a single half-life.

Hemorrhaging - the flow of blood out from a blood vessel; bleeding. Herpetofauna - amphibians (frogs, toads, salamanders and newts) and reptiles (snakes, lizards, turtles, and crocodilians).

Hopper - a piece of equipment used in many types of industry to discharge products at a steady rate.

Immigration - the movement of an organism to a new area from elsewhere, assisted or unassisted. Insectivorous - an animal that eats insects as a primary or exclusive food source. Invasive - a non-native species whose introduction causes or is likely to cause economic or environmental harm, or harm to human health.

Ionic strength (of seawater) - a measure of the concentration of ions in a solution which affects important properties such as the dissociation or solubility of different salts.

9-2 FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT GLOSSARY OF TERMS

Island Biosecurity - The policies and protocols put in place to protect island ecosystems from non-native species by preventing, detecting and responding to introductions.

Lambda - Rate of growth (lambda) is the ratio of population size at the end of one interval to population size at the end of the previous interval. When lambda = 1.0 the population density is stable while values >1.0 indicate increasing populations.

LC50 - the concentration of the chemical in feed that kills 50% of test samples; usually administered over a multi-day period (e.g. 5 to 7 days).

LD50 - the amount of an ingested substance that kills 50% of test samples; usually administered as a single dose.

LOAEL - the lowest dose or exposure level of a toxicant that produces a measurable toxic effect on the test group of animals.

Mitigation - steps taken to reduce or avoid negative environmental impacts. Native - a species that occurs naturally (without human agency) in an area. NOAEL - a dose or exposure level of a toxicant that produces no measurable toxic effects on the test group of animals.

Non-native (introduced, alien) - an organism that is not native to the place in which it occurs, having been accidentally or deliberately transported to the new location by human activity. pp_ - parts per thousand (ppt), parts per million (ppm), parts per billion (ppb) are units of concentration for extremely dilute solutions; for example, a concentration of 1 ppm means 1 mg of solute in 1,000,000 mg of solution, or 1 mg of solute in 1000 g of solution, or 1 mg of solute in 1 kg of solution.

Palatability - having an agreeable or pleasant taste that is accepted by the target consumer. Pesticide - any substance or mixture of substances intended for preventing, destroying, repelling, or mitigating any pest.

Pica - displaying an indiscriminate preference for eating non-food items; such as chicks pecking and eating rocks, sticks and other foreign objects.

Pinnipeds - seals; a diverse group of carnivorous semi-aquatic marine mammals. Potable - safe to drink; drinkable. Predation - the act of one organism killing and eating other organisms; can refer to both animals and plants.

Pyranine - a fluorescent dye commonly found in highlighters and used as a biological stain to show ingestion pathways.

Recruitment - the ability of juvenile organisms to survive and add to the population of that species.

9-3 MIDWAY SEABIRD PROTECTION PROJECT FINAL ENVIRONMENTAL ASSESSMENT GLOSSARY OF TERMS

Refuge - (pertaining to wildlife) a naturally occurring sanctuary, such as an island, that provides protection for species from hunting, predation, or competition; it is a protected area, a geographic territory within which wildlife is protected.

Rodenticide – a pesticide formulated to kill rodents. Sublethal - having an effect less than lethal. Threshold - the magnitude or intensity that must be exceeded for a certain reaction, phenomenon, result, or condition to occur or be manifested.

Weather - conditions of the atmosphere over a short period.

THIS PAGE INTENTIONALLY LEFT BLANK

9-4 FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT APPENDIX A

APPENDIX A. MIDWAY ATOLL NATIONAL WILDLIFE REFUGE BIO- SECURITY PLAN Island biosecurity results from policies and protocols put in place to protect island ecosystems from non-native species by preventing, detecting and responding to introductions Non-native species (sometimes referred to as non-indigenous, alien, or exotic species) are plants, animals, and micro-organisms (PAMs) transported or established outside of their natural range due to the activities of humans, regardless of whether these actions are intentional or not. Non-native invasive species (NISs) have the potential to establish populations and cause unacceptable harm. The main points found within the Midway Atoll Wildlife Refuge Biosecurity Plan are briefly summarized in this appendix. They address prevention measures, reporting protocol, education, and rapid response. Preventing the introduction of NISs is the most time- and cost-effective way to protect island ecosystems. The focus of this summary is rodents, particularly mice and rats.

I. Prevention Measures Use a rodent-proof facility to hold supplies and equipment prior to shipping. Pack consumable goods in rodent-proof containers, and check non-food items carefully. Do not leave cargo outside over-night. On-island preparation includes setting traps/bait stations in key areas like the dock, airport, and cargo staging/storage areas, elevating them to avoid interference with crabs. Check all detection devices regularly and maintain them; have spare traps/bait on hand and replace as needed. Train personnel handling cargo to identify rodent signs, e.g., feces in containers, holes chewed in packaging, etc.

a. Boat Arrival All docking vessels must: (i) have rodent inspections prior to off-loading, (ii) deploy collared rat- guards to all dock lines, and (iii) have pest detection and control devices in use on board, e.g., snap-traps and glue-traps. Carry out rodent control measures 2 weeks before departure, 2 weeks prior to estimated arrival, and for the duration of stay.

b. Plane Arrival All planes must have rodent traps/bait stations deployed near the wheels. Those planes stored inside (e.g., hangers) must additionally deploy control devices around the edges of each structure.

II. Reporting Protocol Report any detections or suspicions of NIS presence immediately to the Refuge Manager who will coordinate appropriate follow-up.

III. Education Update residents and educate visitors on refuge biosecurity measures. Provide guidelines, signs, and brochures containing the basic protocols and procedures, e.g., how to detect, identify, and report NISs.

IV. Rapid Response if Animals Are Detected

A-1 FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT APPENDIX A

a. Rodents i. Use a variety and combination of removal methods, e.g., snap traps, bait stations, flavored sticky traps, rodenticides, and cage traps. Exact types of devices and methods will be determined by target rodent. ii. Bait station grid should cover all habitat types across the island. Bait stations and traps should be placed at a higher density around key habitat and detection sites. iii. All trap and bait station locations should be numbered, visibly marked, and mapped. Any member of the response team should be able to easily locate every location. iv. Place traps in locations with plenty of natural cover, and where animals are likely to be active. Place additional traps near any footprints or scat. v. Traps should be covered and/or placed in locations (e.g. attached to tree limbs) that reduce the chance of interference by non-targets. vi. Bait traps with known attractants. Check all traps daily and bait stations daily or every other day. Peanut butter mixed with rolled oats makes good rodent bait. vii. Keep detailed records. Any sign should be recorded and analyzed. viii. Any specimens caught are to be aged, sexed, breeding status obtained and have samples collected for DNA. DNA analysis may help determine source population. The specimen is then to be frozen (well labeled) in case required for later analysis. ix. Staff should continually search for signs and new trap locations.

b. Mammalian Predators i. A variety of trap types should be used. Possibilities include snares, foothold traps, conibear traps, and box traps. Exact size and trapping techniques will be dictated by the target animal. ii. Bait traps with known attractants. Check all traps daily and bait stations daily or every other day. Peanut butter mixed with rolled oats makes good rodent bait.

c. Reptiles and Amphibians i. Active search and capture of individuals by hand, with sniffer dogs, or using nets or nooses. ii. Drift fences with funnel traps

d. Post-Animal Review

A-2 FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT APPENDIX A

i. Monitor island-wide to verify that all individuals have been removed and a population was not established. During and after an introduction, review the current biosecurity practices and identify how the animal arrived on island. Any failures identified in the biosecurity protocols must be re-evaluated to prevent similar situations from occurring.

A-3

FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT APPENDIX B

APPENDIX B. SUMMARY OF PROPOSED MONITORING Documentation of the best operational approach, the efficacy of the operation, the environmental and ecological effects of the proposed action requires monitoring of a number of parameters before, during, and after the implementation of an eradication operation will be undertaken by the Restoration partners.

I. Eradication Planning To inform eradication planning efforts where possible (e.g., baiting methods, evaluation of methods to reduce non-target impacts of the eradication procedure) we will evaluate bait uptake rates and persistence in representative habitats and use DNA Metabarcoding to provide direct evidence for mice diet components. We will collect mouse diet samples via stomach contents and mouse feces in stratified habitat types; collect samples of diet items if DNA hasn’t been previously characterized. These data can directly inform eradication planning and baiting strategy. We will monitor phenology of plants suspected of being important in mouse diets.

II. Conservation Measures To measure recovery of target conservation species and ecosystems and the unintended consequences of the removal of a ubiquitous species, a number of ecosystem components will be characterized prior to and at 1, 3, 5 and 10 years post-eradication. Good monitoring will enable land managers to predict and prepare for positive and negative impacts of the eradication action (pre-eradication) and clearly document conservation gains (post-eradication). Taxa identified for monitoring, parameters to be measured, and methods include:

4. Seabirds, especially those burrow nesters and ground-nesters particularly vulnerable to rodent predation such as Laysan Albatross, Tristram’s Storm-petrel, Bulwer’s Petrel, and Bonin Petrel for which we will measure breeding population size using counts and acoustic activity and reproductive performance and impacts of mice to Bonin petrels by using camera traps and tracking cards to document mouse-seabird interactions. We will deploy 10-13 Songmeters across Sand and Eastern (control) at known and potential breeding locations to record call activity to record population densities of cryptic burrow nesting species such as Tristram’s Storm-petrels and Bulwer’s Petrels. We will observe and quantify impacts of mice on Laysan and Black-footed Albatross mice and chicks within and outside of mouse control treatment areas by monitoring albatross nest density and reproductive success within a subset of existing albatross demography 20 x 20 m plots as well as additional plots (e.g., zones where no baiting is being conducted as well as Eastern Island as a control).

5. Laysan Duck (also, see Appendix C), for which we will monitor population size, reproductive performance, and behavior. We will quantify direct and indirect impacts of mice on duck population using camera traps on nests and conducting measures of reproductive performance and foraging behavior before and after implementation. After live-capture and holding ducks on Sand or Eastern Island, birds will be released when bait pellets are no longer available to be consumed or biological samples

B-1

FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT APPENDIX B

collected post-application indicate little to no risk (Appendix C). Monitoring will also include recapture of ducks demonstrating signs of toxicosis.

6. Arthropod community, for which we will document changes in population densities of land crabs, insects, and spiders. We will confirm extant land crab species for Midway Atoll and document density and demography of land crab population by conducting opportunistic searches and observations of crabs and potential crab holes that are not obviously the known species, Ocypode pallidul, and surveying quadrats across potential crab habitat, counting burrows and measuring width of burrows. We will monitor arthropod abundance and richness by deploying pitfall traps in stratified habitats and count sample and identify them to families.

7. Plant community and seed production, especially species known to provide important forage for native species to understand the extent of competition between introduced house mouse and native birds and insects. We will quantify native seed availability (potential food resource for both mice and ducks) before and after mouse removal by counting seeds and looking at plant recruitment rates for selected species. DNA Metabarcoding will help to predict where land managers can expect to see changes in flora and fauna and identify areas of focus for pre-eradication monitoring efforts.

III. Efficacy Monitoring Documentation of bait persistence and availability during the implementation period over all habitat types will inform practitioners of appropriateness of application rates. Telemetry of radio- tagged mice during the implementation will allow us to track fates of a sample of individuals. Post eradication detection methods including chew blocks, trail cameras, and other techniques will evaluate successful removal of all mice from Sand Island. Monitoring populations of fish, birds, and plants and controlling populations of invasive species currently takes place on MANWR for management purposes. The refuge staff and personnel from coordinating external agencies will follow established protocols that have been enhanced in scope for surveying and sample-collecting during pre- and post-eradication. While we are still working out details, we currently plan to work with a qualified lab off-site to test for brodifacoum residue in non-target carcasses, prey species, and in the environment over time. Sampling and analysis will be done at appropriate intervals. Surveys of the beaches, coastal waters, and terrestrial areas will be conducted to monitor various aspects of the operation. These surveys will include assessing bait persistence, removing mouse carcasses, retrieving sick or injured wildlife, and collecting samples (including non-target fatalities) for pesticide testing. Every precaution will be taken to avoid impacting non-target species while maximizing the chance of project success. We will use adaptive management techniques to adjust the project if we experience any significant unintended consequences to non- target organisms.

IV. Environmental Impact and Residue Monitoring 1. Brodifacoum residue monitoring: Environmental brodifacoum residues will be evaluated by testing of soil and seawater samples before and after baiting operations. Brodifacoum residues in living tissues (e.g., food web compartments) will be assessed by collection and euthanasia of appropriate invertebrates, lizards, fishes and birds, with liver tissues (site of

B-2 FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT APPENDIX B

greatest accumulation) harvested and submitted for chemistry (whole-body samples will be shipped for processing, with tissue harvest occurring under residue-sanitary conditions). Any tissues representative of items for human food consumption (mollusks, game fish) will have whole-body or muscle tissues analyzed as well. Cockroaches, which are demonstrated to be a significant consumer of rodenticide baits and in turn are heavily consumed by Laysan Ducks, will be a primary focus of sampling, with diminishment of brodifacoum levels to be confirmed before ducks are returned to Midway Island (see Appendix C). 2. Mortality of all non-target organisms associated with the baiting operations will be assessed by active searching for non-target carcasses on terrestrial and near-shore marine environments (including recording of data on search effort), and by opportunistic sampling during other aspects of operational activities. Carcasses that appear to be within approximately three days of time of death, for which cause of death is not obvious (e.g., aircraft strike), will be collected and submitted for brodifacoum residue testing. 3. Sampling of prey species important to terrestrial vertebrates vulnerable to brodifacoum will continue periodically until residue levels become undetectable or until the levels are deemed not harmful to Laysan ducks and other migratory bird species foraging in the terrestrial environment on Sand Island. Chemical analyses (assay and quantification by liquid chromatography and tandem mass spectrometry, “LC-MS/MS”) will be conducted by the USDA NWRC Chemistry Lab Unit in Fort Collins, Colorado. Detection and quantitation limits for each sample type will be established during analysis. Remaining tissues or homogenates may be made available for confirmatory testing by external agencies. Given that brodifacoum has higher toxicity and a longer half-life (compared to first-generation anticoagulants), and therefore an increased risk of bioaccumulation in the food web, sampling will continue over a long enough timeframe to ensure that appropriate environmental thresholds have been met for actions such as the release of Laysan Ducks from protective captivity.

V. Mitigation Measures and Effectiveness Monitoring Mitigation measures and actions identified for multiple species would be carried out and monitored for their effectiveness. Effectiveness monitoring tracks the success in achieving desired outcomes and evaluating environmental effects. Mitigation includes specific measures or practices that would reduce, avoid or eliminate the effect of the proposed action on non-target species. In this EA, the identified mitigation measures are part of the proposed action (project) and necessary to support a FONSI. Examples of mitigation measures include: training ground-based staff to identify endangered plants and how to avoid stepping on seabird burrows; capturing and moving vulnerable species to avoid rodenticide exposure; measures to minimize bait entering the marine environment; and measures to reduce impacts to sea turtles and monk seals. For detailed descriptions see the mitigation section for each species in Chapter 5 and details of the Laysan duck protection plan in Appendix C.

B-3

FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT APPENDIX C

APPENDIX C. SUMMARY OF RISKS AND PROPOSED PROTECTION PLAN FOR LAYSAN DUCKS AND SHOREBIRDS

C-1

Summary of Risks and Proposed Protection Plan for Laysan Ducks and Shorebirds at Midway Atoll During a Mouse Eradication Campaign

Prepared by:

Hamer Environmental LP P.O. Box 2561, Mount Vernon, WA, 98273

November 2018

SUMMARY OF RISKS AND PROPOSED PROTECTION PLAN FOR LAYSAN DUCK AND SHOREBIRDS FOR THE MIDWAY SEABIRD PROTECTION PROJECT TABLE OF CONTENT

TABLE OF CONTENT

1. PROJECT INTRODUCTION ...... 1 1.1 SUMMARY OF RISKS TO LAYSAN DUCKS ...... 1 1.2 SUMMARY OF RISKS TO BRISTLE-THIGHED CURLEW AND OTHER SHOREBIRDS ...... 2 1.3 EFFECTS OF TIMING OF ERADICATION ON NON-TARGET ORGANISMS...... 5 2. BENEFITS OF MOUSE ERADICATION TO LAYSAN DUCKS AND BRISTLE- THIGHED CURLEW ...... 9 3. SUMMARY OF AVIAN CAPTURE METHODS ...... 10 4. LAYSAN DUCK ...... 12 4.1 LAYSAN DUCK MITIGATION PLAN ...... 12 4.2 PROPOSED LAYSAN DUCK PROTECTION PLAN ...... 13 4.3 INFRASTRUCTURE DEVELOPMENT AND PERSONNEL ...... 14 Infrastructure Development ...... 14 Sand Island ...... 15 Eastern Island ...... 15 Eastern Island Freshwater Infrastructure ...... 16 Personnel ...... 17 Capture Methods and Materials ...... 17 4.4 CAPTIVE HOLDING AND TRANSPORT ...... 19 Aviary Structures ...... 19 Predator Control near Sand Island Aviaries ...... 20 4.4.2 Holding and Transport Boxes ...... 21 4.5 CAPTIVE CARE ...... 22 Examination and Treatment ...... 22 Feeding and managing social interactions ...... 22 4.6 RELEASE ...... 23 Carcass and Bait Monitoring Before and After Release ...... 23 Release Method ...... 23 Pre- and Post-project Monitoring ...... 24 5. BRISTLE-THIGHED CURLEW AND OTHER SHOREBIRDS ...... 26 5.1 BRISTLE-THIGHED CURLEW AND SHOREBIRD MITIGATION PLAN ...... 26 5.2 POTENTIAL MITIGATION OPTIONS ...... 26 5.3 INFRASTRUCTURE DEVELOPMENT AND PERSONNEL ...... 28 5.4 CAPTURE METHODS AND MATERIALS ...... 28 5.5 CAPTIVE HOLDING AND TRANSPORT ...... 31 Aviary Structures ...... 31 Perches and Visual Barriers ...... 33 Predator Control ...... 34 Holding Boxes ...... 34 Transport Boxes ...... 36 5.6 CAPTIVE CARE ...... 36 Examination and Treatment ...... 36 Feeding ...... 37 5.7 RELEASE ...... 38 Carcass and Bait Monitoring Before and After Release ...... 38 Release Method ...... 38 5.8 PRE- AND POST-PROJECT MONITORING ...... 39 Sentinel Birds – Capture, Care and Release ...... 40

ii

SUMMARY OF RISKS AND PROPOSED PROTECTION PLAN FOR LAYSAN DUCK AND SHOREBIRDS FOR THE MIDWAY SEABIRD PROTECTION PROJECT TABLE OF CONTENT

6. SUMMARY ...... 40 7. LITERATURE CITED ...... 42

iii

SUMMARY OF RISKS AND PROPOSED PROTECTION PLAN FOR LAYSAN DUCK AND SHOREBIRDS FOR THE MIDWAY SEABIRD PROTECTION PROJECT TABLE OF CONTENT

LIST OF FIGURES

FIGURE 1. SHOREBIRD SURVEYS RESULTS AT MIDWAY JUNE – DEC 2017...... 4 FIGURE 2. PACIFIC GOLDEN PLOVER NUMBERS AT TERN ISLAND, FRENCH FRIGATE SHOALS FROM 1988 THROUGH 1992, THESE DATA ALSO SHOW THE SPIKE OF BIRD NUMBERS IN SEPTEMBER EACH YEAR THAT MAY REPRESENT BIRDS PASSING THROUGH BUT THEN FLYING FURTHER SOUTH...... 4 FIGURE 3. DUCK DETECTABILITY INCREASED MID-AUGUST AND DECREASED DRAMATICALLY MID- TO LATE FEBRUARY. DUCKLINGS ARE FIRST SEEN LATE MARCH TO EARLY APRIL WITH HIGHEST NUMBERS BETWEEN APRIL AND JULY. HOWEVER, THEY CAN STILL BE SEEN THROUGH OCTOBER AND NOVEMBER (RARE). EASTERN ISLAND: DURING THE WINTER MONTHS, WE TYPICALLY SAW 111-195 DUCKS. FROM THIS SURVEY, WE CAN ESTIMATE MIDWAY’S POPULATION TO BE APPROXIMATELY 600 WITH A RANGE OF 526-686 (95% CI)...... 6 FIGURE 4. USGS BIOLOGIST MICHELLE REYNOLDS HERDS DUCKS TOWARDS A NOOSE CARPET, A TRAP USED TO CATCH MANY OF THE 28 DUCKS TRANSLOCATED FROM MIDWAY ATOLL NATIONAL WILDLIFE REFUGE TO KURE ATOLL STATE WILDLIFE SANCTUARY WITHIN PAPAHĀNAUMOKUĀKEA MARINE NATIONAL MONUMENT ON SEPTEMBER 3, 2014...... 10 FIGURE 5. GUZZLER AND SURROUNDING VEGETATIVE COVER AT GREEN ISLAND, KURE ATOLL...... 14 FIGURE 6. VIEWS FROM INSIDE AND OUTSIDE ONE OF THE 8X14 FT AVIARY COMPARTMENTS THAT WAS USED TO HOLD DUCKS DURING THE 2004 AND 2005 TRANSLOCATION. THE COMPARTMENT HAD A SAND FLOOR, NATURAL COVER, AND ARTIFICIAL COVER STRUCTURES...... 16 FIGURE 7. LAYOUT OF LAYSAN DUCK TRAP. THE COVERING FOR THE HOLDING AREA IS NOT INCLUDED TO SHOW THE DESIGN...... 18 FIGURE 8. LAYSAN DUCK TRAP WITH THE COVERING ON...... 18 FIGURE 9. THE TRANSPORT BOX ALLOWS LIGHT TO ENTER THROUGH BOTH SIDES...... 19 FIGURE 10. NATURAL COVER, ARTIFICIAL COVER, FOOD BOWLS, BATH TUBS, AND WATERER INSIDE A 2004 LAYSAN DUCK AVIARY COMPARTMENT...... 20 FIGURE 11. USCG STAFF OFFLOAD LAYSAN DUCKS TO USGS BIOLOGIST MICHELLE REYNOLDS ON KURE ATOLL DURING A MAJOR TRANSLOCATION TO MOVE 28 LAYSAN DUCKS FROM MIDWAY ATOLL NATIONAL WILDLIFE REFUGE TO KURE ATOLL STATE WILDLIFE SANCTUARY WITHIN PAPAHANAUMOKUAKEA MARINE NATIONAL MONUMENT SEPTEMBER 2014. PHOTO BY: JOHN KLAVITEER/USFWS...... 21 FIGURE 12. USCG ENS KANE HOLDS THE DOOR OF A TRANSPORT BOX OPEN AS USFWS JOHN KLAVITTER PLACES A DUCK INSIDE ITS TRANSPORT CAGE IN PREPARATION FOR TRANSLOCATING 28 LAYSAN DUCKS FROM MIDWAY ATOLL NATIONAL WILDLIFE REFUGE TO KURE ATOLL STATE WILDLIFE SANCTUARY IN 2014. PHOTO BY: ERIC DALE/FWS VOLUNTEER...... 21 FIGURE 13. SHOREBIRD SURVEYS FOR SAND ISLAND (TOP) AND EASTERN AND SPIT ISLANDS (BOTTOM) IN 2017...... 27 FIGURE 14. WES JOLLEY CREATING SHOREBIRD ATTRACTION ZONES TO AID IN CAPTURE USING WATER AND DECOYS...... 29 FIGURE 15. DECOYS WERE TESTED IN SINGLE SPECIES AND MULTI SPECIES ARRAYS...... 29 FIGURE 16. A TRAIL CAMERA RECORDS ATTENDANCE AT THE ATTRACTION SITES...... 30 FIGURE 17. THIS SHOREBIRD AVIARY WAS SET ON A CONCRETE PAD AND DESIGNED WITH VISUAL BARRIERS (ALUMINUM SHEETING AND PALM FRONDS) AND PREDATOR CONTROL DEVICES (ALUMINUM SHEETING AND RODENT TRAPS SET ON 5 GAL. [19 L] BUCKETS)...... 31 FIGURE 18. THE INSIDE OF THIS AVIARY SHOWS A RUBBERIZED FLOOR BEFORE IT WAS DIVIDED UP INTO TEN, 9.8 X 9.8 FT. (3 X 3 M) UNITS...... 32 FIGURE 19. A 5 GAL. (19 L) BUCKET BEING USED AS AN EXTERNAL PUMP FILTER...... 33 FIGURE 20. SALT WATER BEING PUMPED INTO A HOLDING TANK BEFORE BEING PUMPED TO THE AVIARY. THIS ALLOWS AVIARIES TO BE CLEANED DURING LOW TIDE WHEN SEAWATER IS UNAVAILABLE...... 33 FIGURE 21. PADDING CAN BE ADDED TO PERCHES TO PREVENT OR ALLOW FOOT SORES TO HEAL...... 34 FIGURE 22. HOLDING BOXES SHOWING HOLES FOR VENTILATION AND BUCKET LIDS FILLED WITH WATER TO ...... 35 PREVENT ANT INFESTATION...... 35 FIGURE 23. EACH HOLDING BOX HAD A REMOVABLE TRAY TO ALLOW FOR EASY ...... 35 CLEANING...... 35 FIGURE 24. BIRDS WERE OFFERED A VARIETY OF FOOD ITEMS...... 38

iv

SUMMARY OF RISKS AND PROPOSED PROTECTION PLAN FOR LAYSAN DUCK AND SHOREBIRDS FOR THE MIDWAY SEABIRD PROTECTION PROJECT TABLE OF CONTENT

v

SUMMARY OF RISKS AND PROPOSED PROTECTION PLAN FOR LAYSAN DUCK AND SHOREBIRDS FOR THE MIDWAY SEABIRD PROTECTION PROJECT TABLE OF CONTENT

LIST OF TABLES

TABLE 1. LAYSAN DUCKS AND SHOREBIRDS OCCUR IN MODERATE TO HIGH ABUNDANCE AT MANWR (N = 6 SPECIES. MBTA = PROTECTED UNDER THE U.S. MIGRATORY BIRD TREATY ACT OF 1918, US ESA = U.S. ENDANGERED SPECIES ACT, IUCN RED LIST = INTERNATIONAL UNION FOR CONSERVATION OF NATURE RED LIST OF THREATENED SPECIES)...... 3 TABLE 2. THE 2019 STAFFING SCHEDULE FOR NON-TARGET SPECIES MANAGEMENT FOR CAPTURE AND CAPTIVE CARE OF BIRDS. V = WILDLIFE VETERINARIAN, A = AVICULTURIST, C = CAPTURE SPECIALIST AND, P = PROTECTION TEAM MEMBERS. PROTECTION TEAM MEMBERS WILL ASSIST WITH CAPTIVE CARE TASKS INCLUDING FOOD PREPARATION, FEEDING AVIARY BIRDS, CLEANING AVIARIES, OBSERVATION OF MARKED BIRDS ON EASTERN, SUPPLEMENTAL FEEDING ON EASTERN, GUZZLER CLEANING, SURVEILLANCE ON SAND, AND OTHER DUTIES...... 8

vi

SUMMARY OF RISKS AND PROPOSED PROTECTION PLAN FOR LAYSAN DUCK AND SHOREBIRDS FOR THE MIDWAY SEABIRD PROTECTION PROJECT PROJECT INTRODUCTION

1. PROJECT INTRODUCTION Midway Atoll National Wildlife Refuge (MANWR) is part of the Northwest Hawaiian Islands (NWHI) archipelago, and located approximately 1,250 mi (2,000 km) west-northwest of Honolulu, Hawai`i. The refuge lies within the Papahānaumokuākea Marine National Monument (PMNM) and has also been designated the Battle of Midway National Memorial. Midway Atoll hosts shallow lagoons and three low-lying islands (Sand, Eastern, and Spit Islands) that are encompassed by a nearly circular fringing coral reef. This isolated refuge in the Pacific Ocean is the breeding grounds for millions of seabirds, a haven for thousands of migratory shorebirds, and one of only three locations that harbor populations of the Federally endangered Laysan duck (Anas laysanensis). MANWR is rodent-free except for Sand Island, where house mouse (Mus musculus) populations and their negative impacts to native species and habitats continue to occur despite on-going rodent control and management efforts. The Proposed Action by the USFWS to eradicate the mice at MANWR involves aerial and ground applications of bait pellets containing rodenticide (Brodifacoum-25D) into all potential mouse territories on Sand Island, including hand- broadcasting of bait in sensitive areas (e.g., narrow shorelines) and placing bait stations in commensal areas. The impacts from invasive predatory mammals, including mice and rats (Mus musculus and Rattus spp.), are one of the leading causes of species extinction on islands (Blackburn et al. 2004, Duncan and Blackburn 2007). Over three million birds, of 25 different species, can be found at MANWR and all of them are potentially susceptible to predation by mice. Predation of vulnerable populations of native seabirds is a real and ongoing threat on Sand Island that demands an immediate and effective response. Eradication of the house mouse would also facilitate the protection and restoration of all native species and habitats present in the refuge, including federally listed species and their habitats.

1.1 SUMMARY OF RISKS TO LAYSAN DUCKS MANWR supports the federally endangered Laysan duck, a rare, non-migratory, year-round resident of the refuge (Table 1). The Laysan duck persists in the wild only on Laysan Island and at Midway and Kure Atolls. Globally, there are fewer than 1,100 Laysan ducks occurring at these 3 sites and approximately half of this population occurs on Eastern and Sand Islands of MANWR (SWCA 2017). Current threats to Laysan ducks are made more significant by a small population size and extremely limited distribution which make this species highly vulnerable to demographic fluctuation and the chance of adverse environmental influences from droughts, severe storms, epizootics, predators, and invasive species. Non-native invasive species (mice, weeds, and possibly predatory insects such as some ant species) can alter the Laysan duck’s habitat. Habitat degradation and loss within PMNM may be intensified by increased storm severity and sea level rise associated with global climate change.

1

SUMMARY OF RISKS AND PROPOSED PROTECTION PLAN FOR LAYSAN DUCK AND SHOREBIRDS FOR THE MIDWAY SEABIRD PROTECTION PROJECT PROJECT INTRODUCTION

As mouse predatory behavior is likely to spread on Sand Island, the negative impacts to Laysan ducks could be significant to the global population. If mice are tenacious enough to attack and cause mortality to the Laysan albatross, one of the largest ground-nesting seabirds at MANWR, then it is likely that mice are having similar negative effects on other ground nesting birds, including the Laysan duck. In addition, mice could be competing with ducks for invertebrate food resources. Mice are known to be a seed predator on Eragrostis variabilis, an important grass providing cover for Laysan ducks. The proposed mouse eradication would allow recruitment of native grasses useful for Laysan ducks as shelter, nesting, and foraging habitat (VanderWerf 2012). Laysan ducks are non-migratory, year-round residents at MANWR, and those on Sand Island would be exposed to rodenticide bait pellets during and after the eradication action. Initial tests at sites on MANWR where non-toxic bait piles were put out and monitored, indicated that Laysan ducks would readily consume bait pellets. Therefore, Laysan ducks would likely ingest Brodifacoum-25D rodenticide bait pellets if they encounter them. Bait pellets would remain accessible on the ground until they are all consumed by mice or terrestrial invertebrates, or until environmental factors cause the grain-based pellet matrix to disintegrate. If bait pellets are consumed, primary poisoning is likely unless an affected duck is administered vitamin K immediately. Secondary poisoning is possible as long as the toxicant remains available in the ecosystem, or there are invertebrates with brodifacoum residue in their bodies (possibly up to 6 months). Human disturbances also pose a risk; impacts can occur during capture and captive holding, hand-broadcasting of bait pellets, and post-project monitoring efforts. Impacts could include trampling of vegetation and nests and inducing stress. Thus, there is a clear primary route of exposure to the rodenticide as it is assumed that ducks would consume bait. For primary exposure to the rodenticide resulting in illness or mortality, the calculated threshold whole body ED5 is 0.47-1.8 µg total intake of brodifacoum for an average Laysan Duck (Mineau 2018). In addition, the primary food source of Laysan ducks is invertebrates and thus ducks are likely to sustain secondary or tertiary exposure to rodenticides after ingesting these smaller organisms that previously consumed bait. Therefore, secondary exposure to the rodenticide is likely. Without mitigation, a large portion of the population of ducks present on MANWR would very likely succumb to the toxic effects of the rodenticide through primary or secondary exposure. Mitigation measures for Laysan ducks during and following the proposed mouse eradication would ensure that, at minimum, a viable portion of the population is protected for later release back on Sand Island. Based on the life history characteristics of the Laysan duck, the best time to conduct a mouse eradication would be during the non-breeding season (October through February) (SWCA 2017).

