Improving Inferences from Fisheries Capture-Recapture Studies Through Remote Detection of Pit Tags

VOl 35 NO 5 mAy 2010

Fish news legislative update Fisheries Journal Highlights FisheriesAmerican Fisheries Society • www.fi sheries.org calendar Job center

improving inferences from Fisheries capture-recapture studies through remote detection of Pit tags

the challenges of tracking Habitat restoration at See InSeRT Various spatial scales

Fisheries • v o l 35 n o 5 • m a y 2010 • w w w .f i s h e r i e s .o r g 209 New! NMT’s Adult Fish Counter

Northwest Marine Technology (NMT) is pleased to introduce our Adult Fish Counter. The Adult Fish Counter is designed to be installed in fishways and ladders to provide real time counts of fish migrating upstream or entering a hatchery.

The Adult Fish Counter includes a control box (top center) and a sensor unit (bottom right) that is installed underwater within the fish channel. The system accounts for fish passing upstream or downstream through the sensor, and for partially overlapped fish.

This first model of the Adult Fish Counter was developed for salmonids, and tallies adults and jacks separately, based on height. It graphs and displays the number of fish passing on an hourly, daily, and 30-day basis (example screens, above right) and tracks temperature. It can be programmed to send a “Target Reached” signal to an existing alarm system or another available indicator to let the operator know when a target number of fish have passed the sensors. NMT’s new Adult Fish Counter is available at an Please contact us for more introductory price of information. $14,900 USD.

Northwest Marine Technology, Inc. Northwest Marine Technology, Inc. www.nmt.us Shaw Island, Washington, USA Corporate Office Biological Services 360.468.3375210 [email protected] 360.596.9400 [email protected] • v o l 35 n o 5 • m a y 2010 • w w w .f i s h e r i e s .o r g vol 35 no 5 May 2010 217 222 Fisheries American Fisheries Society • www.fisheries.org editorial / subscription / circulation offices 5410 Grosvenor Lane, Suite 110 • Bethesda, MD 20814-2199 301/897-8616 • fax 301/897-8096 • [email protected] The American Fisheries Society (AFS), founded in 1870, is the oldest and largest professional society representing fisheries scientists. The AFS promotes scientific research and enlightened management of aquatic resources for optimum use and enjoyment by the public. It also encourages comprehensive education of fisheries scientists and continuing on-the-job training.

AFS OFFICERS FISHERIES STAFF Editors President Senior Editor Science Editors Contents Donald C. Jackson Ghassan “Gus” N. Rassam Madeleine Hall-Arber Ken Ashley President Elect Director of Doug Beard Wayne A. Hubert Publications Ken Currens COLUMN: Feature: Aaron Lerner William E. Kelso 232 FISH HABITAT First Deirdre M. Kimball 212 PRESIDENT’S HOOK Vice President Managing Editor Beth Beard Dennis Lassuy The Challenges of William L. Fisher Allen Rutherford Nurturing the Garden Jack Williams Tracking Habitat Restoration Second Production Editor Cherie Worth If we think of “gardening” as Vice President Book Review at Various Spatial Scales John Boreman Editors one way of viewing the fisheries Comparing databases allows Francis Juanes Past President Ben Letcher profession’s stewardship of aquatic evaluation of the role of scale in William G. Franzin Keith Nislow resources, then what sort of garden tracking restoration data, including Executive Director Abstract Translation lessons learned about reconciling Ghassan “Gus” N. Rassam Pablo del Monte Luna do we want to have? data at different scales and Donald C. Jackson Dues and fees for 2010 are: recommendations for streamlining $80 in North America ($95 elsewhere) for regular members, $20 in North America ($30 elsewhere) for student members, reporting. and $40 ($50) retired members. News: Katie Barnas and Stephen L. Katz Fees include $19 for Fisheries subscription. Nonmember and library subscription rates are $150 ($190). 213 Fisheries Price per copy: $3.50 member; $6 nonmember. Vision Unveiled for “Mosaic of Column: Fisheries (ISSN 0363-2415) is published monthly by the 242 guest Director’s line American Fisheries Society; 5410 Grosvenor Lane, Habitats” in the New York/New Suite 110; Bethesda, MD 20814-2199 ©copyright 2010. Certification—Promoting Periodicals postage paid at Bethesda, Maryland, and at Jersey Estuary the Importance of Fisheries an additional mailing office. A copy of Fisheries Guide for Authors is available from the editor or the AFS website, JoAnne Castagna Professionals www.fisheries.org. If requesting from the managing editor, please enclose a stamped, self-addressed envelope with As the AFS certification program your request. Republication or systematic or multiple Journal Highlights: approaches its 50th anniversary, a reproduction of material in this publication is permitted only new survey assesses its utility and under consent or license from the American Fisheries Society. 215 North American Journal Postmaster: Send address changes to Fisheries, American value to our membership. Fisheries Society; 5410 Grosvenor Lane, Suite 110; Bethesda, of Fisheries Management MD 20814-2199. Gail Goldberg

Fisheries is printed on 10% post-consumer recycled paper with soy-based printing inks. Update: NEWS: 216 lEgislation and Policy 244 AFS UNITS Elden Hawkes, Jr. Column: Advertising Index 248 Students’ Angle Feature: Excellence in Fisheries Management American Public University . . . . . 251 217 Fisheries RESEARCH Made Possible by the Fenske Emperor Aquatics, Inc...... 257 Improving Inferences from Fisheries Fellowship Heidi Ziegenmeyer Federal Energy Regulatory Commission 256 Capture-Recapture Studies through Floy Tag 251 Remote Detection of PIT Tags Forestry Suppliers ...... 250 AFS Annual Meeting Recent advances in technology 252 Continuing Education Halltech Aquatic Research, Inc. . . . 255 related to PIT tags have made Program HTI ...... 260 it possible to implement robust Hydroacoustic Technology, Inc. . . . 259 capture-recapture methods in a wide CALENDAR: Lotek ...... 243 254 FISHERIES EVENTS range of fisheries studies. Northwest Marine Technology, Inc. . 210 David A. Hewitt, Eric C. Janney, Brian Announcements: Oregon RFID ...... 246 S. Hayes, and Rip S. Shively 256 Job Center Sonotronics ...... 257 YSI ...... 231 Cover: A box housing PIT tag readers and associated equipment is mounted on a platform Tell advertisers you found them through amidst spawning Lost River suckers (Deltistes luxatus) at Sucker Spring, Upper Klamath Fisheries! Lake. Credit: u.S. Geological Survey. Fisheries • v o l 35 n o 5 • m a y 2010 • w w w .f i s h e r i e s .o r g 211 COLUMN: Donald C. Jackson AFS President Jackson PRESIDENT’S HOOK may be contacted at: [email protected].

Nurturing the Garden

This is the season when many of us waste on property where you have vested limestone bluffs and canyons, world- engage in the time-honored activities asso- responsibility. renowned smallmouth bass fisheries and, ciated with working on landscapes, large Gardening certainly is not restricted to as the region was discovered by tourists, a and small, deciding just how our relation- terrestrial environments. Aquaculture is degree of cultural innocence. ships with the earth will be expressed this intensive “gardening.” A capture fishery Fisheries professionals are not the sort year. Within the constraints of space, cli- can, in every sense of the word, be con- of people who mope around lamenting mate, social order (e.g., laws, regulations, sidered as extensive “gardening.” Choices what once was. We are action-oriented ordinances, covenants), and our willing- are made for our “gardens” and we persons with an ingrained sense of ness to remain engaged with our invest- proceed accordingly. We can and do have responsibility to take what we have ment, we make choices. The fundamental our “scruffy” places in fisheries where we and make what is needed. We under- choice is: “Do I garden or do I not?” If our allow the wild to prevail (to an extent). stand clearly that we are charged with choice is to garden, we till, fertilize, trim, We also have aquatic environments that the responsibility of engendering good plant, weed, move things around, and have been created or are highly modified relationships between humanity and living in general tidy up our domains to fit our by human activity. What do we plant in aquatic resources. Oftentimes this means unique suite of values and our long-term them? How do we till, fertilize, weed, cul- that we have to figure out how to fix goals. Some of us garden for aesthetics; tivate, and harvest? What do we do with broken things. some garden for consumptive purposes. the pests that invariably rear their ugly In the case of the lost stream fisher- Some garden for both. Some garden for heads, those unanticipated byproducts of ies for smallmouth bass in the Ozarks, economic purposes. Some gently nudge naivety or accident—and with situations the “fix” was and continues to be the nature. Some force nature to assume very where the best intentions went awry? introduction of exotic fishes. Elsewhere specific vectors. We ask ourselves the How do we, how should we, as profes- in North America, where dams have questions: “What do I plant? How will the sionals, address these “gardens?” Let’s impacted traditional fisheries, one of the things I plant interact with one another? consider an example. “fixes” has been to remove the dams in With what intensity do I intend to cul- During the twentieth century’s era of order to help restore the original fisheries. tivate? Will I leave, maintain, or protect dam building, smallmouth bass and small- But quite often the “fixes,” for broken (or some ‘scruffy’ places for wild things and mouth bass fishing in the Ozark Highlands created) inland systems at least, have been processes to prevail?” region of the central United States took to introduce fishes with hopes that they One thing is for sure, if we don’t their licks. Hypolimnetic discharges from will establish new and hopefully appro- garden, somebody else will have to do dams typically pushed temperature priately-founded fisheries. Sometimes it it for us because contemporary human regimes in the receiving tailwaters below works. Sometimes it doesn’t. We’ve been civilization must garden in order to meet thermal tolerances for smallmouth bass know to poison native fish fauna (think tangible as well as intangible needs. And, (and other fishes) and for many of the twentieth century “rough fish removals”) interestingly, in most societies, we can aquatic invertebrates in these systems. in order to establish stocks of “more desir- tell one another how to garden. We can Mitigation revolved around establishment able” fishes and then, as value systems collectively dictate (through markets, regu- and maintenance of salmonid fisheries, (and perhaps scientific understanding) lations, and other social pressures) frame- primarily for rainbow trout, and cold-toler- evolved, to poison the previously stocked works for production and quality control. ant invertebrate fauna. fish in order to “bring back the natives.” We can require sensitivity to “scruffy” The dams comprehensively changed As stewards of a modified landscape (N places. We can force the “gardeners” to the geographic and socioeconomic land- = 1: good old planet Earth) we truly are, consider scales beyond their individual scapes of the region. The dams brought in every sense of the word, “gardeners.” domains. Just as interestingly, if we per- electricity, some degree of flood manage- We plant what we think will bear fruit: sonally decide to be disengaged, we can ment, productive new trout fisheries, and wheat, rice, potatoes, tomatoes, straw- be forced into engagement. If you are a new opportunity for entrepreneurism berries, apples, longleaf pine, eucalyptus, member of a community (and most of us (primarily through recreation and tour- roses, tulips, cattle, poultry, swine, sheep, are), try going a full year without mow- ism) to an economically-depressed region. ring-neck pheasant (and other Asian ing your yard, ignoring water restrictions, However, many things were lost, including Continued on page 246 or failing to clean up solid or chemical free-flowing coolwater streams, gorgeous

212 Fisheries • v o l 35 n o 5 • m a y 2010 • w w w .f i s h e r i e s .o r g JoAnne Castagna News: Castagna is a technical writer/editor Fisheries with the U.S. Army Corps of Engineers, New York City, and may be contacted at [email protected].

The Comprehensive Restoration Plan includes 11 priority targets for ecosystem restoration.

Vision Unveiled Credit: USACE. for “Mosaic of Habitats” in the New York/New Jersey Estuary The U.S. Army Corps of Engineers tunities, and a boost to the region’s sponsor. The New York-New Jersey unveiled an innovative, comprehensive, economy. Harbor Estuary includes not only the restoration plan that was created in “The primary goal of the New harbor, but also rivers, wetlands, coast- collaboration with the New York-New York-New Jersey Harbor Estuary lines, and open waters, and is located Jersey Harbor Estuary Program and Comprehensive Restoration Plan is to within a complex ecological system more than 60 partnering organizations, develop a mosaic of habitats that pro- inside a metropolitan region with a including federal, state, and local agen- vides maximum ecological and societal population of 20 million people. The cies; non-governmental organizations; benefits to the region,” said Lisa Baron, plan’s boundary covers a large region and regional stakeholders. The plan project manager and marine biologist of the estuary, which is a 25-mile radius involves many partners because the with the U.S. Army Corps of Engineers around the Statue of Liberty National New York-New Jersey Harbor Estuary New York District. Monument. spans 1,600 square miles across New Baron, along with a diverse group of “To perform restoration work in the York and New Jersey. Restoring the technical experts and consultants with estuary, the plan divides the estuary estuary will not only create a healthier the Corps New York District, devel- into eight regional areas associated environment for fish and wildlife, but oped the plan as part of the Hudson with specific watersheds,” said Peter it will also provide the public cleaner Raritan Estuary Ecosystem Restoration Weppler, chief of the Army Corps New waters, healthier fisheries, increased Study with the Port Authority of New York District Coastal Ecosystem Section. flood protection, recreational oppor- York and New Jersey, the project’s local The plan includes 11 priority targets for restoration, recognized as Target With the skyline of Manhattan in the background, Col. John Ecosystem Characteristics, that include R. Boulé II, Commander of the Army Corps’ New York District, methods to restore and create habitats, addresses a group of waterfront leaders aboard the Army Corps’ vessel Hayward in the harbor. ensure these habitats live in harmony (Photo: Keegan O’Connell-Lilly, USACE). and with the surrounding urban infrastructure, and ensure the estuary is safe and accessible to the millions of estuary residents and visitors.