1.2 SUMMARY OF RISKS TO BRISTLE-THIGHED CURLEW AND OTHER SHOREBIRDS Shorebirds are present year-round at MANWR. However, their numbers change seasonally, dropping considerably around April/May when breeders depart, and increasing in September when breeders return for the winter. Summer shorebird populations in the refuge consist of juveniles and non-breeders, with non-breeding due to injury, illness, or senescence, for example. There are 5

2

SUMMARY OF RISKS AND PROPOSED PROTECTION PLAN FOR LAYSAN DUCK AND SHOREBIRDS FOR THE MIDWAY SEABIRD PROTECTION PROJECT PROJECT INTRODUCTION species of shorebirds that occur in moderate to high abundance and MANWR (Table 1), 4 of which are present in any significant numbers during the month of July when the application of rodenticide is expected to occur. These include the bristle-thighed curlew, Pacific golden plover, wandering tattler, and ruddy turnstone. The bristle-thighed curlew and Pacific golden-plover are designated as species of high conservation concern in USFWS National and Regional Shorebird Plans, including the U.S. Pacific Islands Regional Shorebird Conservation Plan (Engilis and Naughton 2004). Both of these species overwinter at Midway, with low numbers of juvenile or non-breeding birds also present through the summer months. All of the shorebird species using MANWR demonstrate a pattern of greater abundance during the winter months, with very few or no individuals remaining throughout the summer when adult birds return to the breeding grounds in the Arctic. The bristle-thighed curlew (Numenius tahitiensis) is specifically addressed in this plan because of its small global population size and the anthropogenic pressures on their wintering grounds, which include habitat loss, habitat degradation, and introduced mammalian predators. It is the only one of MANWR’s shorebirds listed as Vulnerable (BirdLife International 2016) (Table 1). This species has a declining global population of approximately 7,000 breeding adults (BirdLife International 2016). Bristle-thighed curlews occur on Sand Island, with breeding birds absent from about May through the end of July. Juvenile and non-breeding birds are present throughout the summer months (USFWS, unpublished) and a pulse of birds may actually start passing through from late July through September (Figure 1).

Table 1. Laysan ducks and shorebirds occur in moderate to high abundance at MANWR (n = 6 species. MBTA = protected under the U.S. Migratory Bird Treaty Act of 1918, US ESA = U.S. Endangered Species Act, IUCN Red List = International Union for Conservation of Nature Red List of Threatened Species).

Common Name Scientific Name Federal Status IUCN-list

Laysan duck Anas laysanensis ESA Endangered Critically endangered

ruddy turnstone Arenaria interpres MBTA Least concern

wandering tattler Tringa incana MBTA Least concern

Pacific golden plover Pluvialis fulva MBTA Least concern

bristle-thighed curlew Numenius tahitiensis MBTA Vulnerable

sanderling Calidris alba MBTA Least concern Source: USFWS (2017)

3

SUMMARY OF RISKS AND PROPOSED PROTECTION PLAN FOR LAYSAN DUCK AND SHOREBIRDS FOR THE MIDWAY SEABIRD PROTECTION PROJECT PROJECT INTRODUCTION

Shorebird Survey Midway Atoll

1600 1419 1400 1200 1071 880 1000 818779 800 633 600 355 351 416410 400 278 242 186 168 225 145 84 102 200 15 5 6 52 4 15 67 33 8 66 0 6/23/2017 7/27/2017 8/8/2017 8/30/2017 9/15/2017 10/20/2017 12/22/2017

Sum Of Plover Sum Of Turnstone Sum Of Tattler Sum Of Curlew

Figure 1. Shorebird surveys results at Midway June – Dec 2017.

Figure 2. Pacific Golden Plover numbers at Tern Island, French Frigate Shoals from 1988 through 1992, These data also show the spike of bird numbers in September each year that may represent birds passing through but then flying further south. Bristle-thighed curlews walk as they pick up items from the ground, also probing in soil or mud with their long bills, to feed on crustaceans, small fish, insect grubs, and snails. This species has also been observed catching and eating mice. In the non-breeding season bristle-thighed curlews forage primarily in terrestrial habitats consuming spiders, land crabs, insects, seabird eggs, lizards, and carrion (Marks 1993). Additionally, research and direct observation during the Palmyra rat eradication has shown that some bristle-thighed curlews will ingest baits containing the rodenticide directly so the risks to this species are not confined to secondary poisoning but include primary poisoning as well (Pierce et al. 2008). Given the diverse diet of the bristle-thighed curlew there are several potential secondary and tertiary exposure pathways for the species including:

4

SUMMARY OF RISKS AND PROPOSED PROTECTION PLAN FOR LAYSAN DUCK AND SHOREBIRDS FOR THE MIDWAY SEABIRD PROTECTION PROJECT PROJECT INTRODUCTION

 Feeding on crabs that have consumed bait or scavenged mice carcasses;  Feeding on insects that have consumed bait or scavenged mice carcasses;  Feeding on geckoes or other lizards that have consumed insects that have ingested the bait; or  Feeding directly on live mice, moribund mice, or carcasses of mice that have consumed bait. Sanderlings occur in very low numbers on Sand Island. No individuals of this species were recorded on MANWR during avian surveys conducted in June, July, August and September 2017. Pacific golden plover is the most abundant shorebird in some months in late summer and fall and have been sighted regularly on the shoreline of all 3 islets at MANWR, in the forested areas, and in open, grassy areas (Figures 1 and 2). Wandering tattler and ruddy turnstone both occur on Sand Island. Breeding birds are absent at MANWR from about May through July or August (Figure 1), but low numbers of juvenile and non-breeding birds of each of these 2 species are also present throughout the summer months. Pacific golden plover, wandering tattler and ruddy turnstone primarily consume a wide range of invertebrate and possibly vertebrate prey, which makes secondary poisoning a potential risk. Each is also documented to consume foods comprised of vegetable matter (natural and artificial). Research has shown that shorebirds may consume rodenticide bait pellets directly (Pierce et al. 2008) or may sustain secondary or tertiary exposure to rodenticides after ingesting smaller organisms that previously consumed bait. In a study of mortality of shorebirds after the Palmyra rat eradication using brodifacoum, Pitt et al. (2015) found that all bristle-thighed curlew (6 birds), Pacific golden plover (2 birds), ruddy turnstone (2 birds) and wandering tattler (1 bird) mortalities collected after a rat eradication on Palmyra Atoll had detectable brodifacoum residues and most likely died from hemorrhage. Carcass searches on Palmyra Atoll continued daily until 4 August (Pitt et al. 2015). In summary, the mouse eradication project on Sand Island places non-resident migratory shorebirds, including bristle-thighed curlews, at risk of both primary and secondary exposure to rodenticide bait that would be broadcast across the island.

1.3 EFFECTS OF TIMING OF ERADICATION ON NON-TARGET ORGANISMS. The optimum time to apply rodenticide is informed by several considerations having to do with the effects on efficacy, human safety, and effects on non-target organisms. The best timing for each consideration does not always align with that for the others. Some deviations from the best time for any particular consideration can be managed by changing other parameters. At Midway the best time to do the work with respect to Laysan Ducks would be between October and February when the fewest birds are breeding or molting. The best time to do the rodenticide application for shorebirds such as Bristle-thighed Curlews would be in late May as soon as all the adults intending to fly north to breed had departed. The best time to do the application for ensuring that every mouse has access to and interest in taking the bait would be when food resources are declining in abundance and mouse populations have peaked and are starting downward. This part of the regular

5

SUMMARY OF RISKS AND PROPOSED PROTECTION PLAN FOR LAYSAN DUCK AND SHOREBIRDS FOR THE MIDWAY SEABIRD PROTECTION PROJECT PROJECT INTRODUCTION cycle of mouse population size happens at different times during different years but is most typically during the summer and early fall at Midway. Other factors that can affect success of bait applications include timing of rainfall and wind velocity. The constraint most relevant to human safety and thus the one that will put bounds on the range of times that can be considered is the number of adult albatrosses flying into the colony. This colony is the largest in the world and any at particular time may have as many as 500,000 breeding albatrosses that each weigh between 5 and 6 pounds coming and going as they bring food to their chicks. For reduction of bird airstrike hazard to the helicopters spreading the rodenticide, the window chosen to do this work should start in July when many of the chicks have fledged and extend until mid-October, when the earliest breeders start showing up for a new season. Given the window dictated by human safety during flight operations (July to mid-October) the choices for the other efficacy and non-target considerations become more difficult. Figure 3 shows results of Laysan duck counts at Midway Atoll and supports the idea that waiting until mid- September to apply bait would allow more ducks to be captured because they are less cryptic (more detectable) after breeding and molting is completed and would not need to be held in captivity for as long because the team would not need to start capturing early in the year before the breeding season starts.

Laysan Duck Activity on Midway Atoll 400 350 300 250 200 150 100 50 0

Sand ducklings Eastern ducklings AHYs Sand AHYs Eastern Total

Figure 3. Duck detectability increased mid-August and decreased dramatically mid- to late February. Ducklings are first seen late March to early April with highest numbers between April and July. However, they can still be seen through October and November (rare). Eastern Island: During the winter months, we typically saw 111-195 ducks. From this survey, we can estimate Midway’s population to be approximately 600 with a range of 526-686 (95% CI).

Optimal timing for bait application to minimize shorebird impacts is during late May as soon as the adult birds have departed the wintering grounds to travel North to breed. As discussed earlier, this time falls outside the flight safety window of late July to mid–October. The later window that

6

SUMMARY OF RISKS AND PROPOSED PROTECTION PLAN FOR LAYSAN DUCK AND SHOREBIRDS FOR THE MIDWAY SEABIRD PROTECTION PROJECT PROJECT INTRODUCTION is best for Laysan Duck capture (late September start) falls during a period less desirable for shorebird protection both because most winter resident shorebirds will have returned and because there may even be a pulse of birds stopping by on their way further south to other wintering sites (Figures 1 and 2) based on observations of particularly high counts during this period. These shorebird numbers do not persist later in the winter. Results from recent trials to compare bait uptake rates in July and September indicate faster bait disappearance in September 2018 than in July of 2018 arguing for the earlier application period from an efficacy standpoint because even at the highest labelled application rate we couldn’t maintain bait on the ground for more than 2 days in amounts sufficient to ensure bait availability to all mice. In 2018 ducks continued to breed during the month of September with young broods still appearing so the advantage of waiting till ducks were no longer breeding to improve detection and capture probabilities was not evident this year. The proposed date for first bait application is July 1, 2019. Non-target species protection activities will be timed to reflect this starting time. We will start capturing Laysan ducks in February and continue throughout the spring and summer. We will start capturing migratory shorebirds in June as soon as all birds that are going to migrate North have done so. Captive ducks and shorebirds will be kept in aviaries on Sand Island until immediately before the July application period. Then they will be moved to corresponding aviaries on Eastern Island where they will be kept until it is deemed safe to release them using a four-step process described later in this report. We will also capture and hold in captivity on Sand Island canaries and common mynas. These birds will be released in test groups after there are no more pellets visible and residue testing indicates levels of Brodifacoum residue in invertebrates have declined sufficiently. We will track released passerines carefully and also observe the health of returning migratory shorebirds for signs that they are being exposed including lethargy and bleeding. Ducks will not be released until all these other indicators of Brodifacoum toxicity are no longer being seen. Staffing and infrastructure requirements of each of the non-target organisms to be held in captivity will be discussed in detail below. The total number of non-target care personnel needed are shown in Table 2.

7

SUMMARY OF RISKS AND PROPOSED PROTECTION PLAN FOR LAYSAN DUCK AND SHOREBIRDS FOR THE MIDWAY SEABIRD PROTECTION PROJECT PROJECT INTRODUCTION

Table 2. The 2019 staffing schedule for non-target species management for capture and captive care of birds. V = Wildlife veterinarian, A = Aviculturist, C = Capture specialist and, P = Protection team members. Protection team members will assist with captive care tasks including food preparation, feeding aviary birds, cleaning aviaries, observation of marked birds on Eastern, supplemental feeding on Eastern, Guzzler cleaning, surveillance on Sand, and other duties.

First Bait application Last Bait application Jan Feb Mar Apl May June July Aug Sep Oct Nov Dec All captive V V V V V V V V V V V care A A A A A A A A A A A Laysan C - 3 C-3 C-4 C-4 C-4 C-4 C-4 C-3 C-2 C-2 C-2 Duck P - 3 P-3 P-4 P-4 P-4 P-4 P-4 P-4 P-4 P-4 P-4 Shorebirds C-2 C-2 C-2 C-2 C-2 C-2 P-2 P-2 P-2 P-2 P-2 P-2 P-2 P-2 Sentinel C-2 C-2 C-2 P P P P P Passerines P P P Total NT 8 8 13 17 17 15 15 14 13 11 10 staff

8

SUMMARY OF RISKS AND PROPOSED PROTECTION PLAN FOR LAYSAN DUCK AND SHOREBIRDS FOR THE MIDWAY SEABIRD PROTECTION PROJECT BENEFITS OF MOUSE ERADICATION

2. BENEFITS OF MOUSE ERADICATION TO LAYSAN DUCKS AND BRISTLE-THIGHED CURLEW Extirpation of the Laysan duck from the main Hawaiian Islands most likely was caused by a combination of predation by introduced mammals, especially rats, hunting by humans, and habitat destruction and degradation. Current threats in the Northwestern Hawaiian Islands include non- native, invasive species (e.g., mice, and plants) that can alter the Laysan duck’s habitat. Since the Laysan duck is also a ground nesting species, the removal of mice is expected to contribute to successfully maintaining Laysan ducks on MANWR. If mice are able to prey upon adult Laysan albatrosses, one of the largest seabirds on Sand Island, then other ground nesting and burrowing seabirds, adults and chicks, are at risk from similar impacts. Without eradication, introduced mice would continue to prey on ground nesting birds on the island, preventing them from reaching their full population potentials, and if the predatory behavior spreads at the rate observed in 2016/2017, then the predation would likely contribute to accelerated declines in affected populations. Mice could also be competing with ducks for invertebrate and plant food resources. Mice are known to be a seed predator on Eragrostis variabilis an important food source and plant cover for Laysan ducks. Therefore, it is possible that food will be less limiting for Laysan ducks and that an increase in nesting cover could occur after mice are eradicated from Sand Island. Diet overlap of Laysan ducks with mice is currently being studied on Midway using DNA barcoding techniques. Some researchers also believe that bristle-thighed curlews are being negatively affected by introduced mammalian predators including rats on their wintering grounds and atolls (Marks et al. 1990).

9

SUMMARY OF RISKS AND PROPOSED PROTECTION PLAN FOR LAYSAN DUCK AND SHOREBIRDS FOR THE MIDWAY SEABIRD PROTECTION PROJECT SUMMARY OF AVIAN CAPTURE METHODS

3. SUMMARY OF AVIAN CAPTURE METHODS Many of the methods used for capturing both Laysan ducks and bristle-thighed curlews are similar. This section addresses trapping methods that can be, and have been used for, capturing ducks and shorebirds. Specific details on trapping Laysan ducks and bristle-thighed curlews are found in each species’ section. Capture of passerine birds (canaries and mynas) involves some of the same techniques as well. Noose Carpet: A noose carpet at least 30 x 30 cm (Bub 1991) can be made of 0.24 x 0.24 in. (0.6 x 0.6 cm) hardware cloth and 20 lbs. (9 kg) monofilament nooses. A clinch knot can be used to secure each noose to the hardware cloth. Birds are coaxed to walk across the noose carpet by herding them towards it (Figure 4). Noose carpets are designed to typically catch one bird at a time. Traps set near wetlands must be weighted or anchored to prevent captured ducks from dragging the noose carpet into a deep wetland and diving to the bottom where they could drown.

Figure 4. USGS biologist Michelle Reynolds herds ducks towards a noose carpet, a trap used to catch many of the 28 ducks translocated from Midway Atoll National Wildlife Refuge to Kure Atoll State Wildlife Sanctuary within Papahānaumokuākea Marine National Monument on September 3, 2014. Walk-in Trap: A walk-in trap such as those using a funnel fence of net walls with a tapering lead where herded ducks enter traps (or transport boxes) (Bub 1991). Decoys placed on the outside with bait on the inside is used to attract birds. A mist net can also be set up as a walk-in trap (SWCA 2017). This method is effective and may cause less stress than some of the other techniques. This kind of trap can be pre-baited and left open to allow birds to habituate to its presence and to learn to come for food there. Mist Net: Mist nets can be made with black polyester shelves 8.5 ft. high and 19.7-29.5 ft. wide (2.6 m high and 6-9 m wide) for capturing birds. Net poles can be made from 1 x 118 in. (0.0254 x 3 m) PVC-pipe or obtained from Avinet and sized 0.63 x 118 in. (0.016 x 3 m). Attempts to attract birds can be with audio playback of vocalization, herding, bait, and decoys. Mist nets can also be setup in a triangular formation to act as a walk-in trap. Net Gun: A CO2-powered Super Talon net gun with a 49 ft. (15 m) range can be used to capture birds. Each net gun can have four heads, each loaded with a 9.8 ft. (3 m) diameter net. Three types

10

SUMMARY OF RISKS AND PROPOSED PROTECTION PLAN FOR LAYSAN DUCK AND SHOREBIRDS FOR THE MIDWAY SEABIRD PROTECTION PROJECT SUMMARY OF AVIAN CAPTURE METHODS of nets are available: removable weights, permanent weights, and permanent weights with grappling hooks. A Crossman CO2 cartridge is used to propel the net. To capture birds with the net gun, a blind should be set up with decoys and a lure to attract birds to the blind. Whoosh Net: A whoosh net uses four pieces of 0.63 x 24 in. (0.016 x 0.6 m) rebar to anchor a net and bungee cords, two pieces 0.34 x 59 in. (0.01 x 1.5 m) rebar with 1 x 59 in. (0.0254 x 1.5 m) PVC-pipe slid over the top for the lead poles, 1.3 x 36 in. (3.2 x 91 cm) punched angle for securing the trigger, and a pull cord made of parachute cord tied to a 0.24 x 1 in. (0.6 x 2.54 cm) metal pin. Decoys can be placed to attract birds to the path of the whoosh net. Dip Net and spotlight: Both shorebirds and ducks can be captured at night by dazzling them with a bright light and gently putting a dip net over them. A variety of traps have been used to catch mynas, some relying on the birds’ attraction to baits of various kinds (called ‘foraging traps’ in Australia) and others using live decoy birds. In practice, baited multi-catch traps become decoy traps as soon as the first myna is caught (https://www.cabi.org/ISC/search?q=Acridotheres%20tristis). At present not enough is known to predict which traps will be most successful in a given area. Different trap designs seem to vary in efficiency in different places, including different islands within the same group (Feare 2010) and different habitats on the same island (Feare, unpublished). On North Island, Seychelles, commercially available multiple catch traps failed to catch mynas feeding on the island’s grassland but were successful close to the island’s organic waste site, while locally-made decoy traps, each with four catching compartments with a drop door triggered on entry, were most efficient on the grassland (Feare, unpublished). Common mynas can sometimes be trapped on the nest using nooses inside the nest or in the nest entrance. Where this has been used the number caught has been very small compared with the numbers caught by more traditional trapping away from the nest (Millet et al. 2004, Tideman et al. 2007). Canaries can be caught using food-baited walk in traps, sparrow traps with live decoys, and mist nets. All nets and traps that are set would be monitored continuously when deployed.

11

SUMMARY OF RISKS AND PROPOSED PROTECTION PLAN FOR LAYSAN DUCK AND SHOREBIRDS FOR THE MIDWAY SEABIRD PROTECTION PROJECT LAYSAN DUCK

4. LAYSAN DUCK

4.1 LAYSAN DUCK MITIGATION PLAN The Laysan duck is endemic to Hawai`i and is considered the rarest native waterfowl species in the United States. It currently has the most restricted range of any duck in the world (USFWS 2009). In 2004 and 2005, a total of 42 ducks were translocated from Laysan Island to the 1,128 ac. (456 ha) Sand Island, MANWR, following extensive habitat restoration efforts (Reynolds and Klavitter 2006). This population peaked in 2010 at 661 adult and juvenile birds (95% CI 608–714), before the botulism outbreaks that caused population declines in 2012 and 2015 of 38 percent and 37 percent, respectively. In 2015, using a statistical model developed from a radio-telemetry study (Reynolds et al. 2014), Reynolds estimated there were between 314 and 435 Laysan ducks (95% CI for population estimate). The point estimate was 375 individuals (Reynolds et al. 2017). This estimate of Laysan ducks at MANWR is approximately 50% of the global population. In comparison, the most recent estimate on Laysan Island in 2012 was 339 individuals (95% CI: 265– 413). The same statistical model developed by Reynolds was then used to estimate the population of Laysan ducks on MANWR using data from March 2017 to March 2018 that included numbers from the non-breeding season (October-February). Using a maximum count of 372 ducks, the model returned a population estimate of 600 ducks with a range of 526-685 (95% CI) individuals (K. Goodale, Pers. Comm.). Laysan ducks fly between Eastern and Sand Island but it is unknown how often birds move back and forth or whether there are birds that do not cross the water barrier and stay on one island of the other. The Laysan duck (Anas laysanensis) is currently on the Federal Endangered Species List due to its restricted distribution and small wild population. It is a small-sized duck, largely nocturnal and sedentary. Adults are approximately 15-17 in. (38-43 cm) in length and weigh around 15-18 oz. (420-500 g). Overall, it is a chocolate-colored bird with contrasting bi-colored body feathers, an iridescent purplish-green speculum (wing patch), and a prominent white eye ring (USFWS 2009, Moulton and Marshall 1996). Legs and feet are pale dull orange and usually brighter in males. On Laysan Island, courtship behaviors occur most of the year and adult pairing is established by September or October. The nesting season on Laysan generally runs from April through July, but reproductive response is flexible according to habitat conditions (Moulton and Marshall 1996). Laysan ducks can produce clutches averaging 3.8 eggs; the start and duration of egg-laying is highly variable from year to year depending upon stochastic factors like climatic events. Egg- laying typically occurs from April to August and incubation is 28-29 days. Duckling activities are concentrated near sources of fresh water with nearby food and cover. At MANWR, the peak nesting season is March through September. Males start molting in June. Most adult females start molting in August and September. Wing molt takes approximately 3-4 weeks, a pattern typical of mainland ducks (Moulton and Weller 1984). Broods receive female- only parental care (Moulton and Weller 1984, Reynolds and Kozar 2000). If the timing of the Proposed Action occurs in July, it will overlap with the Laysan duck breeding season (March

12

SUMMARY OF RISKS AND PROPOSED PROTECTION PLAN FOR LAYSAN DUCK AND SHOREBIRDS FOR THE MIDWAY SEABIRD PROTECTION PROJECT LAYSAN DUCK through September). The only effective mitigation strategy is to prevent the exposure of the ducks to rodenticide.

4.2 PROPOSED LAYSAN DUCK PROTECTION PLAN Captive care facilities (aviaries) with a minimum capacity of 200 Laysan ducks and the capability to separate smaller groups of animals in ways that minimizes aggression will be built both on Sand and on Eastern Islands. Two hundred ducks will be held in the Sand Island aviaries from their date of capture until immediately before the bait application when they will be transferred to the Eastern Island aviaries. There will also be clinic space inside a building on Sand Island allocated for more intensive care of ducks that need veterinary care and aviary space for holding ducks for at least 10 days in order to administer a vaccine for botulism and a booster shot 10 days later for all ducks whether they are held or just captured and subsequently released on Eastern Island. Capture operations will commence in February and the first set of actions would include the capture of 200 Laysan ducks (40% males and 60% females, and 60% adults). Ducks in excess of 200 will be held for 10 days for 2 doses of botulism vaccine, have their flight feathers clipped, and then released on Eastern Island where they will have access to supplementary water at pre-placed guzzlers, supplementary food spread broadly to eliminate excessive food guarding, and artificial cover structures for shade and rain protection. For a July bait application, capture operations would start in February and continue until all pellets have disappeared. If hens and ducklings are detected we will attempt to capture them together and keep them in a separate compartment. If a hen on a clutch of eggs is located, we will place her in an individual compartment and see if she will incubate the displaced clutch. If not, we will use an incubator for duck eggs and raise the ducklings by hand. As before, ducks captured after the first 200 will be vaccinated twice 10 days apart and then released on Eastern Island and provided supplemental food, water, and shelter. The 200 captive birds on Sand will be held until immediately before the bait application when they will be moved to the Eastern Island aviaries to minimize the chance of them accidentally ingesting pellets or insects that have fed on the bait and to reduce the risk of compromising the eradication operation by allowing access to alternate food sources for the mice. Previous food trials indicated that conversion of Laysan ducks to a captive diet on commercial feed is possible (SWCA 2017). Based on a habitat assessment, Eastern Island has a carrying capacity to currently support about 85 to 102 Laysan ducks (SWCA 2017). To increase the carrying capacity of Eastern Island, Laysan duck habitat on Eastern Island will be enhanced by adding vegetative and artificial cover, guzzlers for freshwater resources (Figure 5), and supplemental feeding stations, spread widely enough to preclude exclusion of subordinate ducks by more assertive individuals. These enhancements would increase the carrying capacity of Eastern Islands to approximately 400 ducks. The proposed action thus involves two redundant levels of protection (ducks held in aviaries and ducks released on Eastern Island) to ensure the survival of as many healthy individuals (~600) as possible. By maintaining the capacity to care for ducks both in an aviary setting and by clipping flight feathers and releasing birds on Eastern, the team can respond to situations in which birds are doing poorly in captivity by wing-clipping them and releasing them on Eastern or bring birds into the aviaries that aren’t competing well out on Eastern Island for more intensive care in captivity. As stated, we

13

SUMMARY OF RISKS AND PROPOSED PROTECTION PLAN FOR LAYSAN DUCK AND SHOREBIRDS FOR THE MIDWAY SEABIRD PROTECTION PROJECT LAYSAN DUCK plan to have the capacity to hold a minimum of 200 ducks in aviaries along with an avian care facility with room for 25 sick or injured birds needing more individualized treatment.

Figure 5. Guzzler and surrounding vegetative cover at Green Island, Kure Atoll.

4.3 INFRASTRUCTURE DEVELOPMENT AND PERSONNEL

INFRASTRUCTURE DEVELOPMENT The infrastructure needed to safely hold wild Laysan ducks in captivity for an extended period (>5 months) is extensive. Eastern Island has the habitat capacity to support about 85-102 Laysan ducks that would be wing-clipped, brought from Sand Island, and released wild. However, with the additional water guzzlers, cover and shade structures, and food that will be provided on Eastern (see Section 4.3.1.2 below), the habitat capacity will be much higher (~400), and the remaining population of ducks not kept in aviaries can be supported for the duration of the eradication action. Further, ducks released on Eastern Island would have to be monitored and recaptured for a second wing-clipping, if needed, to prevent them from flying back to Sand Island. Infrastructure needed on Eastern Island for the additional translocated ducks would include shade structures and supplemental feeding stations and water sources. While aviaries can be built on Eastern Island, the island can only be accessed by boat and has limited resources. Therefore, any infrastructure built and maintained on Eastern Island needs to consider conducting operations in terms of weather

14

SUMMARY OF RISKS AND PROPOSED PROTECTION PLAN FOR LAYSAN DUCK AND SHOREBIRDS FOR THE MIDWAY SEABIRD PROTECTION PROJECT LAYSAN DUCK conditions and shuttling efforts by boat across the channel, in addition to all supplies (e.g., freshwater, food, and equipment) being brought over on a regular basis or stored on-island. Bird care specialists will need to camp on a rotating basis at Eastern to ensure that weather and sea conditions don’t prevent access for personnel to the birds in captive care. Other infrastructure is necessary to support captive holding of birds. Work areas will be needed for administering veterinary care, preparing food, storing food items and utensils, and storing cleaning supplies. In general, aviary sites on both islands will be located where there is adequate air flow, shade, and a low amount of noise. In addition to adequate aviary space, each aviary will be fitted with 5mm thick rubber flooring that prevents ducks from developing foot and leg issues (see Section 4.4.1). Aviaries will be located on cement pads to avoid problems with burrowing birds. Eastern Island aviaries will have fewer burrowing bird problems but will also be placed on available concrete pads. Further, visual barriers (see Section 4.4.1), and rodent control efforts are also important aspects to address in terms of aviary design and maintenance. With Laysan ducks, the aggressive behavior displayed among captive ducks over food might be reduced if aviaries are built to compartmentalize ducks into smaller groups (Warner 1963, SWCA 2017). We plan to compartmentalize ducks into groups of 10 birds. If our groupings of 10 ducks prove to be too large, we will have materials on hand to further subdivide the compartments on Sand Island.

Sand Island Infrastructure on Sand Island will include:

 Aviary structures to house at least 200 ducks (40% males and 60% females, and 60% adults): The large shade house on Sand Island (52 x 96 ft [15.8 x 29 m]) will be partitioned into 16 compartments for 10 ducks each to hold 160 ducks with 31.2 ft2 (2.9 m2) per duck. An additional aviary with 10 compartments of 100 ft2 (9.3 m2) each will hold the remaining 40 ducks in groups of 5 with 25 ft2 (2.3 m2) per duck and have extra compartments for any hens with ducklings to be isolated. We will keep males and females separate to avoid aggression and breeding. Total aviary area for ducks will need to be at least 5,000 ft2 (464 m2) on each island (Sand and Eastern) and there will be enough compartments so there are not more than 10 ducks in any particular section to reduce chances for food guarding. Visual barriers made of 80% shade will be installed. Aluminum flashing will be used around aviaries to prevent mice from chewing into shade cloth and entering aviaries. All ducks will be moved to Eastern Island before the rodenticide drop.  Work areas for duck food prep and storage and cleaning supplies.  Avian care facility with room for 25 birds needing more individualized treatment, for sick or injured birds and those suspected of being exposed to rodenticide bait.

Eastern Island Ducks with clipped flight feathers will have access to 11 water guzzlers distributed around the island and feeding stations adequate to allow access for even the most timid birds. There will also be additional cover stations established. Four aviaries, each with a capacity of at least 50 ducks will be constructed (Figure 6) on existing concrete pads near the pier and boat access, making

15

SUMMARY OF RISKS AND PROPOSED PROTECTION PLAN FOR LAYSAN DUCK AND SHOREBIRDS FOR THE MIDWAY SEABIRD PROTECTION PROJECT LAYSAN DUCK aviary construction and captive husbandry of birds convenient. Rubber matting will be installed in 4 aviaries of 30 x 50 ft (9.1 x 15.2 m) each that have 10 compartments each.

Figure 6. Views from inside and outside one of the 8x14 ft aviary compartments that was used to hold ducks during the 2004 and 2005 translocation. The compartment had a sand floor, natural cover, and artificial cover structures. In summary, infrastructure on Eastern Island will include:

 Aviary structures to house at least 200 ducks  11 water guzzlers total (3 already in place)  11 shade structures  Seawater pump system to have additional water available for cleaning aviaries  Camping facilities for up to 10 aviculturists and duck monitors when bad weather precludes traveling between Sand and Eastern Islands  Work areas for duck food prep and storage and cleaning supplies

Eastern Island Freshwater Infrastructure On Eastern Island, three 500 gal. (1,893 L) water guzzlers are currently placed close to freshwater wetlands (seeps) and installed on top of revetment mounds to prevent potential overflow from tsunamis. These water guzzlers are kept close to the existing wetland seeps to reduce the amount of time taken to search for ducks with botulism. Soil on the revetment mounds are loosely compact and can easily be dug by hand without the need to land heavy machinery on Eastern Island. Areas where water guzzlers are installed should have adequate cover habitat for Laysan ducks and therefore, some additional vegetation may need to be planted. The guzzlers and cover plantings will be installed as early as possible to allow catchment reservoirs to fill during the winter and plants to establish during rainier weather. For Sand Island ducks released on Eastern Island, 8 additional guzzlers to the three already in place will be installed. Given the current suitable habitat on the island, it is highly probable that ducks will congregate around freshwater sources. Crowding around freshwater sources increases the risk

16

SUMMARY OF RISKS AND PROPOSED PROTECTION PLAN FOR LAYSAN DUCK AND SHOREBIRDS FOR THE MIDWAY SEABIRD PROTECTION PROJECT LAYSAN DUCK of potential disease outbreak, with botulism being the primary concern, and increases the likelihood of female mortality from forced copulations and duckling mortality from aggressive adults. The three existing seep ponds on Eastern will be filled with sand or covered with shade cloth to prevent ducks from accessing the water to reduce conditions that foster botulism outbreaks. Fresh water for captive ducks and camping duck caretakers will be brought from Sand Island periodically.