Habitat Restoration

The habitats that are being considered for restoration and creation include coastal wetlands, shellfish reefs,

Fisheries • v o l 35 n o 5 • m a y 2010 • w w w .f i s h e r i e s .o r g 213 News: Fisheries

islands for waterbirds, coastal and mari- Waterfront leaders and Army Corps time forests, and eelgrass beds. personnel assemble on the deck Coastal Wetlands—Coastal of the Corps vessel Hayward for a wetlands are the regional areas that group photograph near the Statue of Liberty National Monument. connect the estuary’s open waters to (Photo: Christopher Gardner, dry land. USACE). “Due to industrialization, nearly 80% of the wetlands have been lost and most of what remains is degraded. Wetlands provide such an incredible benefit to the region, on so many levels. Their vegetation provides a critical habitat for wildlife, fish, and migratory birds. We live in an urban environment where there are lots of hardened sur- faces and surface water runoff. Surface a buffer from flooding for our coastal from the water and improve water water runoff is water, from rain, snow- areas and the communities living clarity. melt, or other sources that flows from there,” said Jodi McDonald, chief of Islands for Waterbirds—The plan the land surface into water ways, which the Army Corps New York District will restore island habitats for birds can bring with it contaminants from the Ecosystem Restoration and Flood Risk whose numbers have declined consid- land. Wetlands filter and detoxify our Management Section. The plan has erably in the estuary due to hunting, water by catching contaminated sedi- identified over 26,000 acres of wet- pollution, and habitat loss. The estu- ments in the water,” said Baron. lands throughout the estuary that is ary supports 300 species of birds that “Wetlands are also nature’s suitable for coastal wetlands creation provide a vital role in the health of the sponges and act as shoreline barriers and restoration. estuary’s environment and help regu- and stabilizers. They diminish wave ShellfishR eefs—One of the plan’s late the population of other species by impact, reduce erosion, and provide priorities is to restore and consuming them for food. create shellfish reefs includ- Coastal and Maritime Forests— ing oysters, the keystone Coastal and maritime forests are found species for the estuary, on the fringe of seacoast habitats mussels, and clams, as well behind the dunes or a wetland and as other shellfish. These provide critical upland habitat. Maritime reefs are intricate underwa- forests have trees that are often stunted ter structures made up of by salt spray and high winds and they live shellfish and layers of may grow in unusual, gnarled shapes. empty shells. Oyster reefs These hardy forests are needed by have been almost eliminated due to poor water quality many species because they provide a conditions, disease, and home, food, and a nesting place for over-harvesting. In fact, the migratory birds. They also contribute estuary supported a thriving to shore stabilization and flood control, oyster industry up until the as well as moderate global climate late 1800s, covering approxi- change. mately 200,000 acres. Eelgrass Beds—Eelgrass used to These reefs provide nooks be plentiful. The plant’s long shoots and crannies and potential provide habitat and food for fish nurs- nursery grounds for other eries and estuary wildlife, and improves species because they pro- water quality through the production vide hiding places, feeding of oxygen in the water. Eelgrass also grounds, and egg attach- reduces shore erosion and improves ment sites for many species. water clarity and quality by acting as The eight regional areas encompassed in the Comprehensive Shellfish also improve water a sediment trap and filtering contami- Restoration Plan. A star in the Upper Bay regional area denotes quality by filtering sediment nants from the water. the site of The Statue of Liberty National Monument. Credit: USACE. Continued on page 250

214 Fisheries • v o l 35 n o 5 • m a y 2010 • w w w .f i s h e r i e s .o r g Volume 30 Issue 1 Journal Highlights: February 2010 North American Journal of Fisheries Management

To subscribe to AFS journals go to www.fisheries.org and click on Publications/Journals.

δ13C and δ15N Signatures in Muscle and Fin Tissues: Nonlethal Hierarchical Modeling of Juvenile Chinook Salmon Survival as a Sampling Methods for Stable Isotope Analysis of Salmonids. Function of Sacramento–San Joaquin Delta Water Exports. Ken B. Justin R. Hanisch, William M. Tonn, Cynthia A. Paszkowski, and Garry J. Newman and Patricia L. Brandes, pages 157-169. Scrimgeour, pages 1-11. Habitat Selection and Spawning Success of Walleyes in a [Management Brief] Comparison of Green and White Mesh Tributary to Owasco Lake, New York. Marc A. Chalupnicki, James H. Trammel Nets and Gill Nets to Assess the Fish Community in a Johnson, James E. McKenna Jr., and Dawn E. Dittman, pages 170-178. Large River. Greg A. Wanner, Robert A. Klumb, Dane A. Shuman, Kirk A Comparison of Seasonal Movement Patterns of Yellow Perch in Steffensen, Sam Stukel, and Nicholas J. Utrup, pages 12-25. Simple and Complex Lake Basins. Nicholas B. Radabaugh, William F. [Management Brief] Effect of Radio-Tagging on Escape Reactions Bauer, and Michael L. Brown, pages 179-190. of Adult Blueback Herring to Ultrasound. Dennis J. Dunning and Performance of Three Alternative Estimators of Stream Residence Quentin E. Ross, pages 26-32 Time Based on Live and Dead Counts of Salmonids. John R. Skalski, Movement and Microhabitat Associations of Guadalupe Bass Shiquan Liao, and Rebecca A. Buchanan, pages 191-208. in Two Texas Rivers. Joshuah S. Perkin, Zachary R. Shattuck, Preston Mandatory Catch and Release and Maximum Length Limits for T. Bean, Timothy H. Bonner, Ekaterina Saraeva, and Thomas B. Hardy, Largemouth Bass in Minnesota: Is Exploitation Still a Relevant pages 33-46. Concern? Andrew J. Carlson and Daniel A. Isermann, pages 209-220. Effects of Redd Superimposition by Introduced Kokanee on [Management Brief] Creel Survey Methods to Assess Catch, Loss, the Spawning Success of Native Bull Trout. Matthew A. Weeber, and Capture Frequency of White Sturgeon in the Snake River, Guillermo R. Giannico, and Steven E. Jacobs, pages 47-54. Idaho. Joseph R. Kozfkay and Jeff C. Dillon, pages 221-229. A Comparison of the Survival and Migratory Behavior of Effects of Water Hardness on Size and Hatching Success of Silver Hatchery-Reared and Naturally Reared Steelhead Smolts in the Carp Eggs. Jeff J. Rach, Greg G. Sass, James A. Luoma, and Mark P. Alsea River and Estuary, Oregon, using Acoustic Telemetry. Steven Gaikowski, pages 230-237. L. Johnson, James Harry Power, Derek R. Wilson, and James Ray, pages Factors Associated with Mortality of Walleyes and Saugers 55-71. Caught in Live-Release Tournaments. Harold L. Schramm Jr., Bruce Population Composition, Migration Timing, and Harvest of Vondracek, William E. French, and Patrick D. Gerard, pages 238-253. Columbia River Chinook Salmon in Late Summer and Fall. M. A. Evaluation of Visible Implant Elastomer (VIE) for Marking Jepson, M. L. Keefer, G. P. Naughton, C. A. Peery, and B. J. Burke, pages Fingerling Blue Catfish. Megan M. Zeller and Stefan H. Cairns, pages 72-88. 254-258. [Management Brief] Estimating the Size of Steelhead Runs by A Six-Decade Portrait of Florida Marine Fisheries via Landings- Tagging Juveniles and Monitoring Migrants. David A. Boughton, Based Trophodynamic Indicators. Joseph Munyandorero and pages 89-101. Cameron B. Guenther, pages 259-280. Rotenone Toxicity to Rainbow Trout and Several Mountain Short- and Long-Term Evaluation of Passive Integrated Stream Insects. Brian Finlayson, William L. Somer, and Mark R. Vinson, Transponder and Visible Implant Elastomer Tag Performance in pages 102-111. Muskellunge. Jerry A. Younk, Brian R. Herwig, and Bruce J. Pittman, Associations between Watershed Characteristics and Angling pages 281-288. Success for Sport Fishes in Mississippi Wadeable Streams. J. Brian [Management Brief] Deriving Acceptable Biological Catch from the Alford and Donald C. Jackson, pages 112-120. Overfishing Limit: Implications for Assessment Models. Michael H. [Management Brief] Survival of Shovelnose Sturgeon after Prager and Kyle W. Shertzer, pages 289-294. Abdominally Invasive Endoscopic Evaluation. Drew G. Trested, Early Life Stage Mortality Rates of Lake Sturgeon in the Peshtigo Reuben Goforth, James P. Kirk, and J. Jeffery Isely, pages 121-125. River, Wisconsin. David C. Caroffino, Trent M. Sutton, Robert F. Elliott, Historical and Current Population Characteristics and Subsistence and Michael C. Donofrio, pages 295-304. Harvest of Arctic Char from the Sylvia Grinnell River, Nunavut, [Management Brief] Comparison of Channel Catfish Age Estimates Canada. Colin P. Gallagher and Terry A. Dick, pages 126-141. and Resulting Population Demographics Using Two Common Structures. Robert E. Colombo, Quinton E. Phelps, Candice M. Miller, Estimating Survival and Migration Route Probabilities of Juvenile James E. Garvey, Roy C. Heidinger, and Nathaniel S. Richards, pages Chinook Salmon in the Sacramento–San Joaquin River Delta. 305-308. Russell W. Perry, John R. Skalski, Patricia L. Brandes, Philip T. Sandstrom, A. Peter Klimley, Arnold Ammann, and Bruce MacFarlane, pages 142- Fishing Technologies and Catch Levels among Small-Scale Fishers 156. in Lagos State, Nigeria. K. A. Akanni, pages 309-315.