PERSONNEL People are needed for infrastructure development, bird capture and captive care, bird health and veterinary care, behavioral monitoring, food preparations, aviary construction, maintenance and cleaning, and post-project monitoring. Personnel needs vary from construction and maintenance of aviaries to specifically-trained (and federally permitted) ornithologists, avian rehabilitators, aviculturists, and licensed veterinarians. A veterinarian will be on site for the duration of the capture period, bait application period, and post application holding period and staff will be ready to handle botulism outbreaks and accidental rodenticide poisonings. Details of how ducks with botulism will be treated and cared for are outlined in Appendix E. Table 2 summarizes staffing needs for the care of each of 3 groups of birds.

CAPTURE METHODS AND MATERIALS This species becomes secretive and difficult to locate during the breeding season and capturing birds during the non-breeding season is preferred. Capture during the non-breeding season will also reduce risk to eggs and ducklings. For a July rodenticide application capturing Laysan ducks on Sand Island should start in February of the implementation year. For the bait drop to proceed, we will need to capture a minimum of 200 ducks for captive care and a minimum of 200 ducks for translocation to Eastern Island. Capture efforts will continue throughout the bait application and continue through implementation and as long as brodifacoum remains accessible in the ecosystem in harmful concentrations. Any ducks not captured for holding that are subsequently exposed to the rodenticide, and demonstrating signs of toxicosis, would be captured and treated with the antidote Vitamin K by a veterinary professional to offset the negative effects of the rodenticide. Those ducks showing signs of exposure would be held in temporary holding pens under observation and any decision-making would occur based on the duck’s on-going health prognosis. In the event that nests or very young broods are found, we would capture the adult, and place the nest and eggs into a dog pet carrier (twice the size of a cat carrier) and place them inside one of the aviary compartments. If ducks with small ducklings are found, we will capture adult and young ducklings which can’t thermoregulate, and then place them inside a smaller, private aviary compartment. SWCA (2017) tested methods on Sand Island to increase capture efficiency of Laysan ducks. Trials were conducted to develop alternative trapping techniques for capturing multiple ducks at one time. For trials, construction fencing was used to make a drift fence, chute, and holding pen. The height of the trap was approximately 16 in. (41 cm), the drift fence was 6 ft. (1.8 m) long, and the chute was 8 ft. (2.4 m) long. The holding area was “V”-shaped and 5 ft. (1.5 m) long with an

17

SUMMARY OF RISKS AND PROPOSED PROTECTION PLAN FOR LAYSAN DUCK AND SHOREBIRDS FOR THE MIDWAY SEABIRD PROTECTION PROJECT LAYSAN DUCK entrance of 5 ft. (1.5 m) wide and end width of 5 ft. (1.5 m) (Figure 7). The holding area was covered with a black plastic tarp and had a capture box (also known as a transport box) attached (Figure 8). The transport boxes were the same boxes used to transport Laysan duck during the 2004-2005 translocations (Figure 9).

Figure 7. Layout of Laysan duck trap. The covering for the holding area is not included to show the design.

Figure 8. Laysan duck trap with the covering on. During the trials, Laysan ducks did not hesitate when herded into the trap. The ducks were reluctant to enter pet carriers, which did not allow light through, but they did enter the transport boxes,

18

SUMMARY OF RISKS AND PROPOSED PROTECTION PLAN FOR LAYSAN DUCK AND SHOREBIRDS FOR THE MIDWAY SEABIRD PROTECTION PROJECT LAYSAN DUCK which allowed light to pass thorough (Figure 9). These traps could be built to scale and could be used to catch up to 10 to 15 ducks at one time by attaching multiple transport boxes to the trap. This trapping method has the potential to be dangerous to ducks if experienced duck handlers are not present during trapping operations. Duck trappers should have experience handling endangered species, identifying signs of avian stress, capture myopathy, and injury and capturing 10 or more waterfowl at one time. Trauma and injury to individual ducks by trampling may occur if the holding area is not large enough to support the number of ducks being herded into it.

Figure 9. The transport box allows light to enter through both sides.

4.4 CAPTIVE HOLDING AND TRANSPORT

AVIARY STRUCTURES Aviaries will be built at least one month prior to the capturing of ducks and sites prepared by excluding albatrosses from nesting there prior to when albatross nesting begins (to avoid having to move albatross nests to erect aviaries). To hold at least 200 ducks comfortably, with room for an additional 50 animals if there is a botulism outbreak, 10 aviaries will be constructed each on Sand Island and Eastern Island. An additional aviary would be constructed for ducklings too small to compete well in aviaries holding adults. Aviaries, each measuring ~500 ft2 (46.5 m2) with five, 100 ft2 (9.3 m2) compartments, will be built using galvanized steel poles and shade cloth. Substrate or flooring that discourages burrowing seabirds and crabs from digging under and into the aviary and prevents ducks from developing foot and leg issues will be used. This will entail siting the aviaries on cement pads. Ducks will be housed in aviaries on 0.2 in (5 mm) thick rubber floor pads with localized feeding stations that can be easily cleaned. A-frame hutches will be provided for cover as needed and aviaries will have shade cloth. Bathing tub, water bowls, and food bowls will be placed in each aviary compartment (Figure 10). Outdoor aviaries in well ventilated areas (not inside buildings) will prevent fungal respiratory diseases. Automated

19

SUMMARY OF RISKS AND PROPOSED PROTECTION PLAN FOR LAYSAN DUCK AND SHOREBIRDS FOR THE MIDWAY SEABIRD PROTECTION PROJECT LAYSAN DUCK weighing platforms will be installed in each section of the aviaries to allow for close monitoring of nutritional status without the disturbance of handling the birds to weigh them. For both the Sand Island and Eastern Island aviary facilities we will have sufficient running water to clean feeding stations, watering devices, and floors daily. Visual Barriers will be installed and the aviary facility designed to aid in behavior management. Two layers of 80% shade cloth (24 in. [61 cm] height) will be attached to each aviary compartment divider wall to serve as a visual barrier to reduce stress between ducks held in each compartment. Closed–circuit cameras will be installed to monitor behavior of the ducks without humans being visible. Aviaries will be designed to keep group size small enough to be able to detect food exclusion and fighting problems.

Figure 10. Natural cover, artificial cover, food bowls, bath tubs, and waterer inside a 2004 Laysan duck aviary compartment.

Predator Control near Sand Island Aviaries To reduce the risk of predation by mice, bait stations containing peanut butter flavored sticky traps and multi-catch live traps will be placed every 10 feet (3 m) around the perimeter of each aviary. Sticky traps will be checked several times a week and sticky traps replaced as needed. The bottom outside edge of the aviary will be covered with a 2-ft tall (0.61 m) aluminum sheeting to prevent mice from chewing through the shade cloth to enter the aviary. Biosecurity measures for transport of gear, supplies, and personnel between Sand and Eastern will be stringent to prevent the accidental introduction of mice to Eastern Island. Protocols will include mouse-proof containers, rigorous inspections, bait stations at both boat landings, and mouse detection gear deployed on Eastern Island.

20

SUMMARY OF RISKS AND PROPOSED PROTECTION PLAN FOR LAYSAN DUCK AND SHOREBIRDS FOR THE MIDWAY SEABIRD PROTECTION PROJECT LAYSAN DUCK

4.4.2 HOLDING AND TRANSPORT BOXES Commercially available cat carriers 19 x 12 x 10 in. (48 x 30 x 25 cm [LxWxH]) will be used to hold sick birds receiving veterinary care or rehabilitating as well as to transport ducks during initial capture attempts and to shuttle them (Figures 11 and 12) to Eastern Island aviaries.

Figure 11. USCG staff offload Laysan ducks to USGS Biologist Michelle Reynolds on Kure Atoll during a major translocation to move 28 Laysan ducks from Midway Atoll National Wildlife Refuge to Kure Atoll State Wildlife Sanctuary within Papahanaumokuakea Marine National Monument September 2014. Photo by: John Klaviteer/USFWS.

Figure 12. USCG ENS Kane holds the door of a transport box open as USFWS John Klavitter places a duck inside its transport cage in preparation for translocating 28 Laysan ducks from Midway Atoll National Wildlife Refuge to Kure Atoll State Wildlife Sanctuary in 2014. Photo by: Eric Dale/FWS Volunteer.

21

SUMMARY OF RISKS AND PROPOSED PROTECTION PLAN FOR LAYSAN DUCK AND SHOREBIRDS FOR THE MIDWAY SEABIRD PROTECTION PROJECT LAYSAN DUCK

4.5 CAPTIVE CARE

EXAMINATION AND TREATMENT Each duck brought into captive care will be examined for signs of botulism or other conditions and birds needing extra care or treatment assigned to the clinic. Laysan ducks are susceptible to avian botulism, and a prevention plan involving botulism vaccine and a response plan will be in place to care for sick ducks and management of outbreaks (Appendix D and E). Botulism antitoxin will be kept on-site under refrigeration for administering to sick ducks exhibiting symptoms of botulism toxicity including:  Inability to fly or poor flight  Depressed behavior, weakness  Labored breathing  Diarrhea and soiled vent  Paralysis of nictitating membrane  Inability to walk  Inability to hold up neck A captive care protocol for treating ducks affected with botulism can be found in Appendix E. A botulism vaccine will be administered to all captured ducks to prevent a major outbreak among the captive flock. Enough captive care facilities sufficient to keep all ducks handled in captivity long enough for 2 doses of vaccine 10 days apart will be constructed. Then, birds will join the captive flock of 200 birds that will be housed on Sand Island until July 1 or go immediately to Eastern Island with flight feathers clipped. Once rodenticide applications have commenced, Vitamin K will also be kept in ample amounts with trained staff on-site to administer at first signs of poisoning from brodifacoum. Ducks exposed to brodifacoum will display a hunched posture, puffed up wings, be moribund, or show blood from nares. The vaccine, anti-toxin, and vitamin K will be stored in the work areas built near the clinic aviary. If duck mortality (take) exceed thresholds determined by formal consultation, we will reopen consultation to discuss and determine a best course of action.

FEEDING AND MANAGING SOCIAL INTERACTIONS When kept together, food aggression and food guarding behavior amongst Laysan ducks was the leading cause of starvation-related mortality (Warner 1963). Measures to reduce this behavior will include supplying the group with multiple food containers and larger food containers or troughs, putting up visual barriers between food containers, and designing the aviary to house groups in low numbers together (SWCA 2017). Captive animals will be monitored using closed-circuit cameras to detect dangerous interactions so animal groupings can be adjusted. Automated weighing perches will also be installed to detect weight loss in birds suspected of not getting adequate access to food. Likewise, for the higher than normal densities of ducks living freely on Eastern Island after clipping of their flight feathers, extra food distribution will be managed to minimize the possibility of dominant individuals precluding access to food or water for subordinate individuals. In addition, during the breeding season, holding ducks in all male and all female

22

SUMMARY OF RISKS AND PROPOSED PROTECTION PLAN FOR LAYSAN DUCK AND SHOREBIRDS FOR THE MIDWAY SEABIRD PROTECTION PROJECT LAYSAN DUCK groupings and separating out first year birds may reduce negative interactions and prevent males from competing for females. Reynolds and Klavitter (2006) used commercial duck mash during the 2004-2005 translocation effort and noted that mash used to feed captive ducks in aviaries should be very fresh and should not be stored for more than one month, as birds find older mash less appealing. To increase the caloric input of the mash, dry cat feed may also be added. In feeding trials conducted by SWCA (2017), the combination of fly larvae, cooked white rice and duck feed was the most successful in encouraging food-take from a container and was also the most consumed in the vicinity of the container. Food presented in water was much more likely to be consumed by ducks than food presented without water. Adults without ducklings seemed to display more foraging behavior and showed more interest in the food items than adults with ducklings. Ducks already displaying foraging behavior before encountering the food items were more likely to consume food. Freeze- dried mealworms are taken well by some ducks and captive ducks are more easily converted to accepting food if there is live food incorporated in the ration. All potential live food sources will be carefully evaluated with reference to biosecurity.

4.6 RELEASE

CARCASS AND BAIT MONITORING BEFORE AND AFTER RELEASE Island-wide monitoring of bait presence, mouse sign, and carcasses will be conducted, and starting 2 months before the first bait drop, known duck resting and foraging sites will be searched for signs of any ducks until no visible bait pellets remain after the final application of the bait. Sites will be searched daily. All carcasses found are to be recorded, and carcasses suitable for necropsy will be collected, labeled, frozen and submitted to U.S. Department of Agriculture for analysis. Resting and foraging sites will be monitored before the first bait application, during and up to 42 days after the last aerial broadcast of bait or until 30 days after bait no longer persists in the environment, whichever is longer. To quantify bait persistence post-eradication, a sample of 82 x 3.3 ft. (25 x 1 m) transects will be surveyed throughout the island. No ducks will be returned to Sand Island while there is any bait detectable anywhere that is accessible to ducks. In the event of a mouse detection more than 42 days after the last bait drop a localized eradication response may be undertaken and duck protection measures will be integrated into any actions.

RELEASE METHOD Ducks that have had their flight feathers clipped will remain on Eastern until they molt again. There is a possibility that some of these birds could molt early and have the ability to fly back to Sand Island. If they return to Sand Island and they become sick from brodifacoum toxicity, we will re-capture these birds, treat them with vitamin K, and when healthy again, return them to Eastern Island. Birds being held in the aviaries on Eastern will be kept there until there is confirmation through a four-step process to determine when it is safe to release ducks back to Sand Island. These four steps include: 1) there are no pellets remaining; 2) brodifacoum residue levels measured in selected invertebrates reach levels deemed safe; 3) migratory shore-birds returning

23

SUMMARY OF RISKS AND PROPOSED PROTECTION PLAN FOR LAYSAN DUCK AND SHOREBIRDS FOR THE MIDWAY SEABIRD PROTECTION PROJECT LAYSAN DUCK from their northern breeding grounds appear to be staying healthy upon arrival, and; 4) sentinel canaries and common mynas are released on Sand Island and show no ill effects. All birds on Sand will be observed for any signs of mortality or other signs of anticoagulant exposure (hunched posture, puffed up wings, blood from nares, moribund). Monitoring of released birds will include radio-tagging sentinel mynas.

Pre- and Post-project Monitoring We will monitor Laysan duck population size, reproductive performance, and behavior. We will quantify direct and indirect impacts of mice on duck population by the use of camera traps on nests and conducting measures of reproductive performance and foraging behavior before and after implementation of the mouse eradication effort. Diet overlap of Laysan ducks with mice is currently being studied using DNA barcoding techniques. After live-capture and holding ducks on Eastern Island during the operation, birds will be released after confirmation through a four-step process described above to determine when it is safe to release ducks back to Sand Island as outlined above. We are exploring the possibility of having an analytical lab set up on Sand Island to ensure quick feedback about samples. Monitoring will be done daily at all known duck aggregation sites and throughout the island. We will recapture all ducks if any are found demonstrating signs of toxicosis. Radio-transmitters will be attached to a sample of common myna birds (sentinels) to monitor their movements, behavior, breeding and survival when they are released. The birds will also be observed at a distance using spotting scopes to prevent disturbance. Invertebrate prey species of Laysan Ducks on Sand Island to be chosen after diet studies are complete (DNA barcoding) will be collected at weekly intervals after the final bait application in areas favored by foraging ducks and brodifacoum residues will be measured to assess when those residues go below values calculated to be dangerous to feeding ducks. To determine these residue levels, we used an approached proposed by Thomas et al. (2011) and improved on by Mineau (2014). This approach provides a relationship between brodifacoum liver residues and the probably of evident coagulopathy at necropsy (Thomas 2011) and uses data on the proportion of a single dose that is retained in the liver over time (Mineau 2014). These two sets of information allow for a calculation of a “field-derived LD50” or more accurately ED50 (dose at which half of the test population will show an effect – but not necessarily death) (Mineau 2018). However, to be protective at the individual level for a rare or threatened species like the Laysan duck, ED50 is not appropriate, and for this analysis we used an ED5 (5% of the exposed population will be affected by coagulopathy – not necessarily death). If we assume a 500g Laysan duck, its liver weight will be approximately 11g (after relationships in broiler chickens – Deaton et al. 1969). Using the approach described above, the calculated threshold whole body ED5 is 0.47-1.8 µg total intake of brodifacoum for an average Laysan Duck (Mineau 2018). Because Brodifacoum can bio-accumulate through multiple exposures, expressing the risk as a proportion of daily food intake is not appropriate (Mineau 2018). Instead, we used an

24

SUMMARY OF RISKS AND PROPOSED PROTECTION PLAN FOR LAYSAN DUCK AND SHOREBIRDS FOR THE MIDWAY SEABIRD PROTECTION PROJECT LAYSAN DUCK

‘accumulation window’ of about 200 days (the mid-point estimate for the half-life of brodifacoum in rat liver tissue) (Horak et al. 2018). Pitt et al. (2015) found that residues in cockroaches averaged 2.3 µg/g 10-15 days after bait drop, and 0.9 µg/g about 45 days later. It is reasonable to assume that, over time, a certain amount of brodifacoum will ‘escape’ the insect food chain to be lost to soil and the abiotic environment more generally. Fitting these two points to an exponential curve (a common curve type when describing residue degradation or loss from a system) and defining the peak residues as day 11 (even though it took 10-15 days for the residues to maximally contaminate the cockroaches) yields the following: Residue level (µg /g) = 2.3496 * exp (-0.0213 * days after peak residue) Predicting forward, cockroach residues at 200 days (our chosen period of residue accumulation) post peak would be approximately 0.03 µg/g. We calculated the cumulative residue intake over a 200-day period which produces a recommended prolonged withholding time for the ducks of 13- 15 months needed to drop the risk level below an ED5.

It is important to note that we are being very protective by choosing an ED5 when the outcome is coagulopathy rather than mortality, representing a conservative approach. It is also reasonable to assume that an increasing part of the duck’s diet over time will be made up of unexposed cockroaches and other invertebrates, so the suggested holding times described above may be over- estimated. Therefore, a prudent, and conservative, first release of Laysan ducks may be at the 4-6- months post bait application (after monitoring of other avian sentinels also indicates the environment is safe), with continued surveillance and monitoring of the ducks (i.e. if any negative effects are detected, adaptive management can be enacted to revert it). Furthermore, field application of this proposed model will require real time estimates of remaining residue levels in the Midway ecosystem. Sentinel diet sampling and analysis would be performed at intervals appropriate to inform the mitigation strategy. It will be important to look at the overall loss rate of residues from the food-chain in the early days, weeks and months following the bait application as to characterise and document the residue decline curves for Sand Island.

1 We are only interested in the declining phase of residues after peak 25

5. BRISTLE-THIGHED CURLEW AND OTHER SHOREBIRDS

5.1 BRISTLE-THIGHED CURLEW AND SHOREBIRD MITIGATION PLAN The bristle-thighed curlew is a medium-sized curlew with a wingspan of approximately 15.7-17.3 in. (40-44 cm) and blue-gray legs. The birds have a moderately long and decurved bill, which is flesh-colored at the base and turns to brown near the tip, and a dark lateral crown is present on the head along with eye stripes. Their name comes from the bristle-like extensions at the base of their legs, although these are generally inconspicuous. Females are heavier than males and have longer wings and a shorter bill. Mean weight of adults was 1.09 lbs. (497.4 g) and ranged from 0.84-1.29 lbs. (393-584 g) for 61 unsexed birds caught in the NWHI from October through November (Marks 1993). Juveniles are similar to adults except for the presence of larger cinnamon-buff spots on the upperparts, and virtually unstreaked underparts (Marks et al. 2002). The highest documented count of this species in the Action Area from 1993 to 2001 (using data from various avian surveys) was 242 birds, or 2.42% of the global population (USFWS unpublished). Surveys were also conducted in 2017 on Sand, Eastern and Spit Islands (Figure 13). It is not known how many of these individuals were winter residents versus migrants stopping before they headed further south. Detailed information on the Pacific golden plover, turnstone, and wandering tattler can be found in the Midway Seabird Protection Project Draft EA (USFWS 2017).

5.2 POTENTIAL MITIGATION OPTIONS It is anticipated that the bait application could sub-lethally affect or incidentally kill 5-52 bristle- thighed curlew on MANWR without mitigation along with a substantial proportion of the Pacific golden plover, turnstone, and wandering tattler at Sand Island (USFWS 2017). Numbers of birds that over-summer as well as the timing of northern migrants’ return vary greatly each year so a better estimate is not available for a July deployment. Mitigation options include, trapping and holding birds in captivity or clipping wings and transporting birds to Eastern Island and releasing them there along with using a combination of hazing and attraction techniques to lure more of the shorebirds to Eastern Island on their own. Without wing-clipping, individuals would be able to fly from Eastern Island back to Sand Island where they would be exposed to the bait. Birds that remain on Eastern Island during the mouse eradication would be unlikely to experience any primary or secondary exposure. In addition, there could be additional bristle-thighed curlew and other shorebirds at risk for as many as 6-9 months after the baiting operation due to secondary exposure from invertebrates and other food items containing brodifacoum residues. Combining the potential estimated mortality from the initial bait drop and over the subsequent year, the total estimated mortality of bristle-thighed curlew (n=123) from primary or secondary poisoning is 1.23% of the estimated global population (USFWS 2017). SUMMARY OF RISKS AND PROPOSED PROTECTION PLAN FOR LAYSAN DUCK AND SHOREBIRDS FOR THE MIDWAY SEABIRD PROTECTION PROJECT BRISTLE-THIGHED CURLEW AND OTHER SHOREBIRDS

Figure 13. Shorebird surveys for Sand Island (top) and Eastern and Spit Islands (bottom) in 2017. Live-trapping bristle-thighed curlews and other shorebirds after the adult birds have left for the breeding grounds at the end of May but prior to the eradication of mice and holding the birds in captivity on Sand Island, or releasing birds with clipped wings on Eastern Island, would eliminate the risk of primary and secondary rodenticide exposure until their molt when they would be able to fly back to Sand Island. Although feasible, it is unlikely that a significant portion of the summer population, the majority of which are likely sub-adult birds (1-3 years old), could be captured before the initial bait drop. Before the bait drop on Palmyra in 2011 for the rat eradication, biologists were able to trap 5-9% of the bristle-thighed curlew population present (USFWS 2011). If this same rate of capture were possible on MANWR, approximately 12 to 22 birds could be captured and held in captivity or released with clipped wings on Eastern Island. However, there would also be additional risks associated with capturing birds, including physical injury and physiological stress that could result in the death of some individuals. Thirteen bristle-thighed

27

SUMMARY OF RISKS AND PROPOSED PROTECTION PLAN FOR LAYSAN DUCK AND SHOREBIRDS FOR THE MIDWAY SEABIRD PROTECTION PROJECT BRISTLE-THIGHED CURLEW AND OTHER SHOREBIRDS curlews were caught and successfully held in captivity on Palmyra without any injuries or mortalities reported, but several different trapping methods would need to be employed on MANWR to maximize success, as bristle-thighed curlew tend to be extremely wary. Capture efforts would continue throughout implementation and as birds returned from the breeding grounds. During the 2011 rat eradication at Palmyra Atoll National Wildlife Refuge (USFWS 2011), 13 bristle-thighed curlews were captured using a net gun and brought into captivity. To reduce stress and the risk of injury or capture myopathy, curlews were taken out of the net immediately, their heads were covered and the sedative Midazolam 7.05-5 oz./35 oz. (2mg/kg) was administered through the nasal cavity. The birds were then placed in a holding box for 5 minutes to allow the Midazolam to take effect. Curlews were held in 18 × 16 × 16 in. (46 × 40 × 40 cm) boxes and tube-fed 0.5 to 0.7 oz. (15 to 20 mm) of Oxbow-brand Carnivore Critical Care twice a day until they ate on their own twice a day. After tube feeding, the birds’ crops were packed by hand with egg balls made of hardboiled egg yolk and cat food to reduce weight loss. Water was applied by syringe to the culmen to moisten the throat after each egg ball was consumed. Once curlews were eating on their own they were released into aviaries. Aviaries measured 100 ft2 (9 m2) with 8 ft. (2.4 m) ceilings that held two curlews each. Feeding methods to be employed for captured curlews are summarized in section 5.6.2.

5.3 INFRASTRUCTURE DEVELOPMENT AND PERSONNEL The same types of veterinary and avicultural expertise required for the Laysan duck team will be necessary for captive care of shorebirds. People will be needed for infrastructure development, bird capture and captive care, bird health and veterinary care, aviary maintenance and cleaning, release, and post-project monitoring. Personnel needs will vary from construction and maintenance of aviaries to specifically-trained (and federally permitted) ornithologists and licensed veterinarians. It is estimated that approximately 4 personnel in addition to the veterinarian and specialist will be needed to handle these activities. See Table 2 for schedule and level of staffing proposed for shorebird capture and care.

5.4 CAPTURE METHODS AND MATERIALS USFWS survey data were reviewed to determine potential capture sites and strategies for bristle- thighed curlews (SWCA 2017). Surveys for curlews were made to determine the likelihood of success for potential capture strategies. Additionally, an atoll-wide survey of shorebirds was conducted on June 23, 2017 to determine the species and numbers present at MANWR. The most bristle-thighed curlews observed at one time was six individuals. This species was routinely observed coming to and from an area near the old Sand Aviary, southwest of the water tanks, at sunset and sunrise. Birds were also observed flying to Bulky Dump approximately 30 minutes after sunset. These are areas that should be investigated further to determine if bristle-thighed curlews are roosting in the area. Figures 14-16 show experiments being conducted to assess the utility of single and multi-species decoy arrays and artificial flooding of areas to attract shorebirds. Trail cameras are recording activity levels of shorebirds around each of the treatments.

28

SUMMARY OF RISKS AND PROPOSED PROTECTION PLAN FOR LAYSAN DUCK AND SHOREBIRDS FOR THE MIDWAY SEABIRD PROTECTION PROJECT BRISTLE-THIGHED CURLEW AND OTHER SHOREBIRDS

Figure 14. Wes Jolley creating shorebird attraction zones to aid in capture using water and decoys.

Figure 15. Decoys were tested in single species and multi species arrays.

29

SUMMARY OF RISKS AND PROPOSED PROTECTION PLAN FOR LAYSAN DUCK AND SHOREBIRDS FOR THE MIDWAY SEABIRD PROTECTION PROJECT BRISTLE-THIGHED CURLEW AND OTHER SHOREBIRDS

Figure 16. A trail camera records attendance at the attraction sites.

Using both whoosh nets and net guns are optimal methods for capturing bristle-thighed curlews. SWCA (2017) used cardboard decoys to lure other curlews in, and a blind was made using 1 in. (2.54 cm) diameter polyvinyl chloride to make a triangular frame. The frame was covered with shade cloth, fastened to the frame, and vegetation was layered on top of the shade cloth to provide camouflage. Zip ties were used to fasten all of the materials to the frame. According to the Midway Atoll airport staff, as many as 50 bristle-thighed curlews congregate around the runway during rain events. Curlews were interested in the cardboard decoys and were able to be lured close enough to the blind to be captured with a CO2-powered Super Talon net gun. To call in birds to the blind, both decoys and bristle-thighed curlew vocalizations (whistling) were used. It is recommended that decoys such as Knuston’s Curlews (model # 28-00) and Knuston’s Golden Plovers (model # 37-00) are used to lure curlews into capture areas for capture.

In previous instances, bristle-thighed curlews have been captured using a spot-light and hand-net method as noted by Marks (1993) and by SWCA field biologists on Palmyra Atoll. On Palmyra Atoll, bristle-thighed curlews were also captured at roost sites by hand using a spot light. If roost or night-foraging areas can be identified, additional curlews can be captured.

30

SUMMARY OF RISKS AND PROPOSED PROTECTION PLAN FOR LAYSAN DUCK AND SHOREBIRDS FOR THE MIDWAY SEABIRD PROTECTION PROJECT BRISTLE-THIGHED CURLEW AND OTHER SHOREBIRDS

5.5 CAPTIVE HOLDING AND TRANSPORT

AVIARY STRUCTURES Shorebird aviaries will be built on Sand Island and on Eastern in proximity to Laysan Duck structures. An aviary with 10 units can house 20-40 curlews, plovers, or turnstones prior to, during, and following baiting operations (Figures 17 and 18). All birds will be housed in single species groupings. An optimal site will have a concrete pad to prevent crabs (that may have consumed brodifacoum) from burrowing or entering the aviary and poisoning birds (secondary poisoning). The aviary will have adequate airflow to keep birds cool and reduce the risk of aspergillosis (Aspergillus sp.) fungal infection of the lungs, shade to keep birds cool, access to saltwater for cleaning, and low human traffic to minimize disturbance. Feather-clipping and releasing on Eastern Island with supplemental food and water provided may prove to be much healthier for adult shorebirds than captive care. The aviaries can be built by 8 people over 2 days (168 people hours). On each island the shorebird aviary will be 50 ft x 30 ft with ten 10ft by 10ft compartments. Eighty percent shade cloth will then be secured into place using a combination of 6.6 ft. (2 m) u-channel and zip ties. The bottom outside edge will be covered with aluminum sheeting 2 x 49 ft. (0.6 x 15 m) or similar material secured by zip ties. The flooring should be a rubberized surface, such as 0.02 x 4 x 49 ft. (0.006 x 1.22 x 15 m). Double layer shade cloth, aluminum sheeting, or similar material will be used as visual barriers between sections. The aluminum sheeting will also serve as a barrier to prevent mice and crabs from entering. Each of the ten units in the aviary will have one 3 gal. (11 L) fresh water pan, one 3 gal. (11 L) salt water pan, and a food dish made from a 6 in. (15 cm) pot tray placed inside of an 8 in. (20 cm) pot tray filled with water, the latter used to keep ants out of the inner food dish.

Figure 17. This shorebird aviary was set on a concrete pad and designed with visual barriers (aluminum sheeting and palm fronds) and predator control devices (aluminum sheeting and rodent traps set on 5 gal. [19 L] buckets).

31

SUMMARY OF RISKS AND PROPOSED PROTECTION PLAN FOR LAYSAN DUCK AND SHOREBIRDS FOR THE MIDWAY SEABIRD PROTECTION PROJECT BRISTLE-THIGHED CURLEW AND OTHER SHOREBIRDS

Figure 18. The inside of this aviary shows a rubberized floor before it was divided up into ten, 9.8 x 9.8 ft. (3 x 3 m) units. On Eastern Island when not enough fresh water is available, salt water will be pumped using a small, ½-horsepower motor (like Little Giant), from the ocean to a 50 gal. (1,324 L) holding tank and then pumped to the aviary in a 0.79 in. (2 cm) diameter garden hose (Figures 19 and 20). Pumps can be powered by a EU2000i Honda generator, which will be kept protected from the weather and housed to reduce noise and exhaust fumes to keep from disturbing birds. The pump can be placed inside a 5 gal. (19 L) bucket, drilled with 0.24 in. (6 mm) holes, which serves as a filter. Water can then be pumped 49 ft. (15 m) from the ocean to a pallet tub holding tank. An additional pump should be placed in the pallet tub, and during high tide, the pumps should run simultaneously to allow the holding tank to remain full, which increases water pressure and allows the holding tank to act as a water reserve when water becomes unavailable at low tide. A garden hose can be used to supply the aviary with water from the holding tank. At the aviary, the garden hose would be attached to a 0.75 in. (19 mm) polyethylene hose that is connected to a series of five plastic hoses with an adapter, to be used for cleaning each unit. Each hose will be marked with the associated unit number and run the length of the aviary. The garden hose will have to be connected to the appropriate plastic hose each time a new pen is visited for cleaning. Roughly 10 gal. (38 L) of fresh water for drinking will be used each day per aviary compartment and carried out to the units using two 5 gal. (19 L) buckets.

32

SUMMARY OF RISKS AND PROPOSED PROTECTION PLAN FOR LAYSAN DUCK AND SHOREBIRDS FOR THE MIDWAY SEABIRD PROTECTION PROJECT BRISTLE-THIGHED CURLEW AND OTHER SHOREBIRDS

Figure 19. A 5 gal. (19 L) bucket being used as an external pump filter.

Figure 20. Salt water being pumped into a holding tank before being pumped to the aviary. This allows aviaries to be cleaned during low tide when seawater is unavailable. Aviaries will be cleaned each morning using a scrub brush and salt water. The outside edge of the flooring will form a gutter and water containing fecal material and old food is pushed to the outside edge of each aviary and drained out. To reduce stress, aviaries will be cleaned in the same order and similar time every day so that birds become accustomed to the routine. Food and water will be replaced after each aviary is cleaned.

Perches and Visual Barriers Perches and visual barriers (shade cloth and aluminum sheeting) will be used to provide stress relief from pacing and flying. Birds in captivity on Palmyra began to develop foot sores from pacing and wing sores from flying into the shade cloth. Perches were created out of rocks and tree branches (Figure 21) that were 2-3 in. (5-8 cm) in diameter. Tree branches will be covered with

33

SUMMARY OF RISKS AND PROPOSED PROTECTION PLAN FOR LAYSAN DUCK AND SHOREBIRDS FOR THE MIDWAY SEABIRD PROTECTION PROJECT BRISTLE-THIGHED CURLEW AND OTHER SHOREBIRDS soft material to encourage healing of foot pads. If padding becomes worn over time it will be changed every 2 to 3 weeks. Pacing areas will be identified by conducting visual observations and using the accumulation of fecal matter on the floor as an indicator. Closed circuit cameras will be installed to enable staff to observe any dangerous or problematic behaviors.

Figure 21. Padding can be added to perches to prevent or allow foot sores to heal.

Predator Control To reduce the small risk of predation by mice, on Sand Island, multi catch mouse traps or sticky traps inside tamper-proof bait stations will be set around the outside perimeter of the aviary, in a way that reduces crab interference. All traps will be baited with a combination of coconut meat and peanut butter and will be checked daily. The bottom outside edge of the aviary will be covered with 2 x 49 ft. (0.61 x 15 m) aluminum sheeting to prevent mice and crabs from chewing through the shade cloth to enter the aviary. Any captive shorebirds not already wing-clipped and transferred to Eastern prior to the bait application would be transferred with the duck population and cared for by the same staff on mouse free Eastern Island.

HOLDING BOXES Holding boxes will be made following instructions similar to that given in Breeden (2011) for bristle-thighed curlew mitigation during the Palmyra Atoll rat eradication effort. Specifically, holding boxes measured 1.5 x 1.3 x 1.3 ft. (46 x 40 x 40 cm). The top, bottom, and sides were made of Baltic Plywood, the door was made of Sintra PVC foam board 0.24 in. (0.6 cm) and the back was covered with hardware cloth 0.24 x 0.24 in. (0.6 x 0.6 cm) and pet screen. A sliding tray was made of Sintra PVC foam board 0.12 in. (0.3 cm) and allowed for easy cleaning. A foam pad 0.51 in. (1.3 cm) was attached to the ceiling to protect birds that attempted to fly. To provide ventilation, each side had ten holes, each 0.51 in. (1.3 cm) in diameter, and each front had 3 holes, each 0.51 in. (1.3 cm) in diameter, in addition to 54 holes, each 0.12 in. (0.3 cm) in diameter. Four electric fans were used to increase air circulation to reduce the occurrence of aspergillosis. A metal frame was made of 0.51 in. (1.3 cm) metal conduit, and holding boxes were placed on the frame. The legs of the frame will be placed on bucket lids filled with water to prevent ants from getting

34

SUMMARY OF RISKS AND PROPOSED PROTECTION PLAN FOR LAYSAN DUCK AND SHOREBIRDS FOR THE MIDWAY SEABIRD PROTECTION PROJECT BRISTLE-THIGHED CURLEW AND OTHER SHOREBIRDS into holding boxes and aggravating birds. Holding boxes will be kept in a quiet place that will provide little noise disturbance. Birds will be kept in holding boxes and released into the aviary once weights stabilize and they are eating on their own. Aviary birds will be monitored and if birds showed signs of not eating, they will be placed back in to holding boxes until their weights stabilize. If birds show signs of stress (pacing and flapping), a towel will be placed over the top to reduce the amount of light entering the inside. These holding boxes took 3 people 24 hours to construct (72 volunteer hours) in previous efforts (Figures 22 and 23).

Figure 22. Holding boxes showing holes for ventilation and bucket lids filled with water to prevent ant infestation.

Figure 23. Each holding box had a removable tray to allow for easy cleaning.

35

SUMMARY OF RISKS AND PROPOSED PROTECTION PLAN FOR LAYSAN DUCK AND SHOREBIRDS FOR THE MIDWAY SEABIRD PROTECTION PROJECT BRISTLE-THIGHED CURLEW AND OTHER SHOREBIRDS

TRANSPORT BOXES Transport boxes will be made following designs similar to those used for the Palmyra rat eradication project (Breeden 2011). For the Palmyra operation, transport boxes were made on site by drilling 6 ventilation holes each 1 in. (2.54 cm) in diameter on the sides of a plastic tote 24 x 11.8 x 15.7 in. (61 x 30 x 40 cm). The top was covered with a cloth that had a bungee cord sewn to the outer edges and a 7.9 in. (20 cm) opening cut in the center. The center opening was covered by a separate piece of cloth that was attached with Velcro. The cloth top reduced injury by providing a soft surface if birds tried to fly and escape. The bottom was lined with indoor/outdoor carpet to prevent slipping. The ventilation holes were made on the upper edge of the tote to prevent birds from trying to put their heads through the hole. Transport boxes had handles that allowed for ease of carrying and transporting of birds.

5.6 CAPTIVE CARE

EXAMINATION AND TREATMENT Examination and treatment methods for curlew will similar to those used for the Palmyra rat eradication project. To reduce stress, the risk of capture myopathy, and injury, curlews will be taken out of the net immediately, their heads were covered and the sedative Midazolam 7.05-5 oz./35 oz. (2mg/kg) will be administered through the nasal cavity. Birds will then be placed in a holding box for 5 minutes to allow the Midazolam to take effect. Holding boxes will be located in a quiet, well-ventilated area away from loud noises and human activity. Once birds are sedated, they will be taken out of the box for examination, weighing, hydrating, measurements, bleeding and banding. A 6 in. (15 cm) SPI plastic dial caliper will be used to take measurements of culmen length, culmen depth (inner, middle and outer), total head and tarsal length. A 6 in. (15 cm) stainless-steel wing ruler will be used to measure the right wing-chord. Weight will be measured with a 21 oz. (600 g) Pesola spring scale by placing individual birds into a cotton bird bag 9.5 in. x 15 in. (24 x 38 cm). Injections of saline solution will be given to each leg for hydration. The keel score will be taken by palpating the breast bone to determine body condition. In addition, each bird will have a picture taken of its culmen and primary feathers against graph paper to determine sex. A blood sample will be taken from the jugular vein of each bird before they are placed into an individual holding box and monitored. Blood will be used to measure cortisol levels, to confirm sex and for population genetics information. Birds will be marked with federal stainless-steel bands on one leg and unique color-flags (black with white alpha/numeric symbols) on the opposite leg. Processed birds will be examined for signs of injury, bait consumption and stress. If brodifacoum bait has already been applied to the island, all captive shorebirds will be administered vitamin K 7.05-5 oz./35 oz. (2mg/kg) by mouth to treat potential exposure and possible previous exposure to brodifacoum. To minimize heat stress birds will be sprayed with rubbing alcohol on the neck, feet and under the wings. Curlews will then be placed back in the holding box and monitored. If birds show signs of stress such as attempting to

36

SUMMARY OF RISKS AND PROPOSED PROTECTION PLAN FOR LAYSAN DUCK AND SHOREBIRDS FOR THE MIDWAY SEABIRD PROTECTION PROJECT BRISTLE-THIGHED CURLEW AND OTHER SHOREBIRDS fly and pacing, the holding boxes will be covered with towels to make the inside dark. Fans will be setup to create airflow to reduce the likelihood of aspergillosis. Birds manifesting the inability to move their legs or stand will be treated for capture myopathy by exercising and massaging the birds’ legs for short periods every 4-5 hours until the bird is able to stand on its own. The weight of the bird will be supported by placing a hand under the birds’ breast and belly, allowing its legs to dangle with feet touching the floor. The bird will be slowly lifted up and down allowing its leg muscles to go through a full range of motion. This physical therapy will be repeated until the bird can support its own weight unassisted (Rogers et al. 2004). Once the bird is able to stand on its own it will be monitored and moved to the aviary. Birds showing signs of ingesting bait, bait in and blue fecal material, will be given vitamin K orally and placed in a holding box that provided food, water and ventilation. Affected birds will be given vitamin K and monitored for 5 days before moving them to the aviary. During the Palmyra rat eradication project, two affected bristle-thighed curlews recovered from the poisoning after treatment. Foot and wing sores will be cleaned when birds are handled. Silver sulfadiazine cream will be applied to sores and the feet will be wrapped with padding. Styptic powder will be applied to wing sores. During the Palmyra rat eradication project, one bristle-thighed curlew was examined after it was seen scratching its face. Examination revealed eye flukes. To remove the eye flukes the bird was sedated with Midazolam and the flukes were pulled out with tweezers. All other captive birds were subsequently examined revealing that 7 out of 13 Bristle-thighed Curlews had eye flukes. Captured Pacific golden plover did not have eye flukes.

FEEDING Feeding methods for curlews will be similar to those used for the Palmyra rat eradication project. Captive curlews in aviaries during the Palmyra rat eradication project (USFWS 2011) were fed once a day in the morning between 0700 and 0900 hours. Birds in holding boxes were fed twice a day between 0900 and 1000 hours and 2100 and 2300 hours. Food items were placed in a 6 in. (15 cm) round tray. Bristle-thighed curlews in holding boxes were tube-fed 0.5 to 0.7 oz. (15-20 ml) of Carnivore critical care Oxbow twice a day until they ate on their own. After tube feeding, crops were packed by hand with egg balls made of hardboiled egg yolk and cat food to reduce weight loss. Water was applied by syringe to the culmen to moisten the throat after each egg ball was consumed. Birds in the aviary and boxes were offered a combination of food items including freeze dried meal worms or crickets, frozen blueberries, hardboiled egg yolk, hot dogs, steak, raw fish, Naupaka (Scaevola sp.) seeds, dried cat food, wet cat food and trout chow (Figure 24). To encourage eating, empty sooty tern eggs and hermit crab shells were packed with food. One Bristle-thighed Curlew only started eating after putting three geckos in its box. Rocks were placed in holding boxes and aviaries to provide tools for opening eggs and hermit crab shells. Unconsumed food and water were removed and replaced with fresh supplies each day. Fecal material was monitored for bile which would indicate birds were not eating. Once weights

37

SUMMARY OF RISKS AND PROPOSED PROTECTION PLAN FOR LAYSAN DUCK AND SHOREBIRDS FOR THE MIDWAY SEABIRD PROTECTION PROJECT BRISTLE-THIGHED CURLEW AND OTHER SHOREBIRDS stabilized (generally 2 to 4 days) birds were transported to the aviary. If birds stopped eating in the aviary, they were placed back into holding boxes and force fed until they started eating and weights had stabilized.

Figure 24. Birds were offered a variety of food items.

5.7 RELEASE

CARCASS AND BAIT MONITORING BEFORE AND AFTER RELEASE To ensure a safe environment for shorebird release the same four-step criteria established for ducks will be followed. Known roosting and foraging sites will be searched for signs of poisoned birds, carcasses and bait presence. Sites will be searched daily if non-target shorebird carcasses are found and every other day if no carcasses are present. All carcasses will be recorded, and carcasses suitable for necropsy will be collected, labeled, frozen and submitted to U.S. Department of Agriculture for analysis. Roost and foraging sites will be monitored before the first bait application, during and up to 42 days after the last aerial broadcast of bait and/or until 30 days after bait no longer persisted in the environment, whichever is longer. To detect bait persistence post eradication, 82 ft. x 3.2 ft. (25 x 1 m) transects at five forage sites will be surveyed. In conjunction with a general search for bait (which is more likely to identify leftover bait if any is present).

RELEASE METHOD After live-capture and holding shorebirds on Sand or Eastern Island, birds will be released when it is deemed to be safe using a four-step process that includes: 1) ensuring all bait pellets are gone; 2) residue levels in insects are testing at levels below what is deemed dangerous for shorebird; 3) shorebirds arriving back from the Arctic appear to be surviving, and; 4) sentinel passerines (canaries and mynas) are surviving. Monitoring will also include attempts to recapture shorebirds demonstrating signs of toxicosis. When it is deemed safe to release birds, birds will be released into areas with suitable habitat where other mixed flock of shorebirds are present. Birds will be captured in the aviary, passed to a bird handler and transported to the release areas. Birds’ heads will be covered and their legs sprayed with isopropyl alcohol to minimize stress during transport.

38

SUMMARY OF RISKS AND PROPOSED PROTECTION PLAN FOR LAYSAN DUCK AND SHOREBIRDS FOR THE MIDWAY SEABIRD PROTECTION PROJECT BRISTLE-THIGHED CURLEW AND OTHER SHOREBIRDS

When risk is deemed low, birds will be released all at once near pooled water into a mixed flock of shorebirds.

5.8 PRE- AND POST-PROJECT MONITORING Successful implementation of a complex management action with broad ecological implications requires a monitoring program that covers a wide suite of measurements. These efforts are already underway and fall into 3 separate but interacting categories. They are:

1. Ecological Effects Monitoring or Conservation Measures – measures of the structure and function of the ecosystem at Midway Atoll as affected by the current population of introduced mice and parallel measurements made after mice are removed. These include diet studies of mice and other species with which they may compete (including ducks), and measures of the population success of native species at Midway before and after implementation including the effects of the toxicant on non-target species at MANWR. 2. Implementation Monitoring – Measures of factors important to designing an effective approach to carrying out the operation. These factors include weather patterns, mouse population cycles, studies of the timing of alternate food sources such as dead albatross chicks and Ironwood seeds, studies of bait competitors that may vie for available bait after it is applied, whether or not all mice are successfully exposed to the bait, the pathways and fate of the toxicant in the environment and the rate at which it becomes unavailable to non- target organisms that may be vulnerable, and the genetic make-up of the mouse population in the case that mice are discovered later, so it can be determined whether it was a failure of the eradication or a re-introduction and failure of biosecurity. 3. Efficacy Monitoring – measures during and after implementation that inform us about the effectiveness of the operation including actual rates of bait on the ground after broadcast and how long it remains available in all habitats, fate of radio-tagged mice during implementation, and measures of mouse activity after operation is complete to evaluate success of the eradication attempt. Data useful for protection of protected species is coming from all three categories of monitoring and will be used to identify pathways by which they could be exposed to the rodenticide and the length of time that protection measures such as captive care must be kept in place. Food web studies currently underway using gut contents and DNA barcoding techniques will inform us as to which invertebrates we will choose to sample for brodifacoum residue following the bait applications. Results from previous eradications at different sites will guide our sampling time frame and frequency. We will endeavor to identify analytical capabilities in advance to ensure feedback is received as quickly as possible to provide decision-makers with the information that will inform us when it is safe to allow captive Laysan ducks and shorebirds to be released back at Sand Island following the eradication.

39

SUMMARY OF RISKS AND PROPOSED PROTECTION PLAN FOR LAYSAN DUCK AND SHOREBIRDS FOR THE MIDWAY SEABIRD PROTECTION PROJECT BRISTLE-THIGHED CURLEW AND OTHER SHOREBIRDS

SENTINEL BIRDS – CAPTURE, CARE AND RELEASE The two passerine species living at Midway are canaries and common mynas. Neither of these birds are protected under U.S. law and so we will use them as indicators of the relative safety of the environment for Laysan ducks and migratory shorebirds after the bait application. Both species have some diet overlap with the other two groups of birds and therefore may help us understand when it is safe to release the captive ducks and shorebirds. Prior to the operation we will capture as many of each species as is feasible and keep them in indoor aviaries on Sand Island for the duration. This will allow us to keep better control over their waste food and reduce the chance of feeding mice. Common mynas can locate invertebrate food at or just beneath the surface of grassland (Feare and Craig 1998), which is often their main source of dietary animal protein. Grassland is thus a preferred habitat and common mynas frequently feed on managed lawns, in park grassland, sports fields, grass airstrips and similar areas. Mynas are generally less attracted to dense forest, except when trees are fruiting, which is often localized in time and space in mixed forest. Captive care for mynas is described by Bockheim and Congdon (2001) and commercially available food for soft bills will be used. Canaries are also widely kept in captivity so there are products for their health and nutrition commonly available. Both canaries and mynas will be released in cohorts once residue testing shows diminished and safe levels of brodifacoum in the environment. Their fates will be monitored using radio-telemetry and observation. 6. SUMMARY Island-wide application of Brodifacoum-25D rodenticide at Sand Island, Midway Atoll will pose a great risk to resident Laysan Ducks and migratory shorebirds of several species. Actions to protect Laysan Ducks will include capturing the majority of ducks at the atoll prior to the bait application and holding a minimum of 200 ducks in aviaries on Sand Island until the bait application and then transferring them to aviaries on Eastern Island. Then, the remaining ducks captured would be held for 10 days to administer two doses of a botulism vaccine, before clipping their flight feathers and releasing them on Eastern Island where there are no mice. Birds being held in the aviaries on Eastern will be kept there until there is confirmation through a four-step process to determine when it is safe to release ducks back to Sand Island. These four steps include: 1) there are no pellets remaining; 2) brodifacoum residue levels measured in selected invertebrates reach levels deemed safe; 3) migratory shore-birds returning from their northern breeding grounds appear to be staying healthy upon arrival, and; 4) sentinel canaries and common mynas are released on Sand Island and show no ill effects. Flight-feather-clipped ducks living on Eastern with food and water supplementation will be supported there until all have undergone a molt of their flight feathers and they can again move between islands. Shorebirds remaining at Midway Atoll after the end of May when most birds depart for their northern breeding grounds will be captured and either be held in captivity on Sand and then transferred to Eastern Island near the time of bait application or have their flight feathers clipped and be released on Eastern Island. Shorebirds returning after

40

SUMMARY OF RISKS AND PROPOSED PROTECTION PLAN FOR LAYSAN DUCK AND SHOREBIRDS FOR THE MIDWAY SEABIRD PROTECTION PROJECT BRISTLE-THIGHED CURLEW AND OTHER SHOREBIRDS the breeding season will be captured throughout the operational period, wing clipped, and transferred to Eastern Island. Non-target organisms held in captive care will be released back to Sand Island using the same four-step process for Laysan ducks described above.

41

SUMMARY OF RISKS AND PROPOSED PROTECTION PLAN FOR LAYSAN DUCK AND SHOREBIRDS FOR THE MIDWAY SEABIRD PROTECTION PROJECT LITERATURE CITED

7. LITERATURE CITED Birdlife International. 2016. Numenius tahitiensis. The IUCN Red List of Threatened Species [cited 07 Oct 2017]. Available from http://dx.doi.org/10.2305/IUCN.UK.2016- 3.RLTS.T22693182A95218288.en. Blackburn, T.M., P. Cassy, R.P. Duncan, K.L. Evans, and K.J. Gaston. 2004. Avian extinctions and Mammalian introductions on oceanic islands. Science 305:1955-1958 Bockheim, G. and S. Congdon. 2001. The Sturnidae Husbandry Manual and Resource Guide Breeden, J.H. 2011. The Capture and Captive Maintenance of Bristle-thighed Curlews during a rat eradication, Palmyra Atoll, 2011. Unpublished report. Bub, H. 1991. Bird trapping and bird banding: A handbook for trapping methods all over the world. Translated by F. Hammerstrom, K. Wuertz-Schaefer.Cornell University Press, Ithaca, New York. CABI. 2018. Acridotheres tristis (common myna). Datasheet. Invasive Species Compendium. https://www.cabi.org/isc/datasheet/2994 Deaton, J.W., F. N. Reece, E.H. Mcnally and W.J. Tarver. 1969. Liver, heart and adrenal weights of broilers reared under constant temperatures. Poultry Science 48(1): 283–288. Duncan, R.P. and T.M. Blackburn. 2007. Causes of Extinction in Island Birds. Animal Conservation 10:149-150. Engilis, Jr., A., and M.B. Naughton. 2004. U.S Pacific Islands Regional Shorebird Conservation Plan. In U.S. Shorebird Conservation Plan. Portland, OR: U.S. Department of the Interior, Fish and Wildlife Service. Feare, C.J. 2010. Invasive bird eradication from tropical oceanic islands. Aliens: The Invasive species Bulletin 30:12-19. Feare, C., and Craig, A. 1998. Starlings and mynas. Christopher Helm (A & C Black Publishers Ltd.), London, UK. 285 p. Horak, K., Fisher, P. and B. Hopkins. 2018. Pharmacokinetics of Anticoagulant Rodenticides in Target and Nontarget Organisms. USDA National Wildlife Research Center - Staff Publications. 2091. https://digitalcommons.unl.edu/icwdm_usdanwrc/2091 Marks, J.S., R.L. Redmond, P. Hendricks, R.B. Clapp, and R.E. Jr. Gill. 1990. “Notes on longevity and flightlessness in Bristle-thighed Curlews.” Auk 107:779-781. Marks, J.S. 1993. “Molt of Bristle-thighed Curlews in the northwestern Hawaiian Islands.” Auk 110:573-587.

42

SUMMARY OF RISKS AND PROPOSED PROTECTION PLAN FOR LAYSAN DUCK AND SHOREBIRDS FOR THE MIDWAY SEABIRD PROTECTION PROJECT LITERATURE CITED

Marks, J.S., R. Tibbitts, R.E. Gill Jr., and B.J. McCaffrey. 2002. Bristle-thighed curlew (Numensius tahitiensis). Cornell Lab of Ornithology [cited September 2017]. Available from http://bna.birds.cornell.edu/bna/species/705. Millett, J., Climo, G. and J. Shah. 2004. Eradication of the common myna Acridotheres tristis populations in the granitic Seychelles: successes, failures and lessons learned. Advances in Vertebrate Pest Management 3: 169-183. Mineau, P. 2014. A review and scientific opinion of the ecotoxicological data from Pinzon and Rabida Island lava lizards and the risk this represents to Galapagos hawks. Report prepared for Island Conservation. 16 pp. Mineau, P. 2018. A review of Midway Island’s mouse eradication mitigation plans. Report prepared for Island Conservation. 20 pp. Moulton, D.W., and A.P. Marshall. 1996. Laysan Duck (Anas laysanensis), version 2.0. In The Birds of North America, edited by P. G. Rodewald. Cornell Lab of Ornithology: Ithaca, New York, USA. https://doi.org/10.2173/bna.705. Moulton, D.W., and M.W. Weller. 1984. "Biology and conservation of the Laysan Duck (Anas laysanensis)." Condor 86:105-117. Pierce, R., N. Anterea, U. Anterea, K. Broome, D. Brown, L. Cooper, H. Edmonds, F. Muckle, B. Nagle, G. Oakes, M. Thorsen, and G. Wragg. 2008. Operational work undertaken to eradicate rats and rabbits in the Phoenix Islands, Republic of Kiribati, May-June 2008. Pacific Expeditions Ltd. Pitt, W. C., A.R. Berentsen, A. B. Shiels, S. F. Volker, J.D. Eisemann, A.S. Wegmann, and G.R. Howald. 2015. “Non-Target Species Mortality and the Measurement of Brodifacoum Rodenticide Residues after a Rat (Rattus rattus) Eradication on Palmyra Atoll, Tropical Pacific.” Biological Conservation 185:36-46. Reynolds, M. and J. Klavitter. 2006. “Translocaiton of wild Laysan duck Anas laysanensis to establish a population at Midway Atoll National Widlife Refuge, United States and US Pacific Possession.” Conservation Evidence 3:6-8. Reynolds, M., and K. Kozar. 2000. Laysan duck (Anan laysanensis) population status and translocation feasibiity report. USFWS unpublished report: Honolulu, HI. 88 pp. Reynolds, M.H., Courtot, K.N., Brinck, K.W., Rehkemper, C.L., and Hatfield, J.S. 2014. Long- term monitoring of endangered Laysan ducks: index validation and population estimates 1998– 2012. Journal of Fish and Wildlife Management 6(1):92–101; e1944-687X. doi: 10.3996/032014- JFWM-017. Reynolds, M.H., K. Courtot, and J. Hatfield. 2017. How many Laysan Teal (Anas Lysanensis) are on Midway Atoll? Methods for monitoring abundance after reintroduction. Wildfowl 67:60-71.

43

SUMMARY OF RISKS AND PROPOSED PROTECTION PLAN FOR LAYSAN DUCK AND SHOREBIRDS FOR THE MIDWAY SEABIRD PROTECTION PROJECT LITERATURE CITED

Rogers, D. I., P. F. Battley, J. Sparrow, A. Koolhaas and C. J. Hassell 2004. Treatment of capture myopathy in shorebirds: a successful trial in northwestern Australia. J. Field Ornithol. 75:157– 164. SWCA 2017. Laysan Teal And Bristle-Thighed Curlew Assessment Report For The Mouse Eradication Project, Midway Atoll National Wildlife Refuge. SWCA Environmental Consultants Project No. 042077 prepared for Island Conservation. 43 pp. Thomas, P.J., P. Mineau, R.F. Shore, L. Champoux, P. Martin, L. Wilson, G. Fitzgerald, J. Elliott. 2011. Second generation anticoagulant rodenticides in predatory birds: probabilistic characterization of toxic liver concentrations and implications for predatory bird populations in Canada, Environment International 37(2011): 914-920. (Erratum: 40(2012): 256). Tidemann, C. 2007. Common Indian Myna Website. Australia: Australian National University. http://fennerschool-associated.anu.edu.au//myna/index.html USFWS. 2009. Revised Recovery Plan for the Laysan Duck (Anas laysanensis). USFWS Region 1. Portland, OR, 127 pp. ---. 2011. Palmyra Atoll National Wildlife Refuge Rat Eradication Project Final Environmental Impact Statement. U.S. Fish and Wildlife Service report: Portland, Oregon. ---. 2017. Draft Environmental Assessment for the Midway Seabird Protection Project. U.S. Fish and Wildlife Service report: Portland, Oregon. ---. Unpublished. “Regarding seabirds and shorebirds of Midway Atoll National Wildlife Refuge.” VanderWerf, E.A. 2012. Hawaiian Bird Conservation Action Plan. Pacific Rim Conservation, Honolulu, HI. Warner, R. 1963. Recent History and Ecology of the Laysan Duck. Condor 65(1).

44

FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT APPENDIX D

APPENDIX D. SEEP CHECK PROTOCOL LAYSAN DUCK GUZZLER and ARTIFICIAL POND CHECKS (Revised August 2018, Kelly Goodale, Wildlife Biologist USFWS)

1. Overview

GOAL To effectively and efficiently manage the risk of large-scale avian botulism type C epizootics at Midway Atoll

OBJECTIVE Water and surrounding vegetation adjacent to artificial ponds (aka “seeps”), guzzlers, ephemeral wetlands with a history of botulism, and other manmade water-holding structures that have a history of botulism are patrolled in order to detect and remove sickened or dead ducks and other sources of animal protein (such as dead seabird eggs/chicks or mice) that could contribute to or escalate an outbreak of avian botulism type C. The secondary goal is to get an accurate assessment and movement of the population. This will be done in conjunction with seep checks on a weekly basis. In order to get the most accurate assessment of the population and movement of ducks, at least one person will be deployed to one of four locations. By checking these areas at the same time, it reduces the chances of double counting and will give us the most accurate population assessment and movement. For the sake of brevity and established Midway nomenclature, patrols of Laysan duck habitat will hereafter be referred to as “seep checks.”

LOCATIONS TO BE CHECKED • On Sand Island, there are currently 8 artificial ponds, 4 guzzlers, and 5 low areas that become temporarily flooded after heavy rains. Though the likelihood of botulism being present at each site is highly variable, each location can attract ducks. Additionally, with the exception of guzzlers, all of these locations have at least some history of botulism and support widely variable water levels. There are four ponds that were dug to create Laysan duck habitat prior to the 2004 translocation: Big and Little Ballfield Ponds and Radar Hill East and West Ponds. Brackish Pond is a permanent pond consisting of brackish water in a former sand/gravel borrow pit. Catchment is a large, shallow concrete basin that was part of the water collection infrastructure for the Naval Air Facility at Midway. R2 Ditch is a small but relatively deep drainage ditch that is a part of the operating runway infrastructure. In addition to the artificial ponds, there are four guzzlers (plastic water basins with sloped floors and a rain catching roof that are intended to provide additional water sources for ducks) installed on the island. Guzzlers are found near the FWS Office, in Parade Field, and at two sites formerly supporting constructed ponds, named Communications and Sunrise Seeps, respectively. These are now called Communications Guzzler and Sunrise Guzzler. Additionally, there are five

D-1

FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT APPENDIX D

major ephemeral ponds that pose a potential botulism risk after heavy rains: Mauka- Makai (a former artificial pond that was back-filled in 2013), a drainage ditch adjacent to the former Mauka-Makai pond, Sunrise East and the low area adjacent to Brackish Pond. • There are three constructed ponds and three guzzlers on Eastern Island. The three ponds are named Monument, Sunset, and Rolando. Each pond has one guzzler installed in its vicinity.

EQUIPMENT NEEDED

• Binoculars • Stick or golf club for turning vegetation • Duck kit (includes: zip lock bags, nitrile gloves, a small towel, and a holding bag that cinches closed). • Radio (essential for your safety and for timely notification and treatment of a sick duck).

2. Methods

GENERAL GUIDELINES: • When conducting a seep check, finding dead or sick ducks is the first priority. Other activities such as band re-sights, duck counts, or vagrant sightings are always secondary. • In general, it is preferable to check the seeps in the morning. However, during cooler weather the importance of morning seep checks is reduced. The primary reasons to perform seep checks before other work takes place are to get the disruption of seep checks out the way as soon as possible and to detect and remove carcasses before maggots have a chance to develop, especially during hot weather months. Maggot formation typically takes approximately 24 hours. • When more than one refuge staff is available to check seeps, rotate weekly or bi-weekly visits so that a new set of eyes checks each seep as often as possible. Don’t check a seep the same way each time; go clockwise one day and counter-clockwise the next. Also, alternate high then low or rotate who checks which seep.

DETAILED PROTOCOL: 1. When first approaching a seep, note any LADU immediately fleeing the area, particularly if they are having difficulties moving. Assess these birds for botulism symptoms (see below, #3).

2. Visually survey the water surface, the edge of the water, and the vegetation surrounding the water’s edge out to 15 feet. At many locations, it will be necessary to

D-2 FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT APPENDIX D

enter the water in order the check under vegetation along the water edge. Some ponds have duck-attracting vegetation features outside of the 15 ft. search area. Because sick or dead ducks have been found in these areas in the past, it is essential that these spots are checked during patrols of these locations. The specific locations and vegetation features are as follows:

• Radar Ponds: Look under naupaka shrubs near the edges of each seep, scan hillsides above both seeps. • Catchment Basin: Check under seagrape trees and in duck trails through the tall grass. • Mauka-Makai Ditch: Spot check vegetation throughout ditch when there is standing water. Check perimeter of the former seep to the north (now filled in) when it is flooded. • Brackish Pond: Check the perimeter of any ephemeral ponds created during rains. • Rolando Pond: Check under naupaka shrubs east of the seep.

3. Observe LADU on land and in the water for behavior that may indicate early signs of botulism, including: . Squinty eyes . Weak, slow walking or swimming (may “wing-walk” if legs are paralyzed) . Inability to hold head erect . Inability to fly away ***Note: Laysan ducks on Midway are extremely habituated to humans due to the near daily presence of people in the artificial habitat. Do not use the fact that a duck is not fleeing from you as singular indicator of botulism.***

4. Walk completely around each seep. Use binoculars to scan the shoreline from different locations.

5. Thoroughly check dense vegetation and surrounding areas by moving leaves aside with a stick (ducks will try to hide when they become sick). Also check under trees where they may seek shade.

6. During duckling season, walk from the outside of the vegetation in towards the seep, so that if you flush any broods you will not be blocking their access to the water. Be careful when females with broods are present. Keep your distance and give them time to scurry away. Avoid separating the mothers from their broods.

7. Listen and look for flies. This may be an indication of a dead duck. Sense of smell is also important, however, the amount of dead albatross chicks (particularly during peak botulism season) makes smell a less obvious indicator.

8. Remove any other protein sources (such as seabird carcasses and eggs) from the water and along embankments to prevent a build-up of substrate on which the

D-3

FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT APPENDIX D

botulism bacteria can grow. Deposit them on high ground at least 20’ from seeps where they will not drain back into the seep if it rains.

IF YOU FIND A SICK LADU • Contact the Refuge Biologist via radio so they can meet you in the field with the antitoxin.

• Approach a sick LADU quietly. Even lethargic LADU can make a last ditch effort to fly, so consider using a net. Anticipate its movement and avoid hitting the LADU with the net rim.

• Handle a sick LADU gently. Minimize talking and use quiet voices.

• Bring the sick LADU back to USFWS office for treatment. Use a pet carrier if possible, but a duck bag (cinched tight to prevent escape) works if this is the only thing available. o Note: There is a pet carrier with other supplies stored in the shed on Eastern Island.