Fisheries • v o l 35 n o 5 • m a y 2010 • w w w .f i s h e r i e s .o r g 215 Update: Elden Hawkes, Jr. lEgISlATION ANd POlICy aFs Policy coordinator Hawkes can be contacted at ehawkes@fi sheries.org.

new northeast groundfi sh member vessels. Member vessels with Mexico’s loss of its certifi cation Management Measures are exempt from many area and in April to export its wild-harvest Noaa fisheries service has gear restrictions, but must agree to shrimp to the United States. The announced new measures that stop fi shing once the sector catches certifi cation was withdrawn after the will establish new catch limits and its allotment of fi sh. Solo fi sher- u.s. National Marine fisheries service major changes in how the Northeast men will still be allowed to fi sh, but determined that Mexico’s turtle groundfi sh fi shery (which includes must comply with strict limits on the excluder devices no longer meet U.S. cod and fl ounder) will be managed. number of fi shing days, trip limits standards. U.S. rules require that for some species, and seasonal and The new measures are intended to exporters use excluders comparable area closures. all vessels can also end the overfi shing of Northeast to those used by U.S. shrimpers. increase that allotment by leas- groundfi sh and to continue the Mexico’s National Fisheries Council ing and trading shares of catch or rebuilding of the fi shery. is working with U.S. experts to fi shing days. The new measures for The new measures include a recertify its shrimp fl eets follow- Northeast groundfi sh will be effec- ing new inspections in August and fi rst time cap on the amount of all tive 1 May 2010. groundfi sh of any species that are September. permitted to be caught. In addition, there will be measures to mitigate shrimp bans hit Mexico on both european union subsidies linked if the catch limits are exceeded. sides of border to overfi shing In addition to the new caps, the Mexico’s National Commission for An EU Financial Instrument for measures include fundamental Aquaculture and Fisheries instituted Fisheries Guidance March 2010 changes to the way that the fi shery a ban on shrimp fi shing in March study links EU fi sheries subsidies is managed. One of these changes that is designed to contribute to the and the overfi shing of valuable fi sh includes allowing fi shing vessels renewal of stocks and to protecting stocks. The study was carried out to fi sh in “sectors.” Sectors are the reproductive process to ensure in 10 European countries including voluntary groups for fi shing vessels, future harvests. The temporary ban Britain, Denmark, France, and Spain. that can formed each year and are on shrimp fi shing includes all marine The examined countries account for given a portion of the total avail- waters under federal jurisdiction almost all of the 4.9 billion euros able groundfi sh catch based on the in the Pacifi c Ocean, including the of fi shing subsidies handed out combined fi shing history of their Gulf of California, Lagunari estuarine marshes, by Brussels from 2000–2006. The and bays of report states that 29% of the EU the states of handouts went to measures which Baja California contributed to overfi shing, while sur, sonora, just 17% were dedicated to measures to support healthy fi sheries. Fast-track your paper by submitting it to Marine and Coastal Fisheries. Sinaloa, and The report also indicated that Spain, AFS’ open access, online, international journal.. Nayarit. The ban France, Portugal, and Germany in imposed by particular have benefi ted from sub- Mexico’s sidies for “negative measures” as National far as sustainable fi sheries are con- Commission cerned. The fi ndings of the report www.fi sheries.org/mcf for state that among the key European Aquaculture fi sh stocks where overfi shing has and Fisheries been enabled are southern hake, is coupled monkfi sh, sharks, and prawns.

216 Fisheries • v o l 35 n o 5 • m a y 2010 • w w w .f i s h e r i e s .o r g Feature: T upper B lake , U.S. F ish and W ildlife S ervice Fisheries RESEARCH

Improving Inferences from Fisheries Capture-Recapture Studies through Remote Detection of PIT Tags Lost River sucker (Deltistes luxatus)

ABSTRACT: Models for capture-recapture data are commonly used in analyses of the dynamics David A. Hewitt, of fish and wildlife populations, especially for estimating vital parameters such as survival. Eric C. Janney, Capture-recapture methods provide more reliable inferences than other methods commonly used in fisheries studies. However, for rare or elusive fish species, parameter estimation is Brian S. Hayes, often hampered by small probabilities of re-encountering tagged fish when encounters are and Rip S. Shively obtained through traditional sampling methods. We present a case study that demonstrates how remote antennas for passive integrated transponder (PIT) tags can increase encounter Hewitt is a research fishery probabilities and the precision of survival estimates from capture-recapture models. Between biologist and Janney and 1999 and 2007, trammel nets were used to capture and tag over 8,400 endangered adult Lost Hayes are fishery biologists River suckers (Deltistes luxatus) during the spawning season in Upper Klamath Lake, Oregon. at the U.S. Geological Despite intensive sampling at relatively discrete spawning areas, encounter probabilities from Survey (USGS), Western Cormack-Jolly-Seber models were consistently low (< 0.2) and the precision of apparent Fisheries Research annual survival estimates was poor. Beginning in 2005, remote PIT tag antennas were deployed Center, Klamath Falls at known spawning locations to increase the probability of re-encountering tagged fish. We Field Station, Klamath compare results based only on physical recaptures with results based on both physical recaptures Falls, Oregon. Shively is a and remote detections to demonstrate the substantial improvement in estimates of encounter fishery biologist at USGS, probabilities (approaching 100%) and apparent annual survival provided by the remote Columbia, Missouri. detections. The richer encounter histories provided robust inferences about the dynamics of Hewitt can be contacted at annual survival and have made it possible to explore more realistic models and hypotheses [email protected]. about factors affecting the conservation and recovery of this endangered species. Recent advances in technology related to PIT tags have paved the way for creative implementation of large-scale tagging studies in systems where they were previously considered impracticable. Mejoramiento de las inferencias obtenidas en estudios de captura-recaptura mediante detección remota de marcas internas