• Don’t forget to finish checking the rest of the seep after ensuring the sick LADU is being attended to.

IF YOU FIND A DEAD LADU OR OTHER WATERBIRD NEAR A SEEP 9. Remove carcasses of any waterfowl or shorebird species using gloves and Ziploc bags provided in your duck kit.

10. Be sure to include all maggots and eviscerates in the Ziploc when collecting dead ducks. This is important because a single maggot can contain enough toxin to sicken or kill a duck.

11. Bring all dead ducks back to the office and place in the appropriate section in the duck freezer. Put older, rotting carcasses in the receptacle designated for incineration. Put fresh carcasses (no maggots or odors) into the one designated for necropsy candidates.

12. If you haven’t already, inform Refuge Biologist of any dead LADU. Record the appropriate information on the LADU Intake Log in the lab (be sure to assign LADU carcasses a new log #). Record log #, species, location, band number if applicable, date, time, and your name on the Ziploc.

13. Do not remove bands from freshly dead LADU as they may be sent off for examination and it is important that the band remain on the bird for this process.

Seep check Safety and Ecological Integrity Issues

• Do not risk personal injury or injury to the animal when trying to capture a sick duck. Many ponds (particularly Brackish Pond and all the Eastern Island ponds) contain thick

D-4 FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT APPENDIX D

layers of pond muck several feet deep. Getting mired in the muck while doing a seep check alone poses a serious safety risk. If you must enter interior sections of the seep, proceed with extreme caution or avoid this altogether by using a long-handled net to retrieve a duck or other dead item. Ask for additional help if needed.

• Many ponds also have sharp objects such as branches, nohu thorns, metal or plastic debris, or buried military waste that may not be visible from above. Never enter a pond barefoot. The minimum footwear to enter a pond are sandals with straps around the and heel. Slippers are not suitable.

• Destruction or degradation of native plants is an adverse effect of seep checks. Repeated ground disturbance in some of the most intact native plant areas of the Refuge, such as Brackish, Monument, and Sunset Ponds set back Refuge habitat restoration efforts by trampling or otherwise destroying desirable plants and opening habitat for aggressive weeds. Be careful where you walk and be as gentle as possible on the plants, particularly in the sensitive areas mentioned above. Sesuvium portulacastrum is one species in particular that is relatively rare at Midway Atoll and is being out-planted or encouraged at certain constructed ponds. As much as possible, please avoid stepping on this sensitive plant.

• The spread of invasive weeds is an additional serious, adverse effect of seep checks. While a few ponds have native-dominated plant communities, most of the seeps are densely infested with some of the most aggressive and difficult to control weeds on the Refuge. Many of these weeds, such as Phylla nodiflora began their spread across the Refuge starting in 2008, when botulism patrols in the artificial duck habitat were first initiated. Since then, many of these plants have become deeply entrenched and widespread across the Refuge. When conducting a seep check, take care to ensure that you are not transporting seeds or any other plant material from the pond or guzzler to other locations on the Refuge.

3. Determining seep check frequency

DEFINITIONS (for the purposes of determining seep check frequency)

Emaciated: Body condition score of 1.0 or less. Refer to laminated chart taped to wall in duck lab. Poor body condition: score of 1.0-2.0 Good body condition: score of 2.0-4.0 Trauma: any broken bone (leg, wing, , other) or open wound. Large patches of missing feathers (with or without broken skin) are also considered a sign of trauma Other disease/illness: non-traumatic condition asymptomatic of botulism, including severe dehydration, as described in the Midway Botulism Captive Care intake exam protocol) Botulism suspected: The duck is in moderate to good body condition, there are no signs of trauma or other disease, and it exhibits signs of botulism including: squinty eyes, slow

D-5

FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT APPENDIX D

walking or swimming (may “wing-walk” if legs are paralyzed), inability to hold head erect, inability to fly away Duckling: a duck age 1-60 days old (see Table 2. development and age classification of ducklings for further information) Hatch-year: Fully feathered duck, less than one year of age Botulism season: The period of year when botulism conditions are most consistently present at Midway and when, based on historical data (2007-2015) of confirmed and suspected botulism cases, there is the highest risk of epizootics. Botulism season is June- September at Midway. August is historically the peak botulism season (see Table 1). Non-botulism season: Though there has been at least one botulism case every month of the year at Midway, there is a distinct, off-peak season during the fall, winter, and early spring months, when ambient temperatures are cooler, water temperatures in the ponds are lower, and when ducks are more concentrated in terrestrial areas away from water (see Table 1). When this season overlaps with periods of resource scarcity (i.e. winter at Midway), the naturally elevated level of duck mortality can lead to significant amounts of crew time spent patrolling for botulism when there are, in fact, few to no actual botulism cases.

BASELINE FREQUENCY OF SEEP CHECKS

WEEKLY: Patrols of every constructed pond and guzzler once per week is the default schedule during non-botulism season, when environmental conditions favorable to botulism are not present. Based on verified and suspected botulism detections at Midway from 2007-present, weekly checks are sufficient manage the risk of botulism epizootics.

BI-WEEKLY: During botulism season from June through September, two patrols per week are warranted due to the preponderance of ideal habitat for botulism and increasing frequency of environmental conditions conducive to bacterial production.

Seep checks on Eastern Island are subject to weather, equipment availability, and availability of certified and qualified boat operators. When conditions are safe enough and all other requirements met, seep checks on Eastern may be conducted, as scheduled.

Seep check frequency above the baseline is determined according to conditions specific to each detection of a sick or dead duck. Refuge biological staff, in consultation with the Refuge Manager, will use the Situation-based Seep Check Frequency Decision-support Tool to evaluate each carcass or sick duck detection individually and determine at what spatial scale and frequency any elevated patrols will occur. Search intensity may also be limited by available staffing resources, safety considerations, and competing Refuge priorities.

INTENSE: If more than two sick or dead ducks are found within a 24-hour period in a given area, it may be warranted to conduct an intensive search centered on that “hotspot” because of the possibility of an undetected carcass supporting growth of potentially toxic maggots. The exact search perimeter depends on site-specific conditions and staff time available to dedicate to carcass searches. During an intense search effort, approximately doubling the search radius to 30’ from the current water line is recommended. Detailed

D-6 FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT APPENDIX D

checks around each bunch of grass or sedge at the waterline and within the 30’ radius may be required, however, removal of native plants to enhance carcass searches should be considered a last ditch effort.

Situation-based seep check frequency decision-support tool

Situation Time of Location No action Island-wide Intense check year daily checks at detection for 4 days site(s) ENCOUNTERS OF SINGLE DEAD DUCKS Dead duck: trauma ANY ANY X Dead duck: other ANY ANY X disease/illness Dead adult duck, body Non- >15’ from X condition ≥1.0, no obvious botulism pond or guzzler cause of death season or in town Dead adult duck or hatch Botulism ANY X year duck, body condition season ≥1.0, no obvious cause of death Dead adult or hatch year ANY ANY X duck, body condition ≤1.0 Dead adult or hatch year Non- Constructed X duck, body condition ≥1.0 botulism pond or guzzler season Dead duckling 1-60 days old ANY ANY X (Class IA to Class III) Carcass with maggots, any ANY Constructed X age pond or guzzler ENCOUNTERS OF LIVE, SICK DUCKS Fully feathered hatch year or ANY ANY X adult duck showing signs of botulism ENCOUNTERS OF MULTIPLE DEAD OR SICK DUCKS Two botulism suspected ANY ANY X X adult ducks found within 3 day check period at same site Botulism suspected adult ANY ANY X duck found at second location on same island during island-wide checks initiated above >1 botulism suspected adult ANY Constructed X X duck found at one site on pond or guzzler same day

D-7

FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT APPENDIX D

Table 1. All botulism suspected and confirmed Laysan ducks found at Midway Atoll NWR 2007-2016

D-8 FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT APPENDIX D

Table 2. Average number of days between confirmed or suspected botulism case adult LADU by season.

12

9.78 10

8

6

4 Average Number of of Days Number Average 2.34 2

0 Offseason Botulism Season

Count of Days Between Botulism Cases - 120 OFFSEASON

100

80

60 Counts

40

20

0 0 1 2 3 4 5 6 7 8 9 10 11 12 Number of days between botulism cases

D-9

FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT APPENDIX D

Count of Days Between Botulism Cases - BOTULISM SEASON

120

100

80

60 Counts

40

20

0 0 1 2 3 4 5 6 7 8 9 10 11 12 Number of days between botulism cases

Sequential Botulism Cases Found at Same or Different Island - OFFSEASON

Different 33%

Same 67%

Figure 5. This graph shows whether a sequential, subsequent botulism case was discovered on the same island or different island as the most recent case outside of the botulism season.

D-10 FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT APPENDIX D

Sequential Botulism Cases Found at Same or Different Island - BOTULISM SEASON

Different 19%

Same 81%

Figure 6. This graph shows whether a sequential, subsequent botulism case was discovered on the same island or different island as the most recent case during botulism season.

Table 2. Plumage development and age classification of ducklings (Wildlife Management Institute)

D-11

FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT APPENDIX D

D-12 FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT APPENDIX E

APPENDIX E. CAPTIVE CARE PROTOCOL

E-1

CAPTIVE CARE PROTOCOL For Laysan Ducks Affected with Avian Botulism (Updated August 2018) For trained and qualified personnel only.

CONTENTS 1. General Guidelines ...... p. 2 2. Collection ...... p. 3 3. Intake Exam ...... p. 4 4. Continuing Care ...... p. 8 5. References ...... p. 12 Appendix 1 – Intake Exam: Condensed Version ...... p. 13 Appendix 2 – Body Condition Score ...... p. 14 Appendix 3 – Diet Preparation and Storage ...... p. 15 Appendix 4 – Release Criteria and Banding Procedure ...... p. 19 Appendix 5 – Cleaning and Disinfecting ...... p. 20 Appendix 6 – Disposition of Carcasses ...... p. 22 Appendix 7 – Resources ...... p. 23 Appendix 8 – Datasheets ...... p. 24 Appendix 9 – LADU Captive Care Flowchart ……………………………..…………………………………..p. 28

1

1. GENERAL GUIDELINES

 Before handling a bird, wash your hands thoroughly since sunscreen and hand lotions can affect the bird’s natural waterproofing.

 When handling a duck, work in a team of two: one person to restrain the bird and the other to administer injections or tube-feedings.

 Never leave ducks unattended, even if they appear weak.

 To minimize the bird’s stress level, use quiet voices and limit the bird’s vision (for example, place towels over the pet carrier for a visual barrier).

 Any injections need to be administered with a new, sterile needle and syringe – do not reuse.

 Keep the work area clean and the equipment sanitized.

 Record all information pertaining to captive duck care on the appropriate datasheets.

 Hawaii Wildlife Center is available and willing to provide guidance on any compromised Laysan Duck. Feel free to call them for advice.

 Phone: 808-345-8421  Email: [email protected]

2

2. COLLECTION Do not risk personal health and safety in an attempt to capture affected birds.

1. Prior to beginning collection efforts, check in with the Refuge Biologist. Ask for assistance to capture bird(s) if needed.

2. If the bird is immobile, pick it up by hand. If the bird is still able to move, work in teams of two or more individuals to herd it away from the water and towards a person holding a net or a towel.

3. If possible and depending on location, a qualified FWS staff or Kupu member will meet you in the field to administer the antitoxin, provide a pet carrier, and transport to FWS office. a. For responder: Retrieve one 0.5 cc syringe of antitoxin from wildlife freezer in duck lab and thaw in hand to body temperature. i. Find thickest part of breast muscle, moisten area with a few drops of alcohol, separate feathers with sterile swab to expose skin, and insert needle bevel up into breast muscle (one finger’s width away from keel for ducks). Retract needle to ensure needle not in a vein. If in a vein, you will see blood at base of needle. Pull needle out and re-insert, repeat retraction. If no blood, inject antitoxin intramuscularly into breast muscle. 4. Transport the bird to the duck lab at the FWS office. Use a pet carrier if available, with a towel over it to minimize visual stress. If a carrier is not available, secure the bird in a duck bag for transport, cinching the bag up tight, being extra careful not to drop or injure the bird.

3

3. INTAKE EXAM 1. Prepare a pet carrier, water bath, fluids, and antitoxin (if not administered in the field) in the FWS duck lab. 2. Administer 0.5cc antitoxin if not administered in the field. Find thickest part of breast muscle, moisten area with a few drops of alcohol, separate feathers with sterile swab to expose skin, and insert needle bevel up into breast muscle (one finger’s width away from keel for ducks). Retract needle to ensure needle not in a vein. If in a vein, you will see blood at base of needle. Pull needle out and re-insert, repeat retraction. If no blood, inject antitoxin intramuscularly into breast muscle. 3. Record weight, body condition, temperature, and dehydration level in Captive Care Notes datasheet (Appendix 9). a. Weight: Using the pesola spring-scale, weigh and record the weight of the duck bag by itself. Then weigh the duck in the bag ([weight of duck in bag] – [bag weight] = duck weight). Be sure the bag is cinched tight to prevent escape. Adults typically range from 400-600 g. b. Body condition score: Locate the keel () on the bird and feel its level of prominence relative to the muscle mass on either side. Use the figure in Appendix 3 to assign the bird a score from 1-3. c. Temperature: Gently insert the digital thermometer into the cloaca to get the internal temperature. A normal temperature for a duck is 102°-106° F. i. If less than 102°F, place a heating pad with a towel between the bird and the pad in the pet carrier after the intake exam is completed. Set the heating pad on the low setting. Situate the pad so that only half of it is inside the carrier. Monitor very closely and adjust or remove heat as necessary. Caution must be used, as it is easy to overheat the duck so check on the bird frequently. ii. If greater than 106° F, wipe the feet with alcohol swabs to aid in cooling. Place the bird in the pet carrier and put the carrier in a cool, shaded, well-ventilated area away from human disturbance. d. Dehydration: To assess the level of dehydration, examine the eyes, inside of the mouth and skin of the feet. Determine whether the dehydration level is slight, moderate or severe. i. Well hydrated – bright, glistening eyes, the inside of the mouth will be a pinkish- red color and moist, and the skin of the feet will be smooth and supple. ii. Severely dehydrated – sunken eyes, the tongue and membranes inside the mouth will be white or pale pink with ropy and tacky mucous strands, and the skin of the feet will be dry/flaky. Another sign will be if you pinch the skin near the cloaca and it holds a tent shape relatively long before smoothing out. 4. Administer fluids: Obtain a bag of Lactated Ringers Solution (LRS) or 0.9% Sodium Chloride solution (NaCl). a. Subcutaneous administration (SQ) – Fill a syringe with 20 cc using a sterile 18-gauge needle. Replace the 18-gauge needle with a sterile 25-gauge butterfly needle. Put the

4

syringe and needle in a warm water bath for ~ 5 minutes. The fluid should be around the duck’s body temperature. Be careful not to overheat the fluid or administer fluid that is too cool. Find loose skin on the inner thigh (see Figure 1) and moisten with an alcohol wipe to clean the area. Insert bevel up just under skin, the needle doesn’t need to go in beyond the bevel. Retract the needle to ensure the needle not in a vein. If in a vein, you will see blood at base of needle. Pull needle out and re-insert, repeat retraction. If no blood, slowly inject LRS subcutaneously. Attempt to give 20 cc of fluids, if it is too much; insert the needle in the inguinal fold of the other leg.

Figure 1. On the left, the arrow is pointing at the general region of a duck’s right inguinal fold. On the right is a closer look at the loose skin at the inguinal fold where you want to inject SQ fluids. b. Oral administration (PO) – Oral fluids can be given only when the bird is able to hold its head up and protect its airway. Cut a corner of the LRS or NaCl bag and empty the contents into a container with a lid. Draw up 20 mL of solution into a 35 mL catheter syringe. Always draw up 5 – 10 mL more than required for tubing since at least 5 mL will be lost in the gavage tube. Remove any air bubbles by pointing the syringe upwards and depressing the plunger to move any air out through the tip. Firmly attach the gavage tube to the syringe tip. Place the syringe with the attached gavage tube in the hot water bath for ~5 minutes to heat the solution. Once heated, prime the tube by filling the entire length with solution. This prevents air from being pushed into the bird’s stomach. Continue to depress the plunger until the amount in the syringe equals 15 mL, which is the maximum amount the bird should receive upon intake. Have one person restrain the bird while the other gavages. The restrainer should hold the bird on the table or at waist level. The person gavaging should obtain firm control of the bird’s head. Gently extend the bird’s neck and use your thumb and index finger to open its mouth. Locate the glottis, which is the opening to a bird’s trachea. It lies directly behind the tongue and will open and close as the bird breathes. AVOID inserting the tube in the glottis as this will send the fluids into the lungs or the air sacs, causing pneumonia or death. Insert the

5

gavage tube to either side of the glottis (Figure 2). Fully extend the bird’s neck up and out. Insert the tube on either side of the glottis and slowly slide it down the esophagus. On a duck it is tough to visually check the tube’s placement due to the small size of its mouth, so you can double check your tube placement by gently feeling the ventral area of the bird’s neck. You should be able to feel two distinct tubes: the gavage tube and the trachea. The gavage tube will feel larger and sturdier. If you only feel one tube, the gavage tube may be in the trachea, remove and try again. Once in the esophagus, allow the bird to partly close its mouth. This will be more comfortable for the bird and it will struggle less. Continue to advance the gavage tube gently down the esophagus until the sharpie mark for length is reached. Depress the plunger at a moderately slow, even pace. Tube feed in 5 mL increments while watching for regurgitation. Give no more than 10 mL for the bird’s first PO fluid administration, and no more than 15 mL for any subsequent administrations. Make sure to hold the top of the tube where it meets the syringe since pressure can cause the tube to come off and spray the contents over the bird and handler. Once the syringe is empty, pinch off the tube and slowly pull it out. Pinching off the tube creates a vacuum along the length of the tube and stops any liquid from dripping into the bird’s trachea. Return the duck to its carrier immediately. The stress of restraint may cause it to regurgitate.

Esophagus

Glottis

Figure 2. On the left is a picture of a bird’s mouth, when open. The glottis is indicated by the red arrow. This is where you want to avoid inserting the tube, instead aiming for the esophagus (blue) on either side of the glottis. Photo: Hawaii Wildlife Center. On the right is an example of a duck being gavaged. Photo: Greg Joder.

6

Potential problems encountered when gavage feeding:

. Resistance while advancing the tube. STOP if you feel resistance. Never force the tube as it can rupture the wall of the esophagus and be fatal to the bird. Reevaluate several things at this point:

 Is the bird’s neck straight? A crooked, twisted or L-shaped neck can cause the tube to hit the esophageal wall or the thoracic inlet. It may become stopped by the clavicle.

 Is the stomach full with food? Withdraw the tube and check the tip to see if there are remnants of food there. If this is the case, wait a few more hours for the bird to digest the food before trying to tube feed again.

 Is the tube adequately lubricated? If the tube and bird are positioned correctly, express a small amount of tubing fluid (2-3 mL) into the esophagus to help with lubrication.

 Is the tube clamped in the bird’s bill? This will stop the normal advancement of the gavage tube.

. Regurgitation. If the bird begins to regurgitate, immediately pinch and remove the gavage tube. Do not hold the bird upside down, but gently hold the bird and let it shake out the fluid by itself. Cotton swabs may be necessary to clear food from the bird’s trachea. Never try to keep a bird from regurgitating by holding its head up or keeping its mouth closed. This will cause aspiration that can lead to pneumonia.

 Emaciated, hypothermic birds, newly admitted birds, or those that are severely stressed are prone to regurgitation.

 A sign that the bird has aspirated a significant amount of fluids is a gurgling or a clicking sound when they breathe.

6. Place the bird in the pet carrier, cover it with a large towel, and relocate the carrier to a quiet place where it will not be disturbed by human noise or movement. If it is hot outside, find a shady, well- ventilated place. If it is cool outside, consider leaving the carrier inside.

7. Clean and disinfect the work station. See Appendix 5 for cleaning protocols.

8. Enter the duck on the Sick and Dead LADU Intake Log (Appendix 9) and assign it a log number.

7

4. CONTINUING CARE There is not a standard hydration and feeding schedule that works for every bird, as it depends on the severity of the symptoms and the individual’s body condition. Assess the bird each morning and create a hydration and feeding schedule according to the following guidelines.

Just as in the intake exam, take and record the duck’s temperature, weight, dehydration level, and body condition score every morning in a quick and thorough manner. Create a hydration and feeding schedule based on the bird’s current condition.

SITUATION 1 – Bird is unable to hold its head up:

. Do not administer fluids orally; birds at this stage are very likely to aspirate.

. Give 10 mL sterile LRS or NaCl solution SQ in the inguinal fold of each leg (for a total of 20 cc per administration).

. You can give SQ fluids at 3 to 4 hour intervals for debilitated birds.

. Check if the bird has absorbed the fluids (by seeing if the bolus has disappeared) before giving any more.

. If the bird has not absorbed the fluids, make sure it is within its normal temperature range (102°-106° F); a cold bird will not be able to absorb fluids. If necessary, offer a heating pad (Instructions, p. 11).

. Continue to offer a rolled-up towel on which the bird can support its head.

. The duck’s fecals will be runny– this is normal.

. Do not give dishes of food and water in the carrier. With no control over their head, it is possible they can drown in their own water dish.

. Do not give a second dose of anti-toxin; one dose is the maximum.

. Never inject non-sterile fluids or fluids not approved for SQ treatment.

SITUATION 2 – Bird is able to hold its head up, and is not emaciated (has a body condition score of 1+ or higher):

. Try administering fluids PO. Tube-feed in 5 mL increments while watching closely for regurgitation and give no more than 10-15 mL.

8

. Later, if the status of the bird is the same or improved, the next tube-feeding can be up to 15 mL of Ensure followed by another administration of LRS/NaCl at least four hours later.

. At this time, examine the fecal matter in the carrier. If it looks exactly like it appeared before going into the duck, the bird is not digesting it. In this case, keep administering LRS/NaCl one or two more times before trying Ensure again. Only after the duck shows that it can digest it properly should you transition to feeding Cat Food Slurry.

. Offer a small food and water dish in the pet carrier to see if the bird will self-feed. If the bird is eating on its own, you can minimize tube-feedings unless the bird is losing weight or getting dehydrated. This is ideal since it saves the bird the stress of multiple handlings. However, Laysan Ducks are often too stressed in captivity to self-feed so this is not always an option.

SITUATION 2A. Bird is able to hold its head up, is digesting properly and is thin:

. Tube-feed the bird Cat Food Slurry up to 3 times a day, as long as you think the bird is getting enough water from the water dish on its own (assess hydration). If not, replace one of these tube-feedings with LRS or NaCl. Wait a minimum of 2 hours after LRS/NaCl feedings and 4 hours between slurry feedings.

SITUATION 2B. Bird is able to hold its head up, is digesting properly and is in moderate or good body condition:

. Tube-feed the bird Cat Food Slurry 2 times a day, as long as you think the bird is getting enough water from the water dish on its own (assess hydration). If not, add in a LRS or NaCl tube-feeding. Wait a minimum of 2 hours after LRS/NaCl feedings and 4 hours between slurry feedings.

SITUATION 3. Bird is able to hold its head up, and is emaciated (has a body condition score 1 or less):

 Before feeding a starving or emaciated seabird, ensure that the bird’s temperature is above 100°F before continuing. If the bird’s temperature is dangerously low, place a heating pad underneath the pet carrier (follow instructions on p. 11). Monitor the bird closely by visually watching the bird as well as taking its temperature often.

 Follow one of the suggested protocols from Lafeber:

Treatment Protocol 1 – Starvation Quantity per tube feeding – 7% of body weight

9

For instance a 200-gram bird could be gavaged up to 14 ml of fluid. Administer the fluid very slowly, and stop if fluid begins to well up in the back of the throat. If the bird is unable to hold 7%, then gradually work up to this volume.

Administer hydrating fluids (LRS or NaCL) PO or SQ 1 hour later: 25% dilution of Emeraid Omnivore 2 hours later: LRS/NaCl (PO or SQ) 1 hour later: 35% dilution Emeraid Omnivore 2 hours later: LRS/NaCl (PO or SQ) 1 hour later: 45% dilution Emeraid Omnivore 2nd Day* LRS/NaCl (PO or SQ) 1 hour later: 55% dilution Emeraid Omnivore 2 hours later: 65% dilution Emeraid Omnivore 2 hours later: 75% dilution Emeraid Omnivore 2 hours later: 85% dilution Emeraid Omnivore 2 hours later: 100% Emeraid Omnivore, made as directed per label instructions 2 hours later: LRS/NaCl (PO or SQ) NOTE: If bird is still dehydrated, you may add another LRS mid-day replacing one Emeraid tube feeding.

Treatment Protocol 2- Severe Starvation Quantity per tube feeding – 7% of body weight (kg) For instance a 200-gram bird could be gavaged up to 14 ml of fluid. Administer the fluid very slowly, and stop if fluid begins to well up in the back of the throat. If the bird is unable to hold 7%, then gradually work up to this volume.

Administer hydrating fluids (LRS or NaCL) PO or SQ 1 hour later: 10% Emeraid Omnivore 2 hours later: LRS/NaCl (PO or SQ) 1 hour later: 20% Emeraid Omnivore 2 hours later: LRS/NaCl (PO or SQ) 1 hour later: 30% Emeraid Omnivore 2 hours later: LRS/NaCl (PO or SQ) 2nd Day* LRS/NaCl (PO or SQ) Same as Day 1, except give one LRS tubing first, and then give 40% Emeraid Omnivore feedings all day. 3rd Day* Assess bird’s overall condition Same as Day 2, except give one LRS tubing first, and then give 50% Emeraid Omnivore feedings all day. In rare cases, it may take up to 4-5 days for a severely emaciated bird to tolerate Emeraid Omnivore as directed on the label, and possibly 5-6 days to eat solid food.

10

GENERAL CAPTIVE CARE GUIDELINES:

 When opening the carrier door to bring out the duck, always be inside a contained room such as the duck lab or the outside aviary in case the duck escapes. To prevent escapes, open the door slowly and block the opening with one hand while you reach in with the other hand.

 Keep the plumage clean by using the wire mesh “poop guards” in the pet carrier to keep the bird from sitting in its feces.

. Also, periodically check the vent area. If it is dirty, put the duck in the “bird burrito” (Figure 2), bring it to the duck lab sink, hold it upside down and run water over the cloaca while wiping off the feces. Towel off the excess water near the vent and tail.

 If the bird is weak and immobile, you can help the duck by manually expelling its feces. You can do this on each handling prior to SQ or PO administrations. Put the duck in the “bird burrito,” place it on its back, and press gently on the area anterior to the cloaca to expel any fecal matter present. Clean the bird’s feathers off. Pat yourself on the back, you just helped a duck poop.

 Replace dirty pet carriers, “poop guards,” and absorbent pads with clean ones every morning and evening. Change out food and water dishes at least once a day unless otherwise soiled.

 Examine the bird’s fecals and note on the captive care datasheet the color and consistency to help you keep track of its digestive health.

 Do not keep birds loose in the aviary since they tend to lose weight quickly that way. Continue to keep them in a pet carrier in a quiet, undisturbed place.

 If the bird is improving and you anticipate releasing it soon, give the bird the Botumink vaccine. The bottle of Botumink is kept in the fridge. The dose is 1cc and is given SQ in the inguinal fold. Wipe the top of the vaccine bottle with alcohol before introducing the sterile single-use needle.

Remember: The botulism anti-toxin is given IM (p. 4) while the botulism vaccine is given SQ; it is easy to confuse the two.

 If the bird is improving and at a good weight, consider release. See Appendix 5 for release criteria.

 If the bird is losing weight and/or not improving within a few days, report to Refuge Biologist with any other symptoms and seek guidance as appropriate.

11

5. REFERENCES This protocol drew heavily from existing written protocols and personal communication with Hawaii Wildlife Center, which were extremely helpful in this attempt at creating a tailored procedure for Midway Atoll NWR. Many thanks go to Linda Elliott, Judi Ellal, Tracy Anderson, Dan Clark, Bret Wolfe, Greg Schubert, Meg Duhr- Schultz, and Ann Humphrey for their support and guidance.

Avian Botulism Management at Midway Atoll NWR. John Klavitter and Bret Wolfe, 2014.

Hawaii Avian Botulism Guidelines. Hawaii Wildlife Center in partnership with Hawaii Division of Forestry and Wildlife.

Laysan Duck Avian Botulism Bird Collection, Stabilization, Transportation, and Rehabilitation, Standard Operating Procedures. Hawaii Wildlife Center, 2014.

12

APPENDIX 1 – Intake Exam: Condensed Version This simplified version is designed to help you remember the steps as you perform the intake exam. However, it is not a replacement for reading the full version (Section 2 and Appendix 1). It is essential that you are familiar with the details of each step before attempting them.

Materials Needed: Pet carrier, absorbent pads, poop guard, towel Hot water bath Appropriate syringes with gavage tubes or sterile needles Thermometer Duck bag, pesola spring scale Captive Care datasheet Defrosted anti-toxin, if appropriate

Intake Exam: 1. Prepare pet carrier.

2. If not administered in the field, defrost an anti-toxin syringe in your hand and administer 0.5 cc of it to the duck intramuscularly (IM) in the breast muscle.

3. Prepare and warm fluids. I. Draw up 20 cc in a sterile luer-lok syringe with a sterile 23 or 25 gauge needle to give subcutaneously (SQ). II. Place syringe in hot water bath to warm.

4. Take and record temperature, weight, body condition score and dehydration level. . Is the bird’s temperature lower than 102° F? If yes, consider placing bird on a heating pad after the exam (important instructions on page 10). . Is the bird’s temperature higher than 106° F? If yes, wipe feet down with alcohol pads and place in a shady, well-ventilated place after the exam.

5. Administer SQ fluids to the bird by injecting 10 cc into the inguinal fold of each leg with a sterile 23 or 25 gauge needle.

6. Place bird in pet carrier and relocate to a quiet place and make sure the ambient temperature is appropriate.

7. Clean and disinfect equipment and work station.

13

APPENDIX 2 – Body Condition Score The body condition score is a useful tool for assessing a bird’s overall health. Note a patient’s score on the captive care datasheet and use it to help determine when to release a bird.

(Keel)

NOTE: If the body condition scores in between two numbers on the scale, you can use “+” and “-“ to convey this. For example, if the score is in between a 1 and 2 and

is closer to a 1, use 1+.

Source: Kaytee Diets (http://www.kaytee.com/assets/014/27187.pdf)

14

APPENDIX 3 – Diet preparation and storage It is important for the duck to be well hydrated and maintaining or gaining weight while in care. Use the following diets when caring for a duck. Refer to the Continuing Care section to determine which diet to use.

A. Lactated Ringers Solution or 0.9% Sodium Chloride Solution – These are rehydrating solutions that can be given PO or SQ. Either solution will work; use what is available. Give 15 cc maximum for PO, and 20 cc maximum for SQ. See Appendix 1 for more information on administration.

 Storage:  For bags that have been cut open for oral administration, pour the contents into a container with a lid and refrigerate. Label the container with the open date and throw out the contents after one week.

 For SQ fluid bags that have only been entered by sterile needles via the port at the bottom with the yellow stopper, store the bag in the fridge. Label the bag with the date of first use and throw out after one week.

 For intravenous (IV) administration, the contents in the bag are only good for 24 hours. However, fluid administration via IV is not part of this protocol so you will not have to worry about this.

 NOTE: If you observe tiny particles or floating material in these solutions (even if they have never been opened), they are no longer good. Throw these out.

B. Ensure Nutrition Shake, plain or vanilla flavor, 8 oz bottle – This is given PO. It provides some nutrition in addition to hydration and is a good transition between a diet of fluids only to a diet of fluids and Cat Food Slurry. Gavage up to 15 cc of Ensure per feeding.

 Storage: Label the bottle with the open date, refrigerate after opening, and dispose of after 24 hours.

C. Cat Food Slurry – Before transitioning to slurry, make sure the duck is digesting the Ensure properly. Use potable water and approved cat kibble (Evo or Taste of the Wild brands) to make the slurry. Gavage up to 15 cc of slurry per feeding.

 Store the bag of cat food in the freezer to prevent it from going bad.

 Preparation instructions:

15

 Using the coffee grinder, grind approximately 1 cup of cat kibble as fine as possible without causing it to become heated in the process (this affects the nutrients). The coffee grinder should do a good job of grinding it into a fine powder. However, if there are any large chunks of kibble, sift these out as these can clog the tube when gavaging.