Resumen: En los análisis de dinámica poblacional de peces y poblaciones naturales es común utilizar los modelos de captura-recaptura, especialmente para estimar parámetros clave como la supervivencia. Los métodos de captura-recaptura brindan estimaciones más confiables que otros métodos frecuentemente utilizados en estudios pesqueros. No obstante, para especies de peces raras o particularmente elusivas, la estimación de los parámetros a veces es afectada por la baja probabilidad de reencuentro de peces marcados cuando éste se basa en métodos tradicionales de muestreo. Se presenta un caso de estudio que demuestra cómo las antenas remotas para marcas electromagnéticas internas (PIT tags) pueden incrementar la probabilidad de encuentro y la precisión de las estimaciones de supervivencia derivadas de los modelos de captura-recaptura. Entre 1999 y 2007, mediante una red de trasmallo, se capturaron 8,400 individuos adultos del matalote (Deltistes luxatus) durante la época reproductiva en el Lago Upper Klamath, Oregon. A pesar de haber realizado un intenso muestreo en zonas de reproducción relativamente bien delimitadas, las probabilidades de encuentro calculadas con los modelos Cormack-Jolly- Seber fueron consistentemente bajas (< 0.2) al igual que la precisión de las estimaciones de supervivencia anual. Al inicio de 2005, se colocaron antenas remotas en sitios conocidos de reproducción para rastrear marcas electrónicas internas y aumentar las probabilidades de reencuentro de los peces marcados. Con el fin de demostrar una mejora sustancial en las estimaciones de las probabilidades de encuentro (cerca del 100%) y de la supervivencia anual que se obtienen utilizando detección remota, se compararon los resultados tanto sobre la base de individuos recapturados como de individuos recapturados y detecciones remotas. El grupo de datos con un mayor número de reencuentros sirvió para hacer inferencias más robustas acerca de la dinámica anual de la supervivencia e hizo posible explorar modelos más realistas e hipótesis sobre los factores que afectan a la conservación y recuperación de esta especie amenazada. Los avances tecnológicos recientes relacionados a las marcas electrónicas internas, han preparado el camino para implementar estudios de marcado a gran escala en sistemas en los que dichos trabajos se consideraban impracticables.

Fisheries • v o l 35 n o 5 • m a y 2010 • w w w .f i s h e r i e s .o r g 217 Introduction to include CR studies and modern methods of analysis in their toolbox for estimating survival and other demographic Management and conservation of animal populations parameters (Pine et al. 2003). Perhaps the primary reason for depend on an understanding of key life history parameters the tepid response is that fisheries capture-recapture studies and their roles in regulating population dynamics. For exam- are often difficult to implement. Three concerns are comple, populations of fish species with low natural survival rates monly expressed: are often more productive and can support more intense exploitation than species with higher natural survival rates 1. Cost and effort associated with tagging and recapture sam- (Adams 1980). Such populations are said to be evolutionarily pling is prohibitive; adapted to low survival and compensate through rapid growth 2. Statistical model assumptions about tag retention and the or high fecundity. For long-lived species with high natural effects of tagging on behavior and survival are hard to survival rates, populations can be rapidly depleted by har- meet; and vest and other sources of anthropogenic mortality (Fujiwara 3. Capturing and tagging a subset of fish that can be con- and Caswell 2001), and population growth rates are sensitive sidered representative of the population as a whole is to variability in survival (Pfister 1998; Doherty et al. 2004; difficult. Schmutz 2009). In some situations, populations of imperiled We suggest that the first two concerns about costs and model species may require intervention to increase or stabilize sur- assumptions can usually be overcome by careful planning and vival and reduce the risk of extinction. design, and that the benefits to inference about population Accurate and precise estimates of survival are needed to dynamics from CR studies outweigh the costs. Tag retention evaluate hypotheses about factors that influence population and effects of tagging can be assessed with pilot or comple- dynamics and to develop effective management strategies. Capture-recapture (CR) or tagging studies are arguably the mentary studies, although a reliable tagging method with most reliable methods for generating such estimates, and minimal adverse effects remains a prerequisite for robust development of theory and methods for analysis of CR data inference. Fortunately, a large body of literature is available has been exceptionally rapid in the past few decades (Seber on tag types and tagging techniques for fishes (Parker et al. and Schwarz 2002; Senar et al. 2004; Thomson et al. 2009). 1990; Nielsen 1992; Guy et al. 1996), as well as design and Researchers are now able to use CR data to directly evalu- analysis of CR studies (Burnham et al. 1987; Pollock et al. ate factors affecting not only survival (Burnham et al. 1987; 1990; Williams et al. 2002). Lebreton et al. 1992; Nichols 2005), but also recruitment and Concerns about the representativeness of the tagged subset population growth rate (Pradel 1996; Nichols et al. 2000; of fish are important, but they should not prohibit important Nichols and Hines 2002), movement or migration (Schwarz inferences from being made in most cases. Representative and Arnason 1990; Schwarz et al. 1993; Schwarz 2009), and sampling is important for inferences based on CR, as it is for reproductive success (Nichols et al. 1994; Rotella 2009). any statistical analysis of sample data. However, most fish- Most recent developments can be viewed as special cases of a eries studies require CR models that are applicable to open flexible class of models that treat individual animals as occu- populations; that is, those that undergo change due to births, pying one of a number of states, broadly defined, in any given deaths, or migration during the study period. Inferences about time period (Lebreton and Pradel 2002; White et al. 2006; dynamics in open populations can be restricted to the tagged Bailey et al. 2009; Kendall 2009). subset of fish using models of the Cormack-Jolly-Seber (CJS) Borrowing theory and methods from generalized linear type (Lebreton et al. 1992). The CJS models are conditioned models, CR models can incorporate and evaluate the effects of only on the encounter histories of tagged fish; as a result, CJS variables (covariates) on model parameters, permitting evalua- models avoid the numerous pitfalls associated with estimat- tion of interesting biological hypotheses (Lebreton et al. 1992; ing population size (reviewed in Cormack 1968 and Williams Franklin 2001; Bonner and Schwarz 2004; Nichols 2005; Cam et al. 2002, which contrast with Hayes et al. 2007). Because 2009; Conroy 2009). Models can be fit by maximum likeli- formal statistical inference is restricted to the tagged fish, hood with free software, and competing models that repre- generalization of inferences from CJS models to the popula- sent various hypotheses can be compared in a model selection tion as a whole must be based on the adequacy of the study framework (White and Burnham 1999; Choquet et al. 2004). design. Provided that tagged fish are reasonably similar to the Importantly, a model selection framework avoids inappropri- rest of the population, results should be useful in quantifying ate interpretations of classical statistical tests as strength of dynamics and evaluating hypotheses. Scientific inference is evidence (Royall 1997), leads to a parsimonious interpre- a process, and sampling and modeling can always be adapted tation of the data as represented by models, and provides a to address important sources of variation that are expected or means to account for model selection uncertainty in estimates discovered in the population, such as those that might affect of model parameters and their variances (see Chatfield 1995 the representativeness of the tagged subset of fish. and Buckland et al. 1997; part of multimodel inference sensu A lingering and critical limitation for CR studies of fish Burnham and Anderson 2002 and Anderson 2008). populations is the need for relatively high probabilities of Capture-recapture data have been and remain integral re-encountering tagged fish. High encounter probabilities are to studies of fish stocks in marine and coastal ecosystems, essential for estimating parameters of interest with satisfac- most often in the form of tag returns from fishermen that are tory precision and, of equal or greater importance, for evaluat- used in stock assessment analyses. Despite all of the advan- ing model assumptions (Cormack 1968; Burnham et al. 1987; tages of CR, freshwater fisheries researchers have been slow Lebreton et al. 1992). Unfortunately, recapture probabilities