 Combine equal amounts of water and cat food together. For example, add 3 TB ground cat food to 45 mL water. Mix in the blender. You may have to double the amounts of both ingredients if the blender cannot mix a smaller amount. Blend well, but do not allow the slurry to heat up. Add more water if the slurry is too thick to draw into the syringe and blend again.

 Note: If you are extra careful to mix the slurry thoroughly and break apart any clumps, you can mix by hand with a syringe plunger or spoon in lieu of using the blender. The goal is to not plug up the tube with solid clumps when you are gavaging the duck.

 Draw slurry in to the feeding syringe, attach the gavage tube and remove air and excess amounts of slurry. Warm in hot water bath. Feed as soon as possible after the slurry has warmed since it will swell/solidify with time. Hold the feeding tube onto the syringe when dispensing into the bird so that it does not pop off.

 Leftover slurry should be discarded at the end of each feeding.

D. Self-feeding diet – If the bird is holding its head up, place a small dish of food and a small dish of water in its carrier. Offer a combination of Trip-l duty poultry feed, cat food kibble moistened with water, and dried mealworms. You can also supplement this with Bunchgrass (Eragrostis variabilis) or Grass (Eleusine indica) seeds if you have time to collect them. All the dry food is stored in the biology freezer located in the duck lab. Dishes of food and water can be left in the carrier the entire day unless they become soiled.

 If the bird is stressed, it may not self-feed and will require tube-feedings of slurry to maintain its weight.

 NOTE: Be aware that mice can be attracted to the duck food. To prevent them from gaining access to it, place the pet carrier on a sturdy structure that is more than one foot off the ground, is not touching any walls or cords, and is slick on the surface (to prevent the mice from climbing the sides). A metal table or other similar structure is ideal.

16

E. Emeraid Omnivore/Psittacine – This is a critical care diet meant for critically ill or emaciated birds (birds with a body condition score of 1 or less). It is easy to assimilate because it is in a form that does not require the patient’s valuable energy stores in order to digest. This allows the debilitated animal to quickly get the essential nutrients for survival.

 Storage: Upon arrival, store Emeraid in a cool, dry space, and refrigerate powder after opening package. Emeraid has an 18-month shelf life from the time of production, and should be used within 9 months after opening. Handle product carefully and avoid contamination, which can lead to bacterial overgrowth; specifically, remember to keep the Emeraid dry, use a clean, dry spoon whenever using the product, and do not cross-contaminate with scoops used for non- Emeraid foods.

 Preparation:  Fully concentrated Emeraid Omnivore is essentially a 3:2 powder to water ratio. Therefore, for smaller volumes of Emeraid, measure 6 teaspoons (and blend with 20 ml of water) or, for larger volumes, measure 6 tablespoons (and blend with 60 ml of water) to get a 100% concentration. You can prepare however much you need as long as it works out to a 3 part powder to 2 part water ratio. You can then dilute it down to the appropriate concentration or you can also simply follow the amounts given in Table 1, which will give you the dilutions for commonly used amounts when gavaging. Refer to the Emeraid Protocols on p. 7-8 for instructions.

# Scoops of Appx. End Desired Emeraid Powder (1 Amount Amount Concentration scoop = 1 Water (mL) (mL) tsp) 25% 1.5 26 27 50% 3 25 28 75% 6 30 33 100% 6 20 30 Table 1. Amounts of Emeraid Omnivore powder and water to use for different concentrations.

 Measure level scoops; do not pack the powder. Mix the powder with hot (not boiling) water to help it dissolve. When mixing the Emeraid and water, shake or stir the mixture vigorously to avoid any danger of hot spots. Shaking dry powder in the mixing container will facilitate mixing; shaking helps to break up the clumps and lumps prior to adding the water. Make sure all lumps and clumps are blended.

 Keep the gavage tube and syringe in very warm water until ready to feed. The temperature of the Emeraid mixture should be on the hotter side of warm, but not too hot; if you can dip the tip of your pinky finger in the mixture and it feels comfortably

17

warm, that is good. The hot water also warms up the gavage tube and makes it more supple to work with.

 When ready to feed, fill the syringe with the Emeraid mixture and ensure that the duck is ready to feed immediately. To minimize any risk of the tube plugging, do not load the feeding or gavage needle with formula until immediately before feeding. When feeding, always watch for regurgitation. Prepare fresh food if more than 30 minutes passes. Discard any leftover formula. Make fresh Emeraid for each feeding; do not feed birds Emeraid Omnivore for more than 7 days.

18

APPENDIX 4 – Release criteria and banding procedure After the effects of the botulism toxin have resolved, the most important factors to consider when releasing a bird are body condition, weight, and stress level.

The duck in care should meet the following criteria as closely as possible before release:

1. Able to hold its head up, stand, and walk.

2. Holds wings in the correct position.

3. Able to function nictitating membrane (the bird’s third eyelid).

4. Properly hydrated; eyes are no longer squinting.

5. Good body condition (with ideally a minimum of 2- for the body condition score) and is within the normal weight range according to its gender and age. NOTE: Mass fluctuates significantly with season and reproductive status.

. Average weight ranges for this species are:  Male – 450 g  Female, non-ovulating – 460 g  Female, ovulating – 550 to 600 g  Female with brood – 377 g  Hatch-years and young second-years can be below 350 g

6. Vaccinated with Botumink (instructions on p. 9).

7. Banded with the band combination being recorded on the Captive Care Notes datasheet.

. If the Refuge Biologist is available, an aluminum USGS band and an auxiliary band can be put on the bird. Each of the bands should have directions on which leg they need to be placed on. Record the new band combination and numbers on the banding datasheet.

. If the Refuge Biologist is not available, temporary band(s) can be placed on the bird instead. Double check the resight reference file so that you do not duplicate an existing color combination. If the bird has received the botulism vaccine, secure the temporary bands with super glue so that survival can be tracked.

. Upon release, be sure to update the Sick and Dead LADU Intake Log, the Sick and Date LADU database (excel database), and the resight reference file to ensure that the data have been accounted for.

19

APPENDIX 5 – Cleaning and disinfecting Keeping the duck lab clean and disinfected is important in minimizing risk to human and bird health. It also helps prevent insect infestations of ants, roaches and mice.

 Wash hands frequently.

 Spray and wipe the bench and any other soiled surfaces with 10% bleach solution.

 After an injection, dispose of the needle in the sharps container and the syringe in the garbage can. Neither of these objects should be reused.

 Wipe the top of the Botumink vaccine bottle with alcohol before introducing each sterile needle.

 After an oral administration of fluids or slurry, wash the syringe and gavage tube with dish soap and water to remove all food and oily residues. Rinse thoroughly. Make sure remnants of slurry do not remain stuck in the tube; you may have to use the syringe to clear it out. Use chlorhexidine or vercon solutions to disinfect. Completely submerge each object in the solution for at least 10 minutes and rinse with water. Draw solution into gavage tubes to make sure the disinfectant makes contact inside the tube.

 Chlorhexidine solution: 2 Tablespoons per gallon of water. Avoid contact with eyes, ears and mucous membranes. Replace solution every few days.

 Vercon solution: One tablet per pint of water. Use gloves when reconstituting: while the solution is non-toxic and non-irritating, direct contact with the tablet can irritate skin, mucous membranes, and eyes. Replace solution after one week.

 If neither of these are available you can use a 10% bleach solution, although the bleach will shorten the life of the plastic syringes and feeding tubes.

 Never use the same gavage tube twice without cleaning and disinfecting it first.

 Allow clean syringes and tubes to dry thoroughly before putting them away. Trapped moisture can encourage bacterial growth.

 Syringe plungers will become more difficult to depress over time. Throw out ones that take a lot of effort to push.

 Replace any food and water bowls as they become soiled. Clean and disinfect these as well.

20

 Bring dirty pet carriers outside and use the hose, 10% bleach solution and scrub brush to clean. Let the bleach solution sit for at least 10 minutes before rinsing. Let the carriers dry thoroughly. Throw away the used absorbent pads.

 For poop guards covered with fabric, don’t use bleach to clean them as this will shorten the life of the fabric. Let them soak in a tub with chlorhexidine or vercon solution instead.

 Use the washing machine in the store room to clean the dirty duck bags and towels. There is a limited supply of clean duck towels so make sure you have enough clean ones for the next day.

21

APPENDIX 6 – Disposition of Carcasses Laysan Duck carcasses are periodically sent to the USGS National Wildlife Center – Honolulu Field Station for necropsy at the discretion of the Refuge Biologist (Midway Atoll NWR) and Thierry Work (USGS). All other collected LADU and other waterbird carcasses are stored in the freezer and incinerated at a later date.

Place carcasses in the respective designated containers inside the LADU freezer. There is a container for carcasses to be incinerated and a container for carcasses to be sent for necropsy.

Good candidates for necropsies:

 Were found freshly dead. If carcass is soft and mushy, has skin discoloration, feathers or skin that easily rub off, or if maggots are present then it is too decomposed for most lab tests.

 May have died from causes other than botulism. We are interested in other causes of mortality besides avian botulism.

A few fresh carcasses suspected to have died from botulism should be saved each year for testing and confirmation that botulism is present on Midway. These should be placed in the designated container with the other necropsy candidates.

22

APPENDIX 7 – Resources

 For any rehabilitation support and questions:

Hawaii Wildlife Center Phone: 808-884-5000 Email: [email protected], [email protected] or [email protected]

 For disease and carcass related questions:

USGS – National Wildlife Health Center, Honolulu Field Office Phone: 808 792-9520 Email: [email protected] or [email protected]

 For general Laysan Duck species and monitoring questions:

USGS – Kilauea Field Office Phone: 808-985-6416 Email: [email protected] or [email protected]

23

APPENDIX 8 – Datasheets Sick and Dead LADU (and other waterfowl) Intake Log - MIDWAY ATOLL

Log #

toxin (Y/N)toxin Disposition (assigned Location - Date Time Obs. Species Date & Notes to LADU Found Age Status Location

only) Disposition

Sex F, U) (M,

Left Leg BandLeft Leg

Right Leg BandRight

Botulism Vaccine (Y/N) Vaccine Botulism

Suspect (Y/N) Botulism

Botulism AntiBotulism

24

APPENDIX 8 – Datasheets Sick and Dead LADU (and other waterfowl) Intake Log - CODING SHEET Log #: A unique number given to any live LADU brought in to the lab or dead LADU collected. Format: Year-sequential number (e.g., 15-01 for the first bird of 2015). Birds other than LADU do not get assigned numbers. Date: Date of capture or collection. Time: Time of capture or collection. Obs.: Observer's initials. Species: The four letter species code of bird. Common species include: LADU - Laysan Duck NOPI - Northern Pintail AMWI - American Wigeon GRSC - Greater Scaup For other species, refer to Alpha Code document in 'Vagrant' folder in the Volunteer Drive Location Found: General location where duck was found, e.g. seep name, building name or sector number Status: The condition of the bird when found. Use codes: D - Dead S - Sick I - Injured E - Entangled (e.g. with plastic debris around head or feet) O - Other (explain in notes) Age: Approximate age of bird. Use codes: D - Duckling (still downy) HY - Hatch Year (a young bird hatched earlier in the year that is independent and capable of flight). SY - Second Year AHY - After Hatch Year (includes SY and ASY ages, use AHY if cannot specify) ASY - After Second Year (adult bird) U - Unknown See LADU ageing/sex guide for details on age-specific plumage. Sex: Male, Female or Unknown. Refer to LADU ageing/sex guide for details. Right Leg Band: If bird is banded, record the band for the right leg. This may include the color and symbol for a plastic auxiliary band or simply AL for an aluminum band. Record the aluminum band number in the note column. Left Leg Band: If bird is banded, record the band for the left leg. This may include the color and symbol for a plastic auxiliary band or simply AL for an aluminum band. Record the aluminum band number in the note column.

25

APPENDIX 8 – Datasheets Sick and Dead LADU (and other waterfowl) Intake Log - CODING SHEET Suspect Botulism For live birds: write 'Y' if bird is exhibiting symptoms typical of botulism (drooping wings, squinty eyes, paralysis). Write 'N' if unsure, or if another cause (Y/N): is obvious.

For dead birds: write 'Y' if bird was found in an area that has been a hotspot for botulism within the past few weeks. Write 'N' for if unsure, or if another cause is obvious. Botulism Anti-toxin Applicable only for birds found alive. Write 'Y' if this bird received the botulism anti-toxin (IM injection). For birds that later die in care, this can (Y/N): eliminate them from being considered for necropsy. Botulism Vaccine (Y/N): Applicable only for birds found alive. Write 'Y' if this bird received Botumink, the botulism vaccine (SQ injection). Disposition: Ultimate fate of live bird or carcass. Use codes: 1 - Found dead, kept carcass in freezer for incineration. 2 - Found dead or died in care and is a candidate for necropsy. Note: if given anti-toxin, the bird is not eligible for necropsy. Carcasses qualifying for necropsy include ones that: ◦ Are relatively freshly dead (no maggots). ◦ Were not given botulism anti-toxin while alive. 3 - Held for treatment, was released. 4 - Held for treatment, died in care, put in freezer for incineration.

5 - Other (explain in notes). 6- Sent to USGS for necropsy. Disposition Date Ultimate location of bird and accompanying date. For example, "Date: 9/1/2014, Location: Office Guzzler" for a released duck, or "Date: 10/5/2014, and Location: Location: LADU Freezer" for a carcass kept for incineration, or "Date: 11/5/2014, Location: LADU Freezer for USGS" for carcasses to be sent for necropsy. For carcasses that have been sent for necropsy, change location to "USGS.” Notes: Include any noteworthy details such as bird behavior, presence or absence of maggots, condition/freshness of carcass, keel score of carcass if fresh (refer to sheet on the wall in the lab for scoring), where the carcass was found, band condition and aluminum band number if applicable.

26

APPENDIX 8 – Datasheets

LADU CAPTIVE CARE NOTES Page ____ of _____

Log #______Intake Date/Time/Location ______Sex ______Age ______Band Combination RL/LL ______/______USGS Band # ______

Weight Weight Body Dehydr- Obs. Date Time Temp. (bird in (bird Condition ation Fluids or Nutrition Given Notes bag) only) Score Level

Release Date/Time/Location______27

APPENDIX 89 – DatasheetsLADU Initial Intake Flowchart

11

Release Date/Time/Location______28

FINAL ENVIRONMENTAL ASSESSMENT MIDWAY SEABIRD PROTECTION PROJECT APPENDIX F

APPENDIX F. PUBLIC COMMENTS AND AGENCY RESPONSES

F-1

Appendix F. Comment response matrix. Where appropriate, comments are paraphrased, or the comment subject line is included. In the remaining instances, sections from comment letters are included in entirety and enclosed in quotation marks.

Relevant DEA # Commenter Comment Summary or Quote Sections USFWS Response 1 Barbara Mayer Supports the project. Comment noted. 2 Barbara Mayer One suggestion: On p. E3 (and probably elsewhere Page ES- On p. ES-1, paragraph 4, it states in the document as well) you say, "Proposed Action: 1. that: “The other islands of MANWR, Eradication of invasive mice on Sand Island, Eastern and Spit, are not included MANWR." I think it might be wise to re-word the because mice are not currently statement to say something like, "...Eradication of present on them.” Those islands have invasive mice on Sand Island and all other terrestrial remained mice free for many decades areas where mice are seen, within/at MANWR". and so it is very unlikely that they Perhaps you've given a reason, somewhere in your will have mice any time soon. document, for not including Eastern and Spit...other Appendix A, page A-1, briefly than the fact mice aren't there at the moment. What summarizes the refuge’s biosecurity if mice turn up there before or during your plan. The entire plan, which is too eradication effort? How bureaucratically lengthy to add, sufficiently addresses complicated would it be to broaden your effort at measures for preventing the spread of that point in time? mice to nearby Eastern and Spit Islands. These measures are already in place to ensure mice are not accidentally transported off Sand Island. Importantly, MANWR’s Biosecurity Plan includes early detection measures that are taken to alert staff of a possible introduction, thereby preventing an invasion. 3 Catherine “I believe everything possible must be done to N/A Comment noted. Greenleaf protect these seabirds from predatory behavior by the mice at the Midway Atoll. There are already far too many elements adversely affecting the survival of these seabirds, such as the Albatross, including plastic pollution in the ocean, longline entanglement and depleted fish reserves. Please put this project into action to rid the area of mice and to protect the seabirds.” 4 Caren Loebel- “It is crucial that the Midway mouse eradication N/A Comment noted. Fried move forward quickly. The albatrosses need the safe nesting grounds of Midway Atoll. The albatrosses were nesting on Midway long before we humans brought mice and other destructive invasive species there. It is our responsibility to give to the albatrosses their safe haven on Midway.” 5 Danielle Frohlich “I’m writing in support of the proposed action to 3.3.6.4.3 Comment noted. Lead Botanist/ eradicate house mice from Sand Island, Midway by Invasive Species aerially broadcasting rodenticide. Not only will this Program action benefit the albatross and other seabirds on Specialist, Kauihelani, it will also improve the odds of SWCA restoring coastal habitats by reducing seed predation Environmental on native Hawaiian plant species.” Consultants 6 Dr. Fern P. Duvall Supports the project. N/A Comment noted. II, Department of Land and Natural Resources, Division of Forestry and Wildlife 7 “ “ “Add the results of the successful 2016 Antipodes N/A This reference was added to the FEA. Island eradication effort just assessed and published.” 8 “ “ “I fully concur that use of rodenticides to do the 2.2.2 Comment noted. eradication is the method to employ.”

9 “ “ “I fully agree that the Preferred Alternative where 2.3 Comment noted. multiple distribution methodology is outlined should be the alternative utilized.” 10 “ “ “Timing of aerial drops should coincide with the 2.3.5, The timing of the bait aerial drop has time of year where fewest shorebirds of concern are 3.3.6.6, to balance minimizing the impacts to present – especially for the shorebirds species of 3.3.6.7 multiple natural resources and conservation concern. I think that capture and species, including the Laysan duck feather clipping of shorebirds should be avoided to and Laysan albatross. The timing is the fullest extent possible. IF UTILIZED it must currently planned to occur during the only be done for species, and at a time, which will lower period of abundance for allow for all clipped feather to be shed and replaced shorebirds. In addition, due to the risk with new feathers in time to allow for normal of exposure to Bristle-thighed curlew migration to occur. If this cannot be met, I believe (BTCU), as many birds as possible that shorebirds should be left where they are, and are planned to be trapped and any losses that might occur be attributed to the drop maintained in captivity. Maintaining as incidental take. I think this would have less BTCU in captivity and then releasing negative impact that capture, clipping (or plucking), them was successfully carried out for and relocation to a separate island location or the Palmyra rat eradication project. It holding-captivity for the shorebirds.” is not known how long these birds may need to be kept in captivity since birds will only be released when monitoring of bait residues indicates it is safe for birds to be released. 11 “ “ “The stepwise capture-holding-and release of 3.3.6.8 Comment noted. A combination of Laysan Ducks [see 3.28 of 100% of population (or a mitigation strategies are being majority of the population) with: (i) hold in implemented for the Laysan duck captivity on Sand Island and then Eastern Island throughout and following the project. after operation commences until the risk period passes; (ii) clip wing feathers of birds on Sand Island and translocate to Eastern Island; (iii) a combination of the above strategies] seems like the best alternative to pursue.” 12 “ “ Concurrent to that work with Laysan Ducks it 3.3.6.8 Bait will be not be broadcast over should be assessed whether, for all wetlands where wetlands or bodies of water. Smaller water is artificially maintained, they should be ponds will be covered to prevent any concurrently stepwise drained and kept dry to better application into water. The three control for bait-avoidance of aquatic invertebrates, existing seep ponds on Eastern Island etc., and to observe and manage post-drop bait will be filled with sand or covered survival (e.g. to allow possible collection/removal with shade cloth to prevent ducks of remaining bait before re-wetting occurs.) from accessing the water to reduce conditions that foster botulism outbreaks. 13 E J Peiker, “While I feel that there is merit in the eradication of 3.3.6.8 This project will benefit Laysan Professional mice, I also feel that the risk of the project, as Ducks by removing a known food Photographer specified in the Draft Environmental Assessment is competitor, the house mouse, and too high for the critically endangered Laysan Duck removing a species that may indeed and the mortality rates discussed within the prey upon duck eggs or ducklings. document are unacceptable for the long term Moving ducks to Eastern Island or viability and well being of the species. None of the keeping them in captivity will be mitigation proposals are strong enough, in my safer for individual ducks than opinion, to ensure the long term safety of the moving them to another site because Midway Atoll population of Laysan Duck. I there are few appropriate sites with strongly urge the decision makers on this project to no mammalian predators and/or no either: suitable habitat available. We will not A. Come up with a different strategy to eliminate be able to translocate birds off of the mouse population without changing the Midway Atoll at this time due to ecosystem in a way that the Laysan Duck can be regulatory constraints. harmed either in the short term or long term through introduction of this poison into the food system that this species relies on. B. If A. is not possible, the complete translocation of the species to another viable island away from Midway Atoll should be considered and funded. This could be a permanent translocation or at minimum, one that lasts long enough to reduce the toxin on Midway dramatically below the levels where it can harm this species. That may be for several years. Simply moving the ducks from Sand Island to Eastern Island does not offer enough protection in my opinion. The goal should be to harm as close to zero Laysan Ducks as is humanly possible!” 14 Linda M. B. Paul, “The Hawai1i Audubon Society supports the N/A Comment noted. President, Hawaii objectives and preferred alternative for the Audubon Society eradication of mice from Midway Atoll as set forth in the USFWS Draft Environmental Assessment (DEA).” 15 “ “ “While the use of poisons always raises concerns, 3.3.6.5.9 There are three species of albatross we have reviewed the proposed methodology set on MANWR, the short-tailed, black- forth in the DEA and agree with the steps that the footed, and Laysan, and the potential USFWS proposes to take. Since albatross fledging risk for rodenticide poisoning differs season ends in late July, we trust that the rodenticide among them. Black-footed albatross will not be applied until August 1st.” and the federally-listed short-tailed albatross fledge chicks through June; the young albatrosses that do not leave the island by late June/early July naturally succumb to heat exposure and/or starvation, as the adults no longer return to feed them. Laysan albatross chicks fledge through July with ~50% of chicks gone by mid-July; any chicks not fledged by late July/early August would most likely naturally succumb to the same fate as short-tailed and black-footed albatross chicks in late June/early July.

Laysan albatross chicks that are present would be at risk of ingesting bait pellets off the ground via pica behavior. However, the first application of bait pellets would likely be rapidly consumed or cached by mice and cockroaches. Subsequent applications would occur after most albatross chicks have fledged. Also, not all chicks would pick up the bait, and of those that do, not all would ingest significant amounts because their main diet is fish and squid that is provided by the parents. It would be unlikely that Laysan albatross chicks would be killed or sub-lethally affected by primary poisoning. 16 Jeff S. Hatfield, “I commend you on your plan to eliminate mice on 3.3.6.8 Our intent is to capture every duck PhD. Sand Island, as I believe it will eventually benefit all that we can and keep 100 males and Research of the native species found there. Concerning the 100 females in captivity and put all Ecologist, short-term impact of the mouse eradication on the additional birds captured on Eastern USGS Patuxent Laysan duck, I believe your plan, to save 50 males Island with their flight feathers Wildlife Research and 50 females to repopulate the island after the clipped to preclude them flying to Center mouse eradication, is probably adequate, if the Sand Island until they molt again. population of ducks increases rapidly (as expected) after mouse eradication and the ducks are released back to Sand Island. It would be best if you could capture and protect all of the ducks currently on Sand Island prior to rodenticide use, but this is probably not practical.” 17 “ “ Please consider the following 3 suggestions, if you 3.3.6.8 1. Extensive habitat restoration has would like to do more mitigation to offset the effect been undertaken at Midway since it of the mouse eradication on the population size of was designated a National Wildlife the ducks on Midway Atoll: Refuge in 1988, including the control 1. Restoration of native vegetation that can be and in some cases the eradication of utilized by the ducks, in areas of suitable habitat, non-native plants that competed with would help the species increase more rapidly after native species and diminished the impending bottleneck on the population size due diversity. Ironwood (Casuarina to rodenticide use. This may actually increase the equisetifolia) have been eradicated on carrying capacity of the island for Laysan ducks Eastern Island and over 20 acres have over the present carrying capacity, if enough habitat been removed from Sand and native vegetation are restored. This would Island. 1,045 acres of Golden ultimately lead to a larger global population size of crownbeard (Verbesina ducks, and lower longterm probability of extinction encelioides) have been removed of the species. throughout the atoll. Both of these 2. Avian botulism has been known to be a problem species form monocultures and for Laysan ducks on Midway Atoll for the past few provide little food or cover for years (Reynolds et al. 2017). Before releasing the Laysan Ducks. In concert with the ducks back on Sand Island after mouse eradication, management of invasive non-native please consider vaccinating them for avian botulism. plants MANWR has an active native A vaccine is already available and is expected to plant propagation and outplanting work in Laysan ducks. This would temporarily program. To date they have out- increase survival of the ducks and, again, allow planted and seeded 14 native species them to increase more rapidly after the population over 75 acres at the atoll creating bottleneck caused by rodenticide use. The quicker habitat with cover and much greater the population returns to around carrying capacity, food density for Laysan Ducks, thus the less genetic variability will be lost to the increasing the carrying capacity. population (Weiser et al. 2013). 3. Please consider moving some of the Laysan 2. All ducks captured will be ducks from Sand Island, prior to rodenticide use, to vaccinated for avian botulism and other islands in the Northwestern Hawaiian Islands sites that have been implicated in or to a captive breeding facility. If at least 50 ducks previous outbreaks of botulism are (say 25 males and 25 females) were moved to a being modified to reduce risk. captive-breeding population, for example, this would reduce the probability of extinction for the 3. No translocations or removal to species by an order of magnitude (Weiser et al. captive breeding programs are 2013, Figure 2). Currently, I am working on a possible for Laysan Ducks from Science Support Partnership (SSP) proposal, to fund Midway at this time due to funding, bringing 50 Laysan ducks back from Midway Atoll regulatory and compliance to PWRC in Maryland, constraints. where we recently upgraded our captive duck facilities to house such ducks. (We already have a large number of captive ducks of various species used in research.) We would like to conduct a study of the avian botulism vaccine on these Laysan ducks for 1-2 years post capture, then re-home them to 2-3 captive-breeding facilities after this research is finished at PWRC. Two such facilities have already expressed interest in taking these ducks if they come to PWRC. We would not harm the ducks during the vaccine trials, just test the ducks for antibody response using various vaccination schedules. This information should be useful if the vaccination of the ducks prior to release back on Sand Island is indeed an option. Starting a captive-breeding population, funded by private breeding facilities, will greatly reduce the probability of extinction for the entire species, and would also prevent these particular ducks from being exposed to rodenticides on Sand Island during the mouse eradication. If saving ducks is a priority before and during mouse eradication, please consider this suggestion to re- home 50 of them off island prior to rodenticide use. 18 Heather Johnson Supports the project. N/A Comment noted.

19 Hob Osterlund, Supports the project. N/A Comment noted. Founder, Kauai Albatross Network 20 Joe Mainardi Supports the project. N/A Comment noted. 21 JoEllen Arnold Supports the project. N/A Comment noted.

22 Jeffrey Walters, “There is no doubt that the ducks would ingest 3.3.6.8 We will keep every duck possible in Bailey Professor lethal doses of the toxicant to be employed should captivity or in forced residency on of Biology, they have the opportunity to do so. The mitigation Eastern Island during the operation. Virginia Tech plan is designed to eliminate that opportunity, which We will calculate the probability of is appropriate, but it seems virtually impossible that retaining all the alleles currently it will be as effective in doing so as portrayed in the present in this population as EA. The EA acknowledges that birds will be lost, described by Reynolds et al (2015) but I believe it greatly underestimates the number using the program AlleleRetain as that will be killed. It will not be possible to capture described by Weiser et al 2012. all of the ducks on Sand Island. These are secretive, nocturnal birds, so some are bound to avoid capture. Reynolds, M H, JM. Pearce, P. Some of the birds on Eastern will regain flight Lavretsky, P. Seixas, and K.N. ability unnoticed, and fly back to Sand. On Eastern, Courtot. Microsatellite variation and birds will be kept at unprecedented densities, rare alleles in a bottlenecked presenting both biological and husbandry Hawaiian Islands endemic: challenges. Despite supplementing food and implications for reintroductions. providing water, adverse impacts due to Endang Species Res. 28: 117–122. unsustainable densities are inevitable, some of doi: 10.3354/esr00681. which will result in additional deaths. There is considerable uncertainty about how long secondary Weiser E. L., C. E. Grueber, I. G. exposure risk will remain, and therefore it is likely Jamieson. 201. AlleleRetain: a that some individuals will encounter unanticipated program to assess management lethal exposure once the birds return to Sand. The options for conserving allelic monitoring effort required to totally preclude birds diversity in small, isolated being exposed in the ways I have outlined is populations. Mol Ecol. Resour 12: unachievable. The bottom line is that it is likely that 1161−1167. a large number of ducks will perish.” We will not release ducks back onto Sand Island until residue monitoring indicates brodifacoum levels are no longer detectable in duck prey species. 23 “ “ “It is assumed in the EA that the ducks will 3.3.6.8 We have no evidence that reproduce at high levels and thus recover their invertebrate densities were higher in numbers quickly, despite whatever level of Verbesina encelioides areas. mortality occurs, just as they increased rapidly in Verbesina is considered a poisonous numbers following the initial translocation. It is not plant to ungulates (Keeler et al 1992) reasonable to expect this to happen as Midway is and is allelopathic thus inhibiting all not the same as it was when the initial translocation other vegetation (Inderjit et al. 2000). occurred. Subsequent removal of invasive While many plants with chemical vegetation accomplished many Refuge management defenses against herbivory are objectives, but it appears that an unforeseen associated with herbivorous consequence was a reduction in the invertebrates arthropods that have co-evolved to that serve as food for the ducks. This reduction is withstand those defenses, in cases in presumably the reason why mice turned to attacking which the plant is not native, it may albatross. A likely additional consequence is that the have “escaped” that limit and become duck population is no longer capable of growing as more invasive in areas where its rapidly as it did when the ducks were initially natural herbivores do not occur. This introduced. seems likely to be the case for the Northwestern Hawaiian Islands I suggest that the mitigation strategy be altered to where the plant is extremely include translocation of a portion of the population successful but not the case for the off Midway. This would reduce the density of birds main Hawaiian Islands where it is held on Eastern, as well as spread the risk of more sparsely distributed. Verbesina mortality. Some birds could be moved to Kure, distribution and density was much which currently has a dangerously small population. reduced many years before mice were Better yet, other birds could be moved to Lisianski documented killing albatrosses at to establish another (fourth) population. This latter Midway. action seems to me the only way to provide a benefit to the recovery of the species commiserate Keeler RF, Baker DC, Panter KE. with the likely adverse impact on recovery that will 1992. Concentration of galegine in result from mouse eradication.” Verbesina encelioides and Galega oficinalis and the toxic and pathologic effects induced by the plants. Journal of Environmental Pathology, Toxicology and Oncology: Official Organ of the International Society for Environmental Toxicology and Cancer. 11(2):11-17

Inderjit, Chikako Asakawa, and, KMM Dakshini. 2000. Allelopathic potential of Verbesina encelioides root leachate in soil. Canadian Journal of Botany 77(10): 1419-1424, https://doi.org/10.1139/b99-097.

Oecologia (Berl.) 46, 22-31 (1980) Oecologia 9 by Springer-Verlag 1980. Feeding Patterns of Monophagous, Oligophagous, and Polyphagous Insect Herbivores: The Effect of Resource Abundance and Plant Chemistry, Department of Biology, The University of New Mexico, Albuquerque, New Mexico 87131 USA.

No translocations or removal to captive breeding programs are possible for Laysan Ducks from Midway at this time due to funding, regulatory and compliance constraints. 4 Kathy Dolson “You should read all the comments at the end of the N/A Comment noted. Washington Post article from 3/27. I guess the people offering suggestions there, did not understand how to properly submit their comments.” 25 Kim Rogers “Thank you for quickly addressing the growing 3.3.6.8 The approach of placing vulnerable mice problem on Midway. I support every effort in non-target organisms on nearby this regard. My main concern, however, is the Koloa islands while treating their habitat to duck. Has this kind of temporary remove rodents is a common practice. translocation/containment/wing-clipping been done The native Peromyscus found at Ana elsewhere with other birds, and if so, to what Capa Island was placed on an outcome?” adjacent islet during that operation. A number of species in the Galapagos and in the Caribbean have been held in captivity during rodent eradication operations. 26 “ “ “What about the dead mice? And how long might 3.3.6.8 During rodent eradications the they, if infected with botulism, be a host for the relative scarcity of dead animals on disease? I’m assuming that’s been factored into this the surface is quite surprising. When plan? So Laysan ducks are not released back into an the animal ingests the anticoagulant environment only to have to be faced with a and starts to hemorrhage, they tend to potential botulism outbreak.” go to their burrows and die underground. The influx of dead animal protein may cause a temporary elevation of botulism inducing bacteria but soon the lack of a continuous source of dead mice will possibly lower the baseline levels in the soil. 27 “ “ “My last concern is Wisdom. She has become the 3.3.6.5.9 Wisdom is not in an area that mouse “face” of Midway and Laysan albatross. What is attacks have occurred. Her location is being done NOW to protect her from mice? We in town where the buildings have CANNOT lose her to mice.” rodenticide bait stations around them. Due to this area having bait stations present even before the attacks began, this action has protected Wisdom. The area that Wisdom nests in has been continuously monitored for mouse attacks and none have been observed. 28 Louise Barnfield “I whole-heartedly agree that the suggested action N/A Comment noted. of eradication, and the method proposed, should go ahead as soon as feasibly possible.” 29 Loyal Mehrhoff “An Environmental Assessment or Environmental 3.3.6.8 With the focused efforts of a large Impact Statement. The Environmental Assessment team of duck capture experts, we feel (EA) claims that there will be no significant impact confident that over the period of 4 from this action. But, the project will undoubtedly months prior to the operations we impact endangered Laysan ducks to at least some will be able to capture almost all extent and those impacts could potentially rise be ducks on Sand Island. The entire very significant. For example, some ducks will island is accessible and with probably be injured or killed during capture and persistent and thorough effort, all uncaptured ducks will be at high risk from ducks will be located and captured. brodifacoum poisoning. The EA currently lacks the A combination of captive care and specificity needed to determine the magnitude of release with clipped flight feathers on impacts. The EA states that a maximum of 119 to Eastern Island where food, water, and 477 ducks could be at risk without mitigation (20-80 cover will be augmented will increase percent of the population; which is up to 40 percent the likelihood of high survival during of the world population). Proposed mitigation is too the operation. vague, stating that the worse-case scenario is to have 197 ducks uncaptured and at risk. But, that it is A combination of strategies are being likely that fewer will escape capture. How much implemented to minimize and offset fewer is not estimated – and should be. Even the impacts to Laysan ducks. Habitat loss of 100 ducks is probably a significant outcome. improvements are being made on In addition, the project does not seem to provide any Eastern Island to provide ducks with actions to offset even the minimum potential adequate food, water, and shelter. impacts to Laysan ducks. There are numerous Protocol to prepare and manage seeps options for minimizing impacts and (water bodies) is in place to reduce mitigating/offsetting impacts. For example, the the potential for a botulism outbreak, project could improve habitat for ducks, implement and personnel to care of the ducks better anti-botulism efforts in the future, or bolster will be on-site and will include a duck populations on other islands. To reach a wildlife veterinarian and an finding of no significant effects, the EA and project aviculturist, capture specialists and will need to do a better job of explaining how you protection team members. Currently, are getting to this finding. Otherwise, an no translocations of ducks off Environmental Impact Statement will likely be Midway Atoll are possible for Laysan required.” Ducks due to funding, regulatory and compliance constraints. 30 “ “ “The explanation of why there is not a diphacinone 2.5.3 Island Conservation wrote a alternative is weak and needs to be expanded. The Feasibility Report (2017) for this Palmyra rat eradication used an EIS that did a better project. They specifically job of explaining the rodenticide selection process. recommended the use of a bait This is even more important at Laysan, given the containing brodifacoum. They note much higher potential for significant non-target that first generation mortality from brodifacoum compared to anticoagulant rodenticides, such as diphacinone to Laysan ducks and other species.” diphacinone, are multi-feed rodenticides, meaning rodents need to consume bait over a sustained period of days to achieve mortality, which may vary from 3-12 days of ingestion of a lethal dose. While diphacinone has been successfully used to eradicate rats from islands, it has been used in efforts to eradicate house mice on islands only twice, with one instance being a reported success and one a failure. Second generation anticoagulant rodenticides, such as brodifacoum, can be toxic after a single exposure or feeding event to sensitive rodents. All other operational considerations being equal, the task of delivering enough bait to all rodents on the island to ensure 100 % mortality is less certain to succeed when using diphacinone than when using brodifacoum. Rodents that survive diphacinone bait application may be resistant to first-generation anticoagulants (e.g., diphacinone), or could be unable to find or consume adequate bait to maintain lethal effects of diphacinone dose, avoid bait after a single feeding, or a combination of these. These rodents, and possibly their offspring, could consequently be more difficult to eradicate from the island in any subsequent attempts. The use of the much more field-tested brodifacoum provides a greater degree of confidence that the eradication program would succeed. The only way to attain cost-effective, and ecologically secure, long-term conservation returns on this investment is through a successful and complete eradication. Therefore, the USFWS considered the use of diphacinone to eradicate mice on Sand Island but these were dismissed from analysis since they failed to meet the purpose and need of the project, which is to eradicate invasive mice in a manner that minimizes harm to island residents and the ecosystem while still maintaining a high probability of successful eradication. 31 “ “ “The project seems to lack key management triggers 3.3.6.8, Previous capture trials and the to protect species from unanticipated impacts. For 3.3.6.5.9 inclusion of Laysan Duck experts on example, if Laysan ducks turn out to be very the duck protection team have difficult to catch, what is the minimum number that increased our confidence that it is must be captured for the project to proceed? If likely that almost all ducks will be ducks from Eastern Island begin making routine captured and either placed in flights to Sand Island, what do you do? If ducks do captivity or have their flight feathers poorly in captivity what do you do? If hundreds of clipped and be transported to Eastern albatross chicks die after the first application, what Island where their food water, and do you do? How many deaths does it take to stop or shelter will be augmented for the alter plans? If albatross chicks do die from the duration of the operation until it is treatments, will you intensify anti-botulism efforts deemed safe for birds back on Sand to reduce botulism impacts on ducks? The lack of Island. Laysan ducks held in contingency planning seems to be a serious problem captivity during 2 translocations and needs to be addressed before approving the adapted well and suffered no losses. project.” If captive ducks do poorly in the aviaries we will clip their flight feathers and release them on Eastern Island. We do not anticipate any mortality in young albatrosses from the bait application. 32 Michelle “The DEA anticipates that the proposed action will 3.3.6.8, The Protected Species Protection Reynolds, PhD, have no significant adverse impact, with Page ES-2 Plan calls for and provides resources Wildlife Biologist implementation of mitigation (page ES-2), however for capture of the entire population the substantial risks to the critically endangered prior to the operation. The removal of Laysan teal appear to be underestimated, and the Mus musculus from Sand Island, proposed mitigation has important inadequacies, Midway Atoll will have immediate uncertainties, and feasibility concerns. Broader and benefits to the Laysan Duck more effective mitigation actions are urgently population by eliminating a direct needed to protect the Laysan duck, as the current competitor of the ducks for food. plan substantially increases the species high risk of extinction and may reverse the recovery efforts of the last decade.” 33 “ “ “Implementing actions within the USFWS Laysan 3.3.6.8 Comment noted. Botulism risk will Duck Recovery Plan (USFWS 2004; 2009) would be minimized by vaccination of the provide more comprehensive mitigation and risk entire population during handling reduction. The proposed mitigation plan prior to the operation. Botulism risk concentrates risks at Midway Atoll (holding in will further be minimized by captivity the entire population on a low lying remote modifying sites that have been atoll). The DEA does not provide any mitigation implicated in previous outbreaks of alternatives that would dilute species risk offsite? botulism, monitoring the health of Translocating a quantity of candidate at-risk birds captive birds, and by managing the off of Midway Atoll to Lisianski Island, Kure Atoll, supplemental water and feeding and/or establishing a genetically managed captive stations that will be installed on breeding population would benefit the species. Eastern Island. A veterinarian, an Broader mitigation actions would help preserve aviculturist, and between six and species genetics, reduce the probability of extinction eight care specialists will be on-site due to random event and disasters, and is an for the duration of the project to opportunity to reduce the risks associated with monitor and manage captive ducks toxicant and avian botulism exposure during and and those ducks wing-clipped and after the proposed toxicant application. released on Eastern Island. Taking Additionally, if the mouse eradication effort is not ducks for translocation off Midway successful and needs to be repeated, a proportion of Atoll is not possible nor the translocated endangered population will have no recommended to us at this time. exposure risk. If the secondary exposure risk are greater than expected, the worst case scenario 80- In the event of a mouse detection 90% mortality would be offset with some off-island more than 42 days after the last bait population insurance, and a plan to genetically and drop a localized eradication response demographically improve population persistence may be undertaken and duck can be implemented after the toxicant application protection measures will be and mouse eradication.” integrated into any actions. 34 “ “ “Apparently, bait is expected to be consumed and 3.3.6.8 We will not release Laysan Ducks stored in invertebrates, particularly cockroaches, until residue of brodifacoum is and a favorite food of Laysan Teal on Midway. This undetectable in prey species (see the secondary exposure risk to the ducks is apparently revised Laysan duck protection plan very high but not clearly mitigated. The plan to in Appendix C of the FEA). We will release endangered ducks and monitor them for release other bird species that feed on toxicosis is incredibly risky. The mitigation actions, the same prey as Laysan ducks effectiveness monitoring, post application actions, (canaries, Serinus canaria and and contingency planning (should the effort fail) common myna, Acridotheres tristis) become less well considered, and less complete. The and monitor their health status prior current mitigation plan for the endangered ducks is to releasing any Laysan Ducks. unrealistic and does little to prevent the primary, secondary, tertiary, disturbance, and extirpation and extinction risks to this vulnerable non-target species.” 35 “ “ “The Draft EA fails to provide the level of analysis 3.3.6.8 Extensive habitat restoration has been or even a recent literature review, and fails to use undertaken at Midway since it was the best available data needed to evaluate the designated a National Wildlife consequences of the array of negative impacts to the Refuge in 1988, including the control Laysan ducks at both the species, genetic, and and in some cases the eradication of population level. The EA document contains non-native plants that competed with inaccuracies and false assumptions about the species native species and diminished behavior and current population dynamics (survival diversity. Ironwood (Casuarina and rates of population growth). The population equisetifolia) have been eradicated on growth under ideal conditions during the Eastern Island and over 20 acres have reintroduction in 2004/2005 are assumed after the been removed from Sand toxicant application, despite that conditions have Island. 1,045 acres of Golden changed. Midway Atoll has undergone substantial crownbeard (Verbesina habitat changes at after intensive de-vegetation with encelioides) have been removed weed removal and annual seasonal outbreaks of throughout the atoll. Both these plant Avian Botulism C (a foodborne poisoning produced species form monocultures and by Clostridium botulinum). The current mitigation provide little food or cover for plan does not provide any actions to reduce the Laysan Ducks. In concert with the extinction risks to the Laysan duck, and misses the management of invasive non-native opportunity to improve the species probability of plants MANWR has an active native persistence before, during or after the proposed plant propagation and out-planting toxicant application.” program. To date they have out- planted and seeded 14 native species over 75 acres at the atoll creating habitat with cover and much greater food density for Laysan Ducks, thus increasing the carrying capacity.

All ducks captured will be vaccinated for avian botulism and sites that have been implicated in previous outbreaks of botulism are being modified to reduce risk. The removal of Mus musculus, a direct competitor with Laysan Ducks and a possible nest predator, will increase the carrying capacity of the Atoll for Laysan Ducks. 36 “ “ “The timing of the toxicant application (July) 3.3.6.8 Comment noted. creates additive risks to Laysan ducks. This is during the peak of LADU breeding and molt, peak Results from recent trials to compare of seasonal avian botulism epizootics, and reduced bait uptake rates in July and food availability, and reduced capture probability. September indicate faster bait The mitigation plan ignores both the risk and disappearance in September 2018 feasibility associated with capturing and holding a than in July of 2018, arguing for the large number of wild birds in captivity, and the earlier application period from an scale of the operation… an entire wild population efficacy standpoint because even at (approx. 500 endangered birds) is a massive the highest labelled application rate undertaking for a remote location. Concentrating the we couldn’t maintain bait on the ducks in a flightless state on a remote atoll during ground for more than 2 days in the period of primary and secondary poisoning sufficient amounts to ensure bait exposure risks is also a massive undertaking. The availability to all mice. In 2018 ducks mitigation actions at the beginning of the document continued to breed during the month states 100% of ducks will be caught, but later in the of September with young broods still document describes that worst case scenarios will appearing so the advantage of waiting have only a short term effect.” until ducks were no longer breeding to improve detection and capture probabilities was not evident this year. 37 “ “ “Lastly, there is less scrutiny of the actions that 3.3.6.8 We will have full-time veterinary impact endangered species implemented directly by staffing and recognized experts in the the USFWS compared with other agencies, captive care of waterfowl that will be researchers and non FWS managers. The section 7 responsible for the maintenance of of the Endangered Species Act (vs and Endangered the captive population at all times. Species Permit) is self-approved within the agency, not subject to outside or public review and We will measure brodifacoum mitigation implementation is not usually enforced or residues and release other species reviewed. There is no Institutional Animal Care and (specifically, canaries and mynas) Use Committee (IACUC) requirements for the USFWS. There is essentially no oversight for this with diet overlap with Laysan Ducks remote location, where the priority to eradicate mice as soon as those residues are no may be higher than the priority to prevent take. The longer detectible in the food prey. EA does not lay out the details of a monitoring plan The health of canaries and mynas will to quantify the post toxicant exposure mortality to be monitored carefully prior to re- Laysan ducks. Given that the detection probability introducing Laysan Ducks. of duck carcasses may be very low, a rigorous, intensive and systematic post toxicant application Known duck resting and foraging non-target monitoring plan (perhaps including sites will be searched for signs of any detection dogs) may improve our ability to quantify ducks until no visible bait pellets the direct mortality impacts to the endangered remain after the final application of population. The mitigation concentrates these the bait. Sites will be searched daily. endangered birds and does not acknowledge that Any carcasses found will be stochastic events could result in mass mortality. recorded, and carcasses suitable for Actions to recover the Laysan teal are not included necropsy will be collected, labeled, in the DEA.” frozen and submitted to U.S. Department of Agriculture for analysis. See the revised Laysan duck protection plan in Appendix C of the FEA) for more details. 38 “ “ “In the event that there is failure to eradicate mice 2.3.8 A failure of the eradication will not after this plan is implemented, are future eradication reduce the purpose and need of the attempts planned? The duration of repeat toxicant first attempt A careful review of the exposures certainly increases the species level threat methods and possible causes of a to the Laysan ducks. How will this be mitigated?” failure would be undertaken prior to any decision on whether a repeated attempt would occur and how methods might differ from the previous one. For example, in the event of a mouse detection more than 42 days after the last bait drop a localized eradication response may be undertaken and duck protection measures will be integrated into any actions. 39 “ “ “Post treatment environmental effects – 2.3.7, Environmental effects monitoring is documenting and quantifying the mortality of 3.3.6.8 already underway with pre- Laysan ducks and the detection probability of this eradication surveys of a wide variety mortality is not clearly outlined. These metrics are of ecosystem components. These important to evaluating the impacts to the surveys will continue during the post- endangered ducks. A transparent and systematic eradication period. monitoring plan with oversight should be included so that endangered species take (and other non- A comprehensive monitoring plan target mortality) is carefully documented. Detection that specifically addressing LADU dogs are quite effective when trained for species and the documenting of post- specific carcass detection.” treatment environmental effects for the project is outlined in the Protected Species Protection Plan and is attached as an appendix to the EA. 40 “ “ “Costs are estimated at 3.5 million. Does this cost 2.6 The project cost is estimated at 4.5-5 estimate include the Laysan duck mitigation?” million dollars.

41 “ “ “The importance of the Midway Population, 3.3.6.8 Our protected species plan calls for proportion of the species likely to be effected, extreme efforts to protect the entire sensitivity and exposure of Laysan ducks to the Laysan Duck population at Midway toxicant, and the duration of the ecological effects is Atoll. We have no way of knowing not carefully analyzed.” what proportion of that population will evade our efforts but we do not intend to allow any ducks to be exposed. We are confident that any ducks left on Sand Island will ingest a lethal dose, so we will endeavor to capture them all. 42 “ “ “Table 3.10 Mitigation measures to “capture 100% 3.3.6.8 We do not know how long of the (Laysan duck) population” or with current Brodifacoum will persist at estimates approximately 500 wild birds (pgs. 3-66) dangerous levels in the Midway to hold in captivity until the risk period passes (not ecosystem. We are exploring the defined) or feather trim and release all the birds not possibility of having an analytical lab in aviaries on 133 ha island (marginal habitat) and a set up on Sand Island to ensure quick mean elevation of 2.1 m (SD 0.6) concentrates feedback about samples. See the stochastic risks to a globally significant population. revised Laysan duck protection plan What is duration of captivity? For planning in Appendix C of the FEA for purposes (costs and resources) and risk assessment, estimates of the duration of captivity. the duration of captivity information and alternative scenarios are crucial. The risk period and thus holding period for the Laysan ducks is vague, but I infer that for primary poisoning is expected to be about 40-100 days, and for secondary poisoning about 6 months? Please clarify.” 43 “ “ “Bait longevity (pgs. 3-29) may be high in some 3.3.6.9, Laysan Ducks will be prevented from areas due to other factors (USFWS unpublished). 3.3.6.8 having access to terrestrial What are the factors we cannot look up in that invertebrates on Sand Island until reference? Bait longevity influences toxin exposure residues are no longer detectable and of endangered ducks. Pellet disappearance was other species with similar diets driven by cockroach and invertebrate consumption. (canaries and mynas) are persisting Terrestrial invertebrates and cockroaches in without signs of Brodifacoum particular are important components of the adult poisoning. Laysan ducks and larger stage ducklings’ diet thus secondary exposure may be considerable. Prior to Brodifacoum residues in living risking Anas laysanensis to secondary poisoning, tissues (e.g., food web compartments) couldn’t a small scale experiment be conducted to will be assessed by collection and test these residues a-priori…using baits and euthanasia of appropriate cockroaches at Midway Atoll to reduce secondary invertebrates, lizards, fishes and poisoning uncertainty? It is easier to keep birds, with liver tissues (site of cockroaches flightless and in captivity than wild greatest accumulation) harvested and endangered ducks.” submitted for chemistry. 44 “ “ “Since 2011 (after the 2011 Tsunami) and after 3.3.6.8 See comment #23 for a detailed intensified weed removal efforts targeting the response. invasive Verbesina then other non-native but non- invasive grasses in 2012-2016, vegetative cover decreased by more than 50% (p 3-31). The weeds were utilized as foraging habitat, cover, and nesting habitat for Laysan ducks and the weeds are thought to have contributed to their rapid population growth and high survival of newly translocated birds (harboring abundant arthropod communities). Unintended consequences of rapid weed removal and time lags for native and biodiverse vegetation restoration, likely included reduced food and habitat for both the mice and the endangered ducks at Midway Atoll.” 45 “ “ “Page 3-64 Biologist were able to trap 5-9% of the 3.3.6.7, Each year we cover every square Bristle Thighed Curlews at Palmyra. This section of 3.3.6.8 meter of Sand Island to count all the DEA acknowledges the additional risks albatross nests and we have associated with capturing the birds including essentially eliminated Verbesina physical injury and physiological stress and risk of encelioides by also treating every death. These risks of physical injury are not square meter of ground. We will be acknowledged in the section describing an Anas able to catch almost all ducks on the laysanensis population scale capture and captivity island over the period of 4 months attempt? Wild Laysan teal are quite aggressive and prior to the operation. Aviaries will fight especially when held in captivity. Many will be designed to separate ducks into not feed and will pace incessantly and do not adapt small groups and they will be to captive care. Captivity stress and unsanitary monitored constantly with closed captive conditions can result in respiratory and other circuit TV to ensure that behavioral diseases. Pair-bond dissolution and brood conflicts are addressed immediately fragmentation, and territory loss are also likely. by on-site personnel that includes a wildlife veterinarian and an aviculturist, capture specialists and protection team members. Pair bonds may be disrupted but we will have the option to release to Eastern Island birds that do poorly in the aviaries. 46 “ “ “The literature review for this section is based on 3.3.6.8 Errors will be corrected and older research mostly 20-30+ years old primarily references updated as much as from early studies on Laysan Island and does not possible to include recent work. include the recent research and includes many citation errors. Contemporary threats to the Laysan ducks are largely absent. “Mice and invasive weeds” are listed as current threats to the species, yet highest documented reproduction and survival for the species co-occurred at Midway Atoll with mice and weeds. The translocation to Kure is described as “established” despite that the population is approx. 30 birds and adult mortality (due to avian botulism) has exceeded successful reproduction in 2016 and 2017.” 47 “ “ “The average weight of post-fledgling LADU at 3.3.6.8.3 We are assuming that any ducks left first capture at Midway Atoll was 442 gm range on Sand Island will have access to a 270-600 (SD 35.44 gm) n=459 ducks captured lethal dose of brodifacoum. 2007-2014. Females are heaviest while gravid, but Removing them from Sand Island lose weight rapidly during incubation (about 30 during the operation is essential. days) and brood rearing. Therefore the lethal doses may actually be slightly less based on average body mass, and the number of female breeders in the population will be much lower than 50%.” 48 “ “ You state that “20-80%” of the population could 3.3.6.8.3 Comment noted. See the revised still be exposed to primary or secondary poisoning. Laysan duck protection plan in “Birds that breed on Eastern Island also fly to Sand Appendix C of the FEA for mortality Island and vice versa so 90-100% exposure is also and injury estimates to Laysan ducks possible. Given the mitigation as currently outlined, from each phase of the operation. the maximum level of mortality could still occur, along with other mortality factors (brood fragmentation, emaciation, avian botulism, sudden flooding, extreme weather and random mortality factors) as well as the risks associated with the current mitigation plan to hold 50% of a global population in captivity at a single site.” 49 “ “ In the last paragraph of 3-67 you state “The worst 3.3.6.8.3 Agreed; detection probability during case scenario would be a rate of 33% maximum surveys is not capture probability accessible population for capture?” Detection over 4 months. However, with the probability during a survey is not the same as focused efforts of a large team of capture probability. 197 will not be captured? Why duck capture experts, we feel is this paper cited? (Reynolds et al 2017). Also confident that over the period of 4 capture does not = survival either. Back to table months prior to the operations we 3.10 the no significant impact depends on 100% will be able to capture almost all capture (and 100% survival) of Laysan ducks held ducks on Sand Island. The entire long term in experimental conditions. island is accessible and with persistent and thorough effort, all ducks will be located and captured. A combination of captive care and release with clipped flight feathers on Eastern Island where food, water, and cover will be augmented will increase the likelihood of high survival during the operation. 50 “ “ “On Page 3-68 there are inconsistencies re: the 3.3.6.8.3 We agree that captive care poses number of ducks to be held. This varies from 100% significant risks but during the capture, to 66% to a “significant number” during the translocation project there were no bait drop period “protecting” them helicopter ducks lost to captivity issues. We will collisions and trampling, but risk associated with hold 200 ducks in captivity (100 capture holding wild birds in captivity are not male/100 female) and the rest of the acknowledged. Stress responses (pacing, anorexia, population (~400 animals at this fighting, hiding, and substantial weight loss) were time) will be confined to Eastern observed in wild captive Laysan ducks held for Island by clipping flight feathers. short periods during the translocation. Starvation, There will be no ducks intentionally trauma, and aspergillosis and other fungal and left on Sand Island to be exposed to respiratory infections are also possible. On the next the toxicant. Personnel for Laysan page, 100 birds are planned capture 50 males and 50 duck mitigation will include a females are proposed. This might be more feasible, wildlife veterinarian and an but surely 400 endangered ducks remaining exposed aviculturist, three capture specialists to the toxicant is not considered robust mitigation. and three protection team members. Please clarify. Protection team members will assist with captive care tasks including food preparation, feeding aviary birds, cleaning aviaries, observation of marked birds on Eastern, supplemental feeding on Eastern, 51 “ “ “How many additional aviculture professionals 3.3.6.8.3 Current plans have duck protection experienced with the captive care of wild birds will staff levels at 10 individuals be hired and how long will these aviculturist be including 1 veterinarian, 1 stationed at Midway? Keeping aviaries clean and professional aviculturist, 4 other supplying adequate water for 500 captive ducks in aviculture staff, and 4 capture separate pens to reduce fighting and trauma is a experts. Birds will be held on Sand large undertaking. Who is this interdisciplinary team Island until the bait application then of Laysan duck experts? Supplemental feeding and moved to aviaries on Eastern to avoid aviaries attract mice and ants. How will potentially the risk of invertebrates that have toxic invertebrates (secondary poisoning risks) be consumed bait getting into the kept out the aviaries on Sand Island?” aviaries. The duck care team will rotate to a residential camp on Eastern. Aviary cleanliness on Eastern will be maintained with salt water pumps and aviary floors will be easily cleaned rubber matting that is easy on the birds’ feet. There are no mice on Eastern to be attracted to the aviaries and supplemental feeding stations. Ducks will only be held in aviaries on Sand Island until just prior to the first bait application and thus ducks will be on Eastern Island before toxic invertebrate prey species become available. 52 “ “ “Mitigation – “The only effective mitigation and 3.3.6.8.3 No translocations or removal to minimization strategy is to prevent the exposure of captive breeding programs are the ducks to the rodenticide either through live possible for Laysan Ducks from capture and movement to Eastern Midway Atoll at this time due to Island”….numerous other approaches are possible funding, regulatory and compliance to prevent the exposure of a proportion of Laysan constraints. duck population to the rodenticide. Please read the recovery plan for some additional suggestion that may reduce the extinction risk to Laysan duck. Also see the Structured Decision Making report (National Conservation Training Center 2011). Translocation of birds off of Midway Atoll should not be excluded as mitigation.”

53 “ “ “Page 3-69 critically endangered Laysan teal that 3.3.6.8.3 Laysan Ducks will not be used as are captured and held in captivity will be released “sentinels” We will test for residues during the period of risk for secondary poisoning? of brodifacoum in likely prey species Endangered species will be released and used as and release other captive birds that sentinels (indicators of toxin?). The endangered have a diet that overlaps with Laysan birds should not be used as indicators of Ducks prior to any release of Laysan environmental toxin. This is a huge risk of take. Ducks. Sentinel canaries and mynas Rather, intensive and systematic survey efforts to will be released on Sand Island and find dead birds and estimate the time of death from monitored for any ill effects. Radio- carcasses in the environment. This would not risk transmitters will be attached to a additional birds (i.e. search for those birds that will sample of Indian Mynah birds be missed during the capture effort or those that fly (sentinels) to monitor their from Eastern Island to forage). After residual toxin movements, behavior, breeding and in prey items is low, and only old or no new survival when they are released. The carcasses are detected during intensive searches, birds will also be observed at a should captive held ducks be released and distance using spotting scopes to monitored.” prevent disturbance. 54 “ “ “Page 3-70 these bulleted actions are increasingly 3.3.6.8.3 We will engage in intensive scrutiny unfeasible and untested. The mitigation ignores that of all the free-roaming Eastern Island flightless Laysan teal are highly nocturnal and birds and be able to monitor their secretive. It will be very difficult to notice flight molt condition because of the feather growth of these hiding birds. Even radio relatively open conditions and tagged birds are very difficult to observe and artificial feeding and watering progressively difficult to recapture. Toxicosis may aggregation points. All Laysan Ducks also be difficult to detect. Non-target toxicant handled will receive 2 doses of related mortality of Laysan teal, will likely botulism vaccine. SWCA (2017) contribute to terrestrial outbreaks of avian botulism. tested methods on Sand Island to Spores grow in duck carcass guts (post mortem), fly increase capture efficiency of Laysan larvae on the carcasses bio-concentrate the biotoxin ducks. Carcass searches will be exposing more ducks to avian botulism via the fly conducted daily and known duck larvae directly or the food web. This also occur at resting and foraging sites will be wetlands.” targeted. 55 “ “ “Page 3-70 shouldn’t non endangered birds be used 3.3.6.8.3 Agreed; prey species will be tested as indicators of toxin in the food web (sentinel and no ducks will be released until mynas or canaries)?? Can’t the duration of toxic residues are undetectable. Rates at residues of beetles and cockroaches be tested prior? which brodifacoum residue washes Perhaps these Anas laysanensis prey items could be out of invertebrates and becomes fed toxic baits on Midway Atoll in outdoor captive inaccessible depend on ecological experiments to reduce lethal residue uncertainty in conditions such as rainfall, phenology this ecosystem?” of the species in question and community ecology of the soil organisms which cannot be simulated in a laboratory experiment. A toxicological assessment estimated a minimum holding time for ducks of 4-6 months and a maximum holding time of 13-15 months. 56 “ “ “Page 3-70 “Based on capture success rates in 2017 3.3.6.8.3 Comment noted. SWCA (2017) it is likely the majority of birds will be captured”? tested methods on Sand Island to Please cite reports that present results transparently. increase capture efficiency of Laysan To my knowledge, 12 naïve Laysan teal were ducks. Trials were conducted to captured simultaneously in a 2017 SWCA report. develop alternative trapping Risks associated with capturing multiple ducks techniques for capturing multiple simultaneously and the skill required to do so safely ducks at one time. This report is cited was included in that report. Since it is being in the protection plan for LADU described in the EA, this report should be cited. included in the EA as an appendix. Detector dogs might improve detection probability The non-target species protection of nests and hiding birds.” team will explore the use of detector dogs. 57 “ “ “Page 3-71 states “it is likely there will be short 3.3.6.8.3 Agreed; the entire Midway Atoll term adverse effects to the population on Sand population will be treated as one and Island” There is no separate Sand Island population protection actions will be focused on and separate Eastern Island population. This all birds there. toxicant application is likely to negatively impact the entire population that flies between these island to forage, breed, and socialize. The uncertainty associated with the duration of exposure throughout the food web may exceed 6 months after the bait degrades. There is no clear mitigation plan for the longer duration of secondary poisoning risks.” 58 “ “ “Is the EA suggesting that the population today, will 3.3.6.8.3 We are not familiar with any easily recover from a mass mortality on page 3-71? evidence that foraging habitat has A maximum Lincoln-Petersen population been decreased since 2008. In fact, abundance estimate (that included ducklings) is not native vegetation restoration appears an indicator that species is unthreatened. Typically, to have increased the quality and ducks can produce many ducklings in a good quantity of duck forage. Carrying breeding year and many of these juveniles do not capacity for ducks will likely be survive to breed. Density dependence, chronic further enhanced by the removal of annual botulism and dramatic habitat changes were mice, which are direct competitors evident after this maximum estimate, therefore is for Laysan Ducks. not an indicator that the current population would easily rebound from a massive toxicant induced We do not anticipate a massive mortality event. The translocation candidates were toxicant induced mortality event. selected over two years from the healthiest post fledgling juvenile birds from Laysan. These Collection stations to process candidates were in ideal body condition from captured ducks would be temporarily different families (to maximize genetic founders) at setup for the initial capture efforts. the age with the highest lifetime reproductive Personnel will include a wildlife potential. The birds were treated for parasites and veterinarian and an aviculturist, three released when Midway Atoll was densely covered capture specialists and three in Verbesina weeds with more numerous fresh water protection team members. seeps. The post release population did not suffer avian botulism until late 2008. Die off and seasonal botulism mortality has occurred every year since. Some years the mortality is catastrophic. Additionally, foraging habitat today is much reduced compared to 2004-2008. The population has not increased at the rate it did after the translocation since 2008. Annual survival rate of Laysan teal at Midway has declined every year since the founding birds were released, and is approximately 20-30% lower than annual survival on Laysan Island.” 59 “ “ “Page 3-70 Supplemental feeding is not a mitigation 3.3.6.8.3 Botulism risks will be minimized by for overcrowding, insufficient habitat, or territorial vaccinating the entire population, aggression on Eastern Island. Supplemental feeding filling existing seeps on Eastern, and of free-range (feather trimmed) birds on Eastern providing all water using 11 guzzlers may have unintended consequences. Examples of that can be easily cleaned and unintended consequences (of providing large maintained. quantities of duck chow in the environment) – it might increase invasive ant populations or increase We agree that supplemental feeding avian botulism risks (the toxin producing bacteria will increase the risk that ducks cannot produce its own protein). Supplemental would also take bait pellets if they feeding might influence molt (decrease the period of were on Sand but experiments with flightlessness). Lastly, training Laysan ducks on non-toxic bait pellets have already novel artificial food on Eastern may actually indicated that Laysan Ducks relish influence their foraging behavior and increase them and will take them readily. A consumption of toxic bait on Sand Island (by duck protection team will remain on undetected uncaptured ducks moving between the Eastern Island throughout the islands). These are scenarios that illustrate duration of the project. The team will uncertainty associated with supplemental feeding manage and monitor the wild birds.” supplemental feeding stations and will be able to implement adaptive management strategies to deal with unintended consequences that may arise. 60 “ “ “Page 3-71, the mitigation description confuses that 3.3.6.8.3 Comment noted. We do not anticipate birds on Eastern will not be effected by the toxicants that this action will reduce genetic applied to Sand Island or that feather clipped birds variation in the population. It is will not be exposed to the toxicant or secondary feasible to capture the majorityof poisoning (flightlessness may only last a month). Laysan ducks at Midway Atoll and Kure Atoll’s population is not “established” and therefore, mitigation efforts will be may not be a viable long term in the face of avian attempted to protect all individuals in botulism outbreaks since 2015. Laysan Island no this population, thereby potentially longer has a presence of biologists, and catastrophic safeguarding most of genetic population declines would likely go undetected. The variation remaining in this group of population status of Layasn is often not known. ducks descended from the original Species viability is largely dependent on the translocated population. See the Midway Atoll population, and the toxicant drop revised Laysan duck protection plan may not work the first time, so the duration of in Appendix C of the FEA for more exposure may be quite long. The previous threats to details. the species are still present, thus the toxicant drop is an additive threat. The additive mortality may reduce the probability of persistence of the species via genetic and demographic changes. Furthermore, the Midway population may not be suitable as a source as future translocation founders or founders of captive breeding populations (see FWS recovery plan) if population declines further reduce the genetic variation of this important population.” 61 “ “ “I recommend planning and initiation of capture and 3.3.6.8.3 Translocating ducks off of Midway off-island translocations of the Laysan ducks in Atoll is not possible at this time due 2018 to restore wild population extirpated from to funding, regulatory and Lisianski Island. Avian botulism outbreaks at Kure compliance constraints. have resulted in loss of founding birds (reduced founder genetics) and slow population growth at Kure Atoll. Additional founders are needed at Kure Atoll. Removing as many ducks from Midway Atoll as possible will prevent toxicant exposure, reduce the magnitude of the mitigation at Midway, and potentially benefit the species. Keeping a smaller captive population and feather trimming the birds trapped on Eastern (but not adding many to that habitat) might be more manageable and successful if it were scaled down.” 62 “ “ “There is an opportunity to refresh captive 3.3.6.8.3 Taking ducks for introduction into populations and establish genetically managed captive breeding programs is not captive breeding populations by translocating some possible nor recommended to us at birds into captivity rather than risk mortality from this time due to funding, regulatory primary and secondary poisoning.” and compliance constraints.. 63 “ “ “There is an opportunity to conduct research test 3.3.6.8.3 We will be vaccinating all ducks avian botulism vaccine boosters and the timing of handled during this project for immune response to protect Laysan ducks from preventative and protection reasons. avian botulism mortality via vaccination.” Adding a research project into the eradication project is not feasible nor recommended to us at this time. 64 “ “ “There is an opportunity to include genetic 3.3.6.8.3 We do not anticipate population-scale supplementation as mitigation to offset the genetic losses of Laysan Ducks during this consequences of population scale losses of Laysan operation. It is feasible to capture the teal due to toxicant exposure (additional founders majority of Laysan ducks at Midway from Laysan Island after mice are eradicated and Atoll and therefore, mitigation efforts secondary exposures risks have passed. A small will be attempted to protect the number of supplemental founders from Laysan majority of individuals in this Island if population status there is suitable, could be population, thereby potentially added after Midway Atoll is deemed rodent free.” safeguarding all genetic variation remaining in this group of ducks descended from the original translocated population. 65 Maile Walters Supports the project. N/A Comment noted.