218 Fisheries • v o l 35 n o 5 • m a y 2010 • w w w .f i s h e r i e s .o r g for tagged fish are often low in large water bodies, including studies in situations where they were previously considered many lakes and rivers and most estuarine and marine systems, impracticable. and enormous sample sizes are needed to make inferences in such situations (e.g., Jiang et al. 2007). Low recapture prob- Study species and system abilities are particularly common when the population is diffuse or the species is otherwise difficult to capture with tra- Lost River suckers are long-lived catostomids endemic to ditional gears. The trade-off between encounter probabilities the Upper Klamath River Basin in Oregon and California and sample size in study design has led to what is known as (Miller and Smith 1981; Scoppettone and Vinyard 1991). the “big law” of CR—increase encounter probabilities by any Individuals have been aged to over 40 years and the larg- means possible (Figure 1). As a general guideline, encounter est adult females can grow to 800 mm fork length (FL; probabilities should be 0.2 or higher for modeling and infer- Scoppettone 1988; Scoppettone and Vinyard 1991; Janney et ence to be fruitful without unreasonable sample sizes. al. 2008). Lost River suckers were listed as endangered under In this article, we describe the development of a capture- the U.S. Endangered Species Act in 1988 because of range recapture monitoring program for endangered Lost River contractions, declines in abundance, and a lack of evidence suckers (Deltistes luxatus) in which we had to overcome of recent recruitment to adult populations (USFWS 1988). the problem of low encounter probabilities. Intensive sam- Direct mortality from subsistence and recreational fisher- pling with traditional gears was unable to provide sufficient ies for spawning suckers may have contributed to popula- recaptures of fish tagged with passive integrated transponder tion declines, but fishing for the suckers was banned in 1987 (PIT) tags, but creative use of remote antennas increased (USFWS 1993). Numerous other threats common to imper- encounter probabilities to nearly 100% and provided more iled fishes in the western United States were identified as robust model-based inferences about population dynam- potentially contributing to the declines (e.g., habitat altera- ics. We suggest that recent advances in technology for PIT tion and degradation, nonnative species), but the relative tags and antennas have made it possible to overcome the influence of the various causes is uncertain (NRC 2004). problem of low encounter probabilities in many systems, The most intensively studied remaining population of thus allowing for the implementation of capture-recapture Lost River suckers occurs in Upper Klamath Lake, Oregon (UKL). Two apparently distinct spawning subpopulations of Figure 1. An example of the “big law” of capture-recapture studies— Lost River suckers coexist in UKL (Janney et al. 2008). One increase encounter probabilities by any means possible. The example is subpopulation exhibits a reproductive strategy similar to based on a simple treatment-control experiment to estimate the effect other western lakesuckers (genus Chasmistes) and migrates of dam turbine passage on survival, and is based on the equation in relatively short distances up tributaries to spawn in the Burnham et al. (1987:315). The effect size (E) is the ratio of survival for the treatment group to survival for the control group, and is set to 0.8, spring. Although spawning may have occurred in other trib- representing a 20% reduction in survival due to passing through the utaries in the past, nearly all riverine spawning activity for turbine. The study is a true experiment and the two groups of fish are the suckers is now restricted to the lower Williamson River assumed to be identical except for their probability of survival. The total required number of tagged and released fish, divided equally among and the Sprague River (Figure 2). The other subpopulation treatment and control groups, is plotted against the probability of spawns at upwelling springs along the eastern shore of the recapturing a released fish for three target values of the coefficient of lake below Modoc Rim. The majority of spawning activity variation (CV) for the estimated effect size: 2.5%, 5%, and 10% for both subpopulations occurs in March and April. se(E^ ) [CV(E^) = x 100] Impaired water quality conditions in Upper Klamath Lake E^ have been implicated in reduced survival of adult suckers and Increasing the probability of recapturing a tagged and released fish yields are a concern for recovery efforts. Upper Klamath Lake is the strongly disproportionate reductions in the required number of fish that 2 need to be tagged to achieve a given level of precision on the effect size. largest lake in Oregon (280 km ), but is relatively shallow (aveage depth ca. 2 m). The combination of this bathymetry

20000

15000

10000

5000 2.5% CV U.S. G eological S urvey 5%

10%

Number of tagged and released fish 0

0.0 0.2 0.4 0.6 0.8 1.0 A phytoplankton bloom on Upper Klamath Lake Probability of recapturing a released fish

Fisheries • v o l 35 n o 5 • m a y 2010 • w w w .f i s h e r i e s .o r g 219 Figure 2. Map of Upper Klamath Lake showing the five spawning areas at springs along the shortnose suckers (Chasmistes breviro- eastern shoreline below Modoc Rim. Fringing marshes and other wetlands are shown in dark grey. stris). However, for simplicity of presen- The shaded area in the inset shows the location of the lake within the Klamath River Basin. tation we only describe sampling and analysis methods for the subpopulation of Lost River suckers that spawns at the shoreline springs, and we only include fish tagged in 1999 and later. Fish encountered at the springs are rarely encountered elsewhere in our sampling, so focusing on the spring spawning subpopulation greatly simplifies description. More details of life history, sampling, and analysis for both species are given in Scoppettone and Vinyard (1991), NRC (2004), and Janney et al. (2008). From 1999 to 2008 we captured Lost River suckers for tagging by setting trammel nets at five known spring spawning areas between February and May, beginning soon after ice-out (Figure 2). Nets were set twice per week at each spring, allowing three or four days between sampling events at a given spring. Sampling for the season was continued until only a few fish were captured in a given week. Trammel nets (30 m x 1.8 m; 29 cm bar outer mesh, 3.5 cm bar inner mesh) were deployed from shore by wading in a semicircle around the perimeter of the area of concentrated spawning. Spawning areas are relatively small; size varies among springs, but is typically 500 m2 or less. Nets were deployed for four hours around sunset and checked at least once per hour for newly captured fish. Captured fish were retained in floating net pens for processing and were released when sampling was completed for the night. None of the fish died as a result of handling or retention in the net pens and all released fish appeared vigorous, so with a watershed naturally enriched in phosphorus has led to the conclusion that the we consider short-term mortality asso- lake has been eutrophic since the earliest records in the mid-1800s (NRC 2004; Kann ciated with sampling to be negligible. and Welch 2005). However, water quality in UKL has been markedly altered from Captured fish were scanned for the those historical conditions by various human activities in the watershed, particularly presence of a PIT tag and untagged fish the drainage of marshes and wetlands, timber harvest, and water control and alloca- were injected with a tag using a hypo- tion related to agricultural development (Bradbury et al. 2004). These changes have dermic syringe with a 12-gauge needle. created a hypereutrophic system that experiences massive phytoplankton blooms dom- The PIT tags were injected into the inated by a single cyanobacterium, Aphanizomenon flos-aquae(Lindenberg et al. 2009). abdominal musculature anterior to the Phytoplankton blooms in the summer and fall are associated with direct and indirect pelvic girdle. All fish tagged prior to mortality of adult suckers and can lead to large fish die-offs (Perkins et al. 2000). 2005 were given 12 or 14 mm 125.0 kHz full-duplex tags, and all fish tagged in Methods 2005 and thereafter were given 12 mm 134.2 kHz full-duplex tags. Capture, tagging, and encounters of tagged fish Physical recaptures of tagged fish were obtained from the trammel net sampling Our capture-recapture program in Upper Klamath Lake has been ongoing since from 2000 to 2008, but the number of 1995 and focuses equally on adult Lost River suckers and co-occurring endangered fish recaptured by this method was con-

220 Fisheries • v o l 35 n o 5 • m a y 2010 • w w w .f i s h e r i e s .o r g sistently low relative to the number of fish previously tagged Williams et al. (2002), and Nichols (2005). The fundamental and released (Table 1). In the spring of 2005, as an experi- input to CJS models are the encounter histories (series of mental approach to increase the probability of re-encounter- zeros and ones where the ones indicate an encounter on a ing tagged fish, we opportunistically deployed a single remote given sampling occasion; e.g., 1000010111), and the param- flat plate PIT tag antenna at Cinder Flat spring (Biomark, eters of interest are apparent survival probabilities (Φ) and 30.5 x 66 cm; Figure 3). The antenna was deployed for parts encounter probabilities (p). Apparent survival includes true of 10 days late in the spawning season, between 21 April and survival as well as permanent emigration from the study area, 6 May, yielding a total sampling time of approximately 135 but in our case Φ closely approximates true survival because hours. Beginning in 2006, we deployed flat plate antennas very few adult Lost River suckers leave Upper Klamath Lake for the entire spawning season at four springs: Cinder Flat, (Banish et al. 2009). Ouxy, Silver Building, and Sucker (Figure 2). Two or three Our first model set was developed for encounter histories antennas were used at each spring, and each antenna was that included only physical recaptures from trammel net sam- connected to its own Biomark FS2001F-ISO reader by a 6-m pling, and the second was developed for encounter histories cable. Readers were stored in metal boxes on fixed platforms that included both physical recaptures and remote encoun- near the middle of the spawning areas and antennas were ters. Survival and encounter probabilities were estimated over distributed around the platforms (Figure 3). annual time intervals, so multiple encounters of the same fish in the same sampling season (February-May) were treated as Analysis and modeling a single encounter. Model sets were developed by considering the effects of sex and time (year) on Φ and p, and then Inferences about adult Lost River sucker survival have including models with and without those factors. We mod- benefited greatly from the encounters provided by the remote eled Φ as a function of sex because we expected that greater PIT tag antennas, and we illustrate the benefits by compar- reproductive investment by females would lead to lower sur- ing results from two sets of Cormack-Jolly-Seber models. vival compared to males. Most importantly, we modeled Φ Comprehensive reviews of CJS models were provided by as a function of time because we hoped to detect changes in Seber (1982), Pollock et al. (1990), Lebreton et al. (1992), annual survival and relate those changes to factors such as

Table 1. Summary of releases and re-encounters of PIT-tagged Lost River suckers in Upper Klamath Lake, Oregon between 1999 and 2008. The total annual re-encounters of tagged fish in 2000–2008 are shown at the top, for each encounter method separately (physical recapture with trammel nets [N] and detection on remote PIT tag antennas [R]) and in combination. Re-encounters for 2000–2004 include only recaptures from trammel nets; remote antennas were first used to a limited extent in 2005 (see text). Encounters for one method can include fish also encountered by the other method, so the combined encounters are lower than the sum of the per-method encounters. Re-encounters of fish in the year in which they were tagged are excluded from the counts. The remainder of the table is broken down by release cohort and shows the numbers of tagged fish from a cohort that were re-encountered for the first time in a given year. In this part of the table, successive columns include only fish that were never previously re-encountered, by either physical recapture or remote detection, and remote columns exclude fish that were also recaptured in nets.

Annual Re-encounters of Tagged Fish 2000 2001 2002 2003 2004 2005 2006 2007 2008 N R N R N R N R Total re-encounters 51 152 219 341 383 500 157 182 4,787 471 4,766 345 5,078 by type Total re-encounters 51 152 219 341 383 642 4,794 4,792 5,092 combined

First Re-encounters of Tagged Fish 2000 2001 2002 2003 2004 2005 2006 2007 2008 Individuals Individuals Release Number never encountered cohort released encountered N R N R N R N R again again 1999 744 51 52 34 44 44 18 15 7 181 1 17 0 4 468 276 2000 1,056 91 75 70 46 49 23 17 308 1 24 0 3 707 349 2001 1,262 82 99 68 71 0 17 422 7 53 3 13 835 427 2002 983 74 57 51 0 13 390 6 10 0 0 601 382 2003 1,343 79 83 57 22 753 4 26 2 7 1,033 310 2004 979 82 30 17 665 3 38 1 7 843 136 2005 1,017 27 899 1 6 0 2 935 82 2006 369 30 310 1 5 346 23 2007 698 28 605 633 65

Total 8,451 51 143 191 287 294 354 125 120 3,618 53 484 35 646 6,401 2,050

Fisheries • v o l 35 n o 5 • m a y 2010 • w w w .f i s h e r i e s .o r g 221 water quality conditions. For p, we expected sex to be impor- effect of PIT tag type (125.0 vs. 134.2 kHz) on encounter tant because of differences in reproductive behavior (e.g., probability in 2006–2008. The 134.2 kHz tags were first used males stay at spawning areas longer than females, potentially in 2005 and first re-encountered in our 2006 sampling. The increasing their probability of being encountered), and we higher frequency tags have a greater read range and are more expected time to be important because of annual differences likely to be detected on the remote antennas than lower fre- in sampling intensity and environmental effects on the con- quency tags. In our application, the distance from the anten- dition of the spawning habitat. Past analyses showed that nas at which the 134.2 kHz tags can be read is 15–20 cm, models with some combination of both sex and time effects compared with 5 cm for the 125.0 kHz tags. We included on p were overwhelmingly supported in model selection, so models that constrained the effect of tag type to be the same we only considered models with some combination of both in 2006, 2007, and 2008, as well as models that allowed the effects (Janney et al. 2008). We included models with both effect of tag type to vary by year. additive and interactive effects for Φ and p. Additive mod- We used program MARK to fit the models using maxi- els constrained effects to be the same between groups across mum likelihood (White and Burnham 1999). Models were time (e.g., the difference between male and female survival is specified and passed to MARK using the RMark package the same in each year), whereas interactive models included (Laake 2009; Laake and Rexstad 2009) within the R software more parameters and allowed effects to vary through time environment (R Development Core Team 2009). All model (e.g., separate estimates of survival for each sex in each year). likelihoods were constructed using a logit link function and Note that the last estimates of Φ and p are confounded in the optimized using the default Newton-Raphson algorithm. likelihood and cannot be separately estimated; this is a gen- Models within a set were compared using an informa- eral characteristic of CJS models. As such, we do not report tion-theoretic model selection framework and Akaike’s or discuss estimates of Φ for 2007 or p for 2008. Information Criterion corrected for small sample size and The two model sets were the same with one important overdispersion (QAICc; Burnham and Anderson 2002). For exception: the model set for the encounter histories that each model in a set, we follow Anderson et al. (2001) and included remote detections incorporated models with an report five quantities:

Figure 3. Clockwise from top left: a remote flat plate PIT tag antenna on the substrate; a box for housing the PIT tag readers and associated equipment, mounted on a platform amidst spawning suckers at Sucker Spring; an underwater photograph of spawning Lost River suckers.