66 The Nature “The Nature Conservancy (TNC) strongly supports N/A Comment noted. Conservancy, the Proposed Action in the Midway Seabird Honolulu, Hawaii Protection Project Draft Environmental Assessment (“DEA”) to eradicate introduced house mice (Mus musculus) from Midway Atoll National Wildlife Refuge.” 67 Mark Williamson “It seems to me there is a need to monitor the Appendix While the collection of real-time operation as it is ongoing, and that means some type B surveillance equipment would be of real-time collection of information. The useful for monitoring purposes, deployment of solar- or battery-powered because multiple species need to be surveillance devices, allowing remote monitoring of monitored and monitoring methods the impact on the birds and other non-target species, differ per species and among habitats, might benefit this monitoring process. In this way, real-time surveillance methods are on-site personnel could "monitor and adjust" and not a cost/time-effective strategy to react promptly in case of a serious problem with an employ for this project. Monitoring adverse impact on some of the birds. I saw in the methods outlined in Appendix B plan, in section 3.3.6.2, that vitamin K will be include systematic and routine maintained as an antidote in case non-target animals searches by staff to look for affected ingest the brodifacoum and become ill. Such a species and carcasses on terrestrial surveillance system might allow personnel to notice and near-shore marine environments, a problem and react to it more quickly and more deploying and monitoring camera decisively, for example, by administering that traps and acoustic recorders, antidote more promptly to animals that fall sick, sampling of prey species for thereby helping protect the animals there, especially rodenticide residues, specific and members of the endangered species. Alternatives to stringent monitoring of federally- such surveillance equipment would probably entail listed species, and by opportunistic increased patrolling, which would likely be less sampling and observations made effective and might increase the possibility of during other aspects of the accidentally damaging burrows, etc. operational activities. Last, telemetry So, my suggestion is the deployment of a real-time of radio-tagged mice during the surveillance system to allow for better monitoring of project will track the fates of a the process, if that has not already been sample of individuals. Post- considered/done.” eradication mouse detection methods including chew blocks, trail cameras, and other techniques. 68 Martha Brown, Supports the project. N/A Comment noted. University of California, Santa Cruz 69 Suzanne D. Case, “DOFAW notes that the EA proposes mitigation for 3.3.6.8 We will protect as many Laysan Chairperson, Dept. the Laysan duck, but we are concerned that the Ducks as possible and believe we can of Land and mitigation planned does not go far enough to protect do so for most of the birds at Midway Natural them from the preferred alternative's impacts. The Atoll. We will hold 200 ducks in Resources, current world-wide estimate of Laysan ducks is captivity (100 male/100 female) and Honolulu, Hawaii between 314 and 435 individuals, but it appears that transfer all other ducks caught to the mitigation would include capturing 50 adult Eastern Island after vaccinating and male and 50 adult females to hold and transfer to wing-clipping. Enhancements are Eastern Island before the drop of bait. Given that being made to host additional ducks Midway hosts approximately 50% of the population on Eastern Island, and a duck of the species, DOFAW would encourage the caretaker team that includes a Service to extend mitigation efforts to as many veterinarian and aviculture specialists individuals of the populations as possible.” with Laysan duck expertise will be on-site to monitor the ducks both in the Sand and Eastern Island aviaries and those ducks released on Eastern after wing-clipping. 70 “ “ “DOFAW would also recommend consideration of 2.3 See response to Comment #30. Island an additional alternative that includes the use of a Conservation wrote a Feasibility diphacinone-based rodenticide as part of the Report (2017) for this project. They eradication matrix, or include diphacinone as part of specifically recommended the use of the preferred alternative, in case it is needed in the a bait containing brodifacoum. They mop-up phase of the operation.” note that first generation anticoagulant rodenticides, such as diphacinone, are multi-feed rodenticides, meaning rodents need to consume bait over a sustained period of days to achieve mortality, which may vary from 3-12 days of ingestion of a lethal dose. While diphacinone has been successfully used to eradicate rats from islands, it has been used in efforts to eradicate house mice on islands only twice, with one instance being a reported success and one a failure. 71 “ “ “DAR notes that the analysis has not included an in- 3.3.6 The USFWS did carcass surveys depth consideration of bird carcasses remaining on during the 2017 bait trials to evaluate the island and their potential as a food source the rate of their disappearance during competing with the rodenticide bait applications. and after the fledging season. They Considering this potential may increase the success also used camera traps to document of this eradication effort.” mouse scavenging on those albatross carcasses. During the July 2018 trials, bait palatability and timing directly related to nonhuman-based alternate food sources will be looked into further. So, the USFWS is measuring and considering in their plans bird carcasses and their availability to mice as a potential food source. 72 “ “ “Our Department recognizes that while rodent N/A/ Comment noted. eradication programs pose a potential risk to non•target species, we support the proposed action, as the cost/benefit to wildlife and ecosystems is likely to be strongly positive in the ecosystem's favor in the long run.” 73 Dr. Fern Duvall II, Supports the project. N/A Comment noted. U.S. Co-Chair, Pacific Birds, Habitat Joint Venture 74 William Aila, Jr. The RAC supports the purpose and preferred Appendix The information found in Appendix Chair, Northwest alternative to remove mice from Midway as outlined A A, page A-1, is only a brief Hawaiian Islands in the Draft EA. The bio-security plan could be summarization of the entire Coral Reef more detailed in areas. For example. the plans for biosecurity plan. In the larger 27- Ecosystem ship visits are comprehensive, but plans for aircraft page biosecurity plan developed by Reserve. visits, especially emergency landings, are lacking. the USFWS, the refuge addresses mammals, reptiles, amphibians, plants/seeds, and bird species along with preventative and rapid response methods. The entire plan, which is too lengthy to add to the DEA, sufficiently addresses measures for preventing the introduction and spread of exotic species accidentally transported to MANWR. Biosecurity measures are, and have been in place, for the refuge. Before, during and after the project, a majority of operational gear, personnel, equipment, and supplies will be transported by boat, hence the emphasis here on biosecurity measures is for ships and shipping. 75 “ “ “We appreciate that USFWS is partnering with 1.5 Agencies that USFWS are partnering organizations that have experience with mice with on this project already have eradication efforts. The RAC suggests that the departments and staff that perform USFWS also work with partners who have expertise public outreach and maintain in community relations to gain and maintain support networks in the community and local for the project.” media. In addition, the USFWS has their own “Office of External Affairs” for the Pacific region which provides support to the regional office and field stations to communicate and facilitate information about the Service's programs to the public, media, Congress, Tribes, partners, and other stakeholders. Groups that USFWS does not partner with, but that do support the project, exist. These groups like Friends of Midway Atoll, provide public outreach and information to the media of their own accord. Further, there are local newspapers and radio stations that report conservation-related news for Hawai‘i. Agencies that USFWS are partnering with have key roles in the myriad processes necessary to the project’s success, specifically the planning, implementing, and monitoring. 76 “ “ “Midway is culturally as well as biologically 1.5 Preserving Hawaiian Culture is important. The RAC encourages USFWS to consult already integral to the management of with the Office of Hawaiian Affairs on cultural the Papahānaumokuākea Marine protocols to assist with the success of the project. National Monument. The refuge was expressly created to protect an exceptional array of natural and cultural resources. The refuge’s stated mission is to carry out seamless integrated management of the monument to ensure ecological integrity and achieve strong, long- term protection and perpetuation of the Northwest Hawaiian Island ecosystems, and the native Hawaiian culture and heritage resources for current and future generations. 77 “ “ “USFWS should consider whether climate change N/A Mice were introduced by the military contributed to the problem.” during World War II and was thus the problem was not originally caused by climate change. The current mice attacks on albatross was likely caused by the eradication of the rat population which had previously kept the mice population at low levels. Mice are known to attack seabirds on other islands. For example, mice have been known to attack and eat chicks of other species of albatross on Gough Island in the south Atlantic and on Marion Island in the sub- Antarctic. 78 “ “ “What additional steps will USFWS take should 2.3.8, 2.1 The follow-up actions should one or initial efforts not result in complete eradication? more mice be detected post-project How successful has mouse eradication been on will vary based on the time and similar sized islands? How will USFWS determine circumstances of the detection. that there are no mice remaining on Midway?” Additional baiting may occur in areas where mice are found if the circumstances dictate there is a reasonable chance of eliminating any remaining mice without additional adverse effects. Should the project fail entirely, and a mouse population persists, the partners will undergo reviews to determine the main causes of failure, and to identify possible solutions to those failures to inform a new strategy for future eradication attempts. Mouse population control that was started in 2016 using cholecalciferol would continue proactively and reactively to reduce mouse attacks on albatross while a new eradication strategy is being developed (p. 2-16). These control methods, in combination with post- operation monitoring and surveying efforts, provide a means for quickly detecting if mice still exist on island. The success of mouse eradication on similar sized islands was provided on page 2-2 of the DEA. 79 “ “ “How will the spread of fine bait particles by 2.3.6, No fine particulate rodenticide is helicopter rotor wash be monitored?” Appendix being dispersed. The aerially- B delivered bait is a specially- formulated, hard pelleted rodenticide. However, bait densities after each drop will be monitored throughout the island using sample plots. In addition, as outlined in the Summary of Proposed Monitoring in Appendix B, to inform eradication planning efforts where possible, the USFWS will evaluate bait uptake rates and persistence in representative habitats on Sand Island. 80 “ “ “How will the effects of eradication efforts on 2.3.6, Sea turtles and Hawaiian monk seals nearshore areas, turtle nests and monk seal mothers Appendix are federally-listed species and and pups be monitored? What steps will USFWS B, already closely monitored on take should any of these animals or areas experience 3.3.6.19, MANWR. The refuge carries out unforeseen impacts?” 3.3.6.20.3 monitoring activities along the coastline and nearshore waters of Sand Island for sightings of hawksbill sea turtles, potential Hawaiian green sea turtle nests, and pupping Hawaiian monk seals. Monitoring the impacts post-eradication will be conducted in conjunction with the refuge’s monitoring efforts for sea turtles and monk seals. To measure the recovery of listed species and the unintended consequences of eradication efforts, several key components will be characterized prior to eradication, and evaluated at 1-, 3-, 5- and 10-year post-eradication intervals (p. B-1). Mortality of all non-target organisms associated with the baiting operations will be assessed by active searching for non- target carcasses on terrestrial and near-shore marine environments and by opportunistic sampling during other aspects of operational activities. The USFWS will consult with the Protected Species Division of NOAA, under Section 7 of the Endangered Species Act, to identify the best course of action to ensure protection of Hawaiian monk seals and sea turtles. This process will address monitoring protocols; bait application around seals and turtles; and a response plan in the unlikely event of finding an individual of either species sick or dead during or following the operation. 81 Hugh D. Hinckley Supports the project. N/A Comment noted.

82 Nicki Lorayn Supports the project. N/A Comment noted. 83 Keota Lindsey, “OHA supports the proposed action and we look N/A Comment noted. Program Manager, forward to seeing the project completed and the State of Hawaii, resources of MANWR protected by removing the Office of invasive threat.” Hawaiian Affiars, Papahanaumokuak ea 84 “ “ “One issue OHA requests clarification on is whether 2.3.8 This will be addressed in the non- there is a process and authority structure in target mitigation plan which is place to stop all activities related to the project (i.e. currently being written. The operation the second and third aerial broadcasts of the would be organized using the incident rodenticide and deployment of bait stations), if that command system (ICS) with the becomes necessary, should unanticipated effects to Refuge Manager or their designee (a non-target species occur.” Fish and Wildlife Service employee) as Incident Commander. Any deviation from the expected and allowed effects on non-targets would be grounds for a cessation of operations by the Incident Commander. 85 Mark Rauzon, “We concur with the use of Brodifacoum, to be N/A Comment noted. Vice-Chair for implemented using best practices in concert with the Conservation, experienced consortium of talented biologists in the Pacific Seabird Midway Restoration Partnership Group.” Group 86 Isaac Harp (Paka) “I support the preferred alternative of three bait 2.3, Surveys of the beaches, coastal applications spaced roughly 7-10 days apart, but I Appendix waters, and terrestrial areas will be also have a request. I request that a visual survey of B conducted prior to the second and the beaches, coastal waters, and the land be third applications. These surveys are conducted prior to the second and third applications to monitor various aspects of the to insure no other wildlife is being negatively operation and include assessing bait affected. If other wildlife is found to be negatively persistence, removing mouse affected, I request cancelling additional application carcasses, retrieving sick or injured and seeking another alternative.” wildlife, and collecting samples for pesticide testing. Every precaution is being taken to avoid impacting non- target species. See response #84 above for cancelling additional bait applications if necessary. 87 Prof. Philip “While section 3.3.6.8 sets out a good summary of 3.3.6.8 We currently do not have enough Seddon Director, information about the ducks, I found some details to information to model the Postgraduate be missing, specifically it would have been good to demographic impact of mice on Wildlife Program, know more about the current demographic impact of Laysan Ducks.. We are currently Department of mice on the ducks, especially in relation to other doing diet overlap studies however, Zoology, extant threats and pressures. This would enable a so may be able to predict some University of more quantified assessment of the relative competitive release with the removal Otago, New importance of the mouse eradication for and on the of mice from the ecosystem. Zealand ducks. For example, if the impact of mice was Recovery of the Midway Atoll duck currently severe and limiting, say, reproductive population is expected to be quick output, then it would be potentially possible to given the duck’s ability to rebound anticipate the demographic response to removal of after prior adverse events, such as mouse-related pressure, so called “rebound”. botulism outbreaks and tsunami Currently the document mentions only that the flooding, as well as following “population should recover quickly” - this seems translocation efforts that took place unhelpfully vague. Given the critical importance of from Laysan Island to Midway Atoll, this Laysan duck population it seems worth and from Midway Atoll to Kure attempting a simple modelling exercise that projects Atoll. and compares population trends under different scenarios.” 88 Richard Supports the project and assumes that the USFWS is N/A Comment noted. Schumacher consulting with other governments on their rodent eradication programs. 89 Shane Ferkovich “I propose a habitat specifically made for the N/A Comments noted. Sand Island’s status albatross be constructed in the atoll. I do not believe is that it is a portion of MANWR. eradicating the rodent population is a good idea. Or There are no “crops” being cultivated a fountain made available for the rodents as well as on Sand Island; however, there is a for the albatross. I have contacted the State small hydroponic garden, which Department regarding Midway Island and its status provides fresh vegetables for island as a guano Island and be it true it is still a guano residents. The impacts to the island’s island have stated my desire to be its steward and population is discussed in Chapter 3 laid claim to it. I am not sure how an eradication of of the EA. the rodent population would affect any crops being cultivated but I do not think an eradication would be best until their impact can be determined with a population being on the island.” 90 Valerie Curtis Supports the project N/A Comment noted.

91 Walter M. Supports the project N/A Comment noted. Aikman, Ph. D. 92 Wyl Recommended a certain type of mouse trap to use. 2.3.3 Comment noted.

93 Wieteke Supports the project N/A Comment noted. Holthuijzen 94 “ “ “Is it possible to list what plant species have been 3.3.6.4.3, Mouse diet analysis are currently impacted by mice or sharing other anecdotal 3.3.6.1.3 under way to determine what species, information? What specific terrestrial invertebrate plants and terrestrial invertebrates, species are thought to be impacted by house mice?” that the house mouse consumes. The seeds of native grasses such as Eragrostis variabilis are a preferred food of the house mouse and restoration efforts on Sand Island have been compromised in the past by mouse herbivory of these plants.

95 “ “ “Base operations staff on Midway Atoll NWR are 2.2.1.3 The majority of the Thai Nationals mostly Thai Nationals and complete a wide variety will continue to do their duties to of infrastructure maintenance and related services. I keep Midway Atoll operational. am assuming many of the Thai Nationals that are Meetings will be held periodically in associated with Chugach/DBSI will aid with the the planning and mouse eradication project. How will the Thai implementation phase of an Nationals specifically help with this project? This is eradication to inform Thai Nationals not clear from the Draft Environmental Assessment. and other island residents to ensure More importantly, will there be proper translation the best chance of a successful pieces in place to ensure that they understand the scope, extent, and logistics of the eradication, as eradication. A commensal food plan well as activities to avoid during eradication (e.g., is currently underway in which all properly storing and preparing food, proper waste island residents will be involved in disposal, etc.?” that outlines expectations for properly storing, preparing, and disposing of food items. Instructions and information will be translated into Thai both in writing and verbally at coordination meetings with the island community. This model was successfully implemented during the rat eradication project at Wake Atoll where a large proportion of the island residents were also from Thailand. 96 “ “ “Although helicopter operations were described 2.3.1.1 The risk of helicopter-bird collisions briefly, could you address how USFWS, IC, and was considered in the DEA and partner groups will minimize the risk of bird mitigation measures were collisions in terms of time of day that the helicopter incorporated (p.3-48, 49). The timing will be utilized? What measures will be in place to of the Project during July-August is deter birds from the helicopter once it is in flight as outside the peak breeding season for well as landing or take-off areas?” most seabirds and prior to when shorebirds arrive on MANWR. Due to the significantly lower air speed of the helicopters compared to planes and jets landing at Midway, the risk of bird collisions with a helicopter would be much lower. In addition, all helicopter operations would only occur during daylight, when visibility for birds and the pilot are optimal; nocturnal seabirds would not be at risk. The helicopter will be flying at an average altitude of 164 ft. (50 m) and the bucket >134 ft. (41 m) above the ground, and these heights would allow most birds to flush below the aircraft and temporarily disperse. 97 “ “ “In collaborating with USFWS and IC, no 2.3.2 In the FEA, the protocols and monitoring protocols or methodologies of marine methodologies related to monitoring water, fish, birds, or rodents have been explicitly of non-target impacts will be shared yet that will take place during the eradication augmented to include more detail. to determine if the eradication is proceeding as These protocols are being developed planned or if there are any potential risks or by the USFWS and restoration impacts. How will these risk or impacts be partners. Monitoring populations of quantified? What “team” is drafting these protocols fish, birds, and plants and controlling or methodologies? Who will be collecting these populations of invasive species data?” currently takes place on MANWR for management purposes. The refuge staff and personnel from coordinating external agencies will follow established protocols that have been enhanced in scope for surveying and sample-collecting during pre- and post-eradication. Testing for brodifacoum residues will be conducted and analyzed by the USDA NWRC Chemistry Lab Unit in Fort Collins, Colorado. 98 “ “ “Would glue boards be an issue for Bonin Petrels, 2.3.3 Glue boards will only be used in the Laysan and/or Black-footed Albatross, or Laysan commensal environment as deemed Ducks? Please elaborate how these glue boards appropriate per the terms of each would be used and how potential negative impacts structure’s SMP (see Section 2.3.1). would be studied or quantified.” Therefore, seabirds and ducks would have no exposure to these traps. 99 “ “ “Given that Midway currently does not dispose of 2.3.4, Garbage would be stored in sealed its waste via incineration (only for very specific 3.2.11 containers at all phases (in buildings, materials) and otherwise burns all waste in an open outside, when awaiting incineration), pit or leaves it in bags (which are often ripped open and the garbage collection schedule and the contents consumed by Laysan Ducks, may likely be modified to include Common Mynas, house mice, and other species), more frequent collections and how will waste be treated before/during/after the complete incineration. The compost mouse eradication? Will all trash be collected and pile will likely be eliminated or incinerated rather than stored in the open pit? Where completely enclosed. Rodent-proof will waste be stored before it is treated (to avoid containers will be shipped to the exposure to mice, and moreover, to ensure that all Island. All personnel will be trained mice eat the rodenticide)? What about compost? As to use these waste disposal protocols for the rodent-proof containers, who will provide and these protocols will be enforced these? Will an orientation/training be provided for by the USFWS. everyone (and also be provided in Thai, both verbally and through posters or other communication forms) on-island? Who will enforce proper use of these containers before/during/after the eradication?” 100 “ “ “How will the watershed or water collection system 3.3.2.4 As discussed in Section 3.3.2 of the be closed down during the eradication? How will EA, all water for human consumption water not be collected during and after bait for a 2-year period would be deployment? If the bait remains in the system for up collected in the three enclosed tanks to 6 months, would some bait might still make it (which have a capacity of 12.6 into the drinking water collection?” million-gallons) prior to broadcasting the bait. Therefore, no water will be taken into the system for more than a year after the last bait application and no public health threat is anticipated associated with drinking water or other water use. 101 “ “ You state in Section 3.3.6.1 that it is anticipated that 3.3.6.1, See response to comment 97. In the the primary result of the Proposed Action to 3.3.6.4.5 FEA, the protocols and introduce a toxicant to achieve the purpose and need methodologies related to pre- and would be a shift in the plant and animal post-eradication monitoring will be communities and ecosystem processes on MANWR augmented to include more detail. toward a state more representative of its ecosystem These protocols are being developed before mice were accidentally introduced. by the USFWS and restoration “However, it is not clear from the Environmental partners. In addition, to measure the Draft Assessment as to what longer-term monitoring recovery of listed species and any will take place. Moreover, if the desired result is to unintended consequences of better understand ecosystem response pre-/post- eradication efforts, several key eradication, how will these community shifts be components will be characterized quantified and studied? And by whom? If this is the prior to eradication, and evaluated at specific objective that USFWS hopes to complete, 1-, 3-, 5- and 10-year post-eradication specific baseline data must be collected before the intervals by the USFWS (p. B-1). eradication in order to compare post-eradication results.” 102 “ “ “Bait longevity may be high in some areas of 3.3.6.3 The longevity of bait pellets among Midway Atoll due to other factors (USFWS, habitats can be markedly different, Unpublished).” with the rate of pellet degradation Could you provide more details about bait longevity dependent on many factors including and why it would be high in certain areas? What air temperature, light, humidity, areas? What is thought to be “high” longevity? 6 rainfall, and the presence of months? 1 year? Longer?” scavengers, molds, and soil microbes that potentiate degradation. For example, the bait proposed in this EA is formulated for use in dryer sites, therefore any significant rainfall will cause it to dissolve. A high bait longevity would be greater than 3 weeks. 103 “ “ “Yes, anecdotal evidence for Solanum nelsonii 3.3.6.4.4 Hand broadcasted bait will be exists (in terms of mice predating this speices’ distributed by a team who seeds—per Greg Schubert), but to my knowledge, I systematically walk on parallel have not heard that Pritchardia remota has been transects stopping at predetermined limited by mice, since only a few individuals of this intervals to disperse pellets as evenly species have ever persisted on Midway Atoll NWR as possible by hand. This will only be (see botanical surveys and reports by Forest and done in areas not covered by the Kim Starr). Additionally, because S. nelsonii is helicopter bait drop. The locations of widely used in propagation, outplanting, and federally-listed plants are only seeding, it would be nearly impossible to put bait marked to avoid being trampled around this plant by hand everywhere. How does during hand broadcasting of bait. We USFWS propose doing this?” do not plan to place bait around the specific locations of plant species. 104 “ “ “This list of common invasive plant species are not 3.3.6.4.5 Comment noted. necessarily the most common nor the most invasive. USFWS staff (specifically Ann Humphrey and NWRA consultant Dr. Robert Taylor) spent months compiling a list of the most invasive plant species based on a weed risk analysis (please refer to the “Midway Plant List” produced and compiled by Dr. Robert Taylor). Several USFWS staff (Ann Humphrey, Meg Duhr-Schultz, Dr. Robert Taylor) have worked on creating a more stringent policy for avoiding the invasion of new plant species (i.e. biosecurity), but this effort was ignored and sidelined by current management.” 105 “ “ “How will this ultimately be decided and by what 3.3.6.8 The plan to protect Laysan Ducks has agencies, partner organizations, etc.? Who will be been written with the detailed advice hired to take care of these birds in captivity on Sand of a working group of experienced and/or Eastern Island and ensure that they are safely scientists who have all worked on handled, properly fed, vaccinated with botulism Laysan Ducks for many years. The anti-toxin, etc.? When will an official plan be duck protection team will include produced or be shared publicly for review? Who experts in capture techniques for this will be involved in drafting this plan? How will species and professional veterinarians Laysan Ducks be captured and brought to Eastern and aviculturists that will be on-site Island? Are there enough facilities and boats and available throughout the entire available to do this? Will Eastern Island be able to project. All birds handled will be support all of the Laysan Ducks, if captured and health assessed, vaccinated for brought over? Who will be hired to take care of the botulism, and continually monitored Laysan Ducks? What kind of a contingency plan whether in captivity or released on will be in place if there is a large botulism outbreak Eastern Island. The specific team (which could be more likely if there are several members have not been finalized yet. hundred ducks on Eastern Island with limited water Dr. Michelle Reynolds was consulted resources)? How will Laysan Duck captivity during as part of the Laysan Duck Working the eradication be monitored or documented and Group and has provided extensive how will potential rodenticide impacts on Laysan comment and advice. A veterinarian Ducks be quantified or monitored post-eradication? will be on site for the duration of the Who will be the veterinary professional that will aid capture period, bait application with this complex process? Who is part of the period, and post application holding interdisciplinary team of scientists experienced with period, and staff will be ready to Laysan Ducks that are planning this mitigation handle botulism outbreaks and procedure? Is Dr. Michelle Reynolds part of this accidental rodenticide poisonings. team? I did not see that she was interviewed for the preparation of this proposal, despite her extensive See the revised Laysan duck background and research in Laysan Duck protection plan in Appendix C of the conservation and ecology.” FEA for more details.

106 “ “ “Currently, there is no official invasive plant control 4.3.1 Verbesina encelioides is the correct plan in place other than the Draft Terrestrial Habitat name for the species. Each invasive Management Plan. Also, Verbesina spp. should be plant species is ranked based on Verbesina encelioides.” abundance, invasiveness, ecosystem impact, wildlife impact, native plant impact, and ability to control. The USFWS analyzes each species to determine if the refuge will focus on eradicating, suppressing, no control, or if more research is necessary. This ranking is the foundation of the USFWS invasive species program. All plants and their ranking are listed in the Common Plants of Midway Atoll NWR document. 107 “ “ “I have some grave concerns in regards to Appendix In the larger 27-page biosecurity plan biosecurity, as 1) biosecurity policies have never A developed by the USFWS, the refuge been explicitly shared or described during addresses mammals, reptiles, orientations for visitors, volunteers, researchers, or amphibians, plants/seeds, and bird other transients by Refuge staff, and 2) any species along with preventative and biosecurity policies currently in place are not rapid response methods. enforced by Refuge staff. Biosecurity suggestions, such as cleaning boots and backpacks, are provided by Refuge staff, but are not checked or enforced by Refuge staff. Past Refuge staff and consultants, specifically Meg Duhr-Schultz, Ann Humphrey, and Dr. Robert Taylor had expressed concerns about this issue multiple times; moreover, botantists Forest and Kim Starr have documented through their various botanical surveys and reports that a wide variety of plant species have been brought into Midway without any kind of consultation or approval by Refuge staff. Ann Humphrey and I started to create a draft biosecurity policy specific to plant species to address this problem, but this policy draft was sidelined and ignored by current Refuge staff. The thought process from current staff, as expressed during orientations that I have attended on-island, is that due to Midway’s past and the high amount of exotic species (plants, insects, etc.), the introduction of other species is not a high priority and they do not believe that these taxa pose a threat to the Refuge’s biological resources. However, noting the millions of dollars put into Verbesina encelioides control and ongoing native plant community restoration, biosecurity is absolutely fundament to the broader restoration and protection of Midway Atoll NWR, especially with the pending mouse eradication. The biosecurity plan outlined in the Draft Environmental Assessment is a good start, but simply inadequate, especially when compared to excellent existing biosecurity protocols and policies as developed by New Zealand (refer to the “Biosecurity Strategy for New Zealand”), as well as Gough Island’s Restoration Program (https://www.rspb.org.uk/globalassets/downloads/do cuments/conservation-projects/gough-island- restoration-programme.pdf) and the Agreement on the Conservation of Albatrosses and Petrels Biosecurity and Quarantine Guidelines for ACAP Breeding Sites (https://acap.aq/en/advisory- committee/ac5/ac5-meeting-documents/87-ac5-doc- 19-biosecurity-guidelines-e1/file). The proposed biosecurity plan in the Draft Environmental Assessment needs to focus on the larger issues of invasive species—beyond mice. It should include policies and procedures related to plants, seed, pathogens, insects, and a myriad of other taxa, as well as addressing all forms of transit to/from Midway Atoll NWR through which species could be moved or introduced (to/from Midway Atoll NWR). With this mouse eradication, I hope that biosecurity will be revisited for all taxa and will be properly enforced.

U.S. Fish & Wildlife Service National Wildlife Refuge System Pacific Marine National Monuments 300 Ala Moana Blvd. Honolulu, HI 96850 808/792 9484

January 2019

Front cover: Wisdom’s mate and their chick Kiah Walker/USFWS Volunteer

Back cover: Albatross nesting at Midway Atoll © L. Young/Pacific Rim Conservation