222 Fisheries • v o l 35 n o 5 • m a y 2010 • w w w .f i s h e r i e s .o r g 1. The number of estimated parameters (k), geneity in encounter probabilities (Carothers 1973; Gilbert 2. QAICc, 1973; Carothers 1979). Nonetheless, unmodeled heterogene- 3. The difference in QAICc between a given model and the ity in encounter probabilities will likely remain of concern in Kullback-Leibler (K–L) best model in the set (∆QAICc), most large-scale fisheries capture-recapture studies, and we return to this issue in the Discussion. Heterogeneity can be 4. The probability that the model is the best K–L model in addressed to some extent through goodness-of-fit testing and the set (model probability, or Akaike weight, w ), and i correction for overdispersion (Lebreton et al. 1992; Burnham 5. The value of the maximized log-likelihood function and Anderson 2002). We performed goodness-of-fit tests for (–2logeL). the most general model in each model set and neither showed Evidence ratios are used to compare pairs of models and are any consistent departure from expectations under the CJS simply the ratio of model probabilities. Where possible, we model (see also Janney et al. 2008). However, a small amount account for model selection uncertainty in parameter esti- of overdispersion was evident, presumably caused by a lack mates by calculating estimates and estimated variances as of independence in the fates of tagged fish, and we corrected weighted averages from all models in the set, using the model model selection statistics and inflated parameter variances probabilities as weighting factors. to account for the overdispersion using a variance inflation factor ( cˆ ). We estimated cˆ with the median cˆ procedure in Assumptions of the Cormack-Jolly-Seber model program MARK.

The CJS model makes the following assumptions: Results

1. Tags are not lost, or missed when individuals are Summary of capture, tagging, and encounters re-encountered; 2. Sampling periods are “instantaneous” relative to the Between 1999 and 2007, 8,451 adult Lost River suck- interval between samples; and ers were captured and tagged at the spring spawning areas 3. There is no unmodeled individual variability (heteroge- (Table 1). We recaptured 2,005 of those individuals (24%) neity) in survival or encounter probabilities among the in subsequent trammel net sampling through 2008. However, tagged animals. recaptures of tagged fish in a given year were low considering the intensity of our sampling and the discrete distribution of Assumptions 1 and 2 must be addressed primarily through spawning activity. Trammel net recaptures never exceeded study design. For Lost River suckers, double-tagging experi- 500 individuals in a given year. ments with Floy and PIT tags showed that PIT tag loss rates In contrast, by including the detections of fish on remote were less than 1% over three or more years (U.S. Geological PIT tag antennas in 2005–2008, we re-encountered 6,401 of Survey, unpublished data). For physical recaptures, we the tagged individuals (76%) and total annual re-encounters ensured that tags were not missed when present by scanning increased dramatically (Table 1). In the absence of detections a test tag prior to scanning each fish, and also scanning a test from the remote antennas, 4,396 tagged fish were available in tag after each fish that was found to be untagged. Regarding the population but would not have contributed to our infer- Assumption 2, sampling in our study occurs over a 3 to 3.5 ences about survival. In 2005 alone, when a single antenna month spawning period and is not instantaneous. However, was deployed for a short period late in the spawning season at the vast majority of captures and remote encounters occur one spawning area, we detected 157 individuals that had been over a much shorter time period, and individuals are fairly tagged prior to 2005, which compares with 500 individuals consistent from year to year in the relative times at which recaptured in trammel nets over five spawning areas through- they join the spawning aggregation. Thus, on an individual out the 3-month spawning season. Of the 157 fish detected basis, sampling can be considered nearly instantaneous rela- remotely, 125 had not been previously recaptured in trammel tive to an annual interval used for parameter estimation. In nets despite being at large for as many as 6 years (Table 1). In addition, spawning fish almost always appear to be in excel- 2006, the first year of full remote antenna implementation, lent condition and water quality is good during the spring. we detected 4,787 individuals over the course of the spawn- Thus, we expect that little mortality occurs during the sam- ing season, compared to 182 individuals recaptured in tram- pling period and does not bias survival estimates. mel nets. Of the remotely detected fish, 3,618 had never been Assumption 3 regarding heterogeneity is complex and has re-encountered before by either method. Even in 2008, after received considerable attention in the capture-recapture lit- two years of full remote antenna coverage and eight years of erature (reviewed in Pollock et al. 1990; Williams et al. 2002; trammel net sampling, four individuals that had been tagged Pollock and Alpizar-Jara 2005). This assumption includes in 1999 were re-encountered for the first time only on the the important and well-recognized requirement for negligible remote antennas. effects of tagging on survival. Our observations of Lost River suckers after tagging suggest that mortality related to handling Analysis including only physical recaptures and tagging is negligible, but future analyses will investigate this further. Other aspects of Assumption 3 are often not fully Model selection results and parameter estimates indicated addressed or are completely ignored in fisheries studies (Pine that the data set that included only physical recaptures could et al. 2003). One of the primary advantages of CJS models support rather limited inferences about annual survival of is that estimates of apparent survival are robust to hetero- PIT-tagged Lost River suckers. The top model, which had a

Fisheries • v o l 35 n o 5 • m a y 2010 • w w w .f i s h e r i e s .o r g 223 0.935 probability of being the Kullback-Leibler best model 2001 point estimates for males were on a boundary and in the set, included additive effects of sex and time on Φ five of the eight estimable parameters for females were on and the same additive structure for p (Table 2). The second a boundary. Combined with much wider confidence inter- best model included the same additive structure for p but vals for the female estimates, the bottom line is that few of separate estimates of Φ for each sex in each year; this model the Φ estimates from this model were usable, despite it being had much less support (∆QAICc = 5.8, model probability = the second best model in the set. Estimation problems aside, 0.051). The only difference between this model and the top point estimates of survival probabilities from the top model model was the structure on Φ, so the evidence ratio between appeared to be relatively high for both sexes in all years, con- the two models, 18.3, is direct and fairly strong evidence in sistent with expectations based on life history (Figure 4). support of additive structure on Φ. All other models, includ- Ultimately, limitations to inference about survival from ing the global model [model 9; Φ(sex*time), p(sex*time)], the data set with only physical recaptures could be traced had essentially no support (∆QAICc > 10, evidence ratios back to the “big law” (Figure 1). Recaptures were sparse and compared to the top model > 150). estimated encounter probabilities were consistently very low, Estimates of Φ from the top model were imprecise, par- never increasing to more than 0.2 (Figure 5). Consistent with ticularly for males, and the point estimates for both sexes our expectations, encounter probabilities for male suckers in 2000 and 2001 were on a boundary (1.0; Figure 4). The were more than double the encounter probabilities for female boundary estimates indicate that the data provided insuf- suckers; female encounter probabilities never exceeded 0.05. ficient information to the likelihood in those years. In the Strong decreases in encounter probabilities for both sexes in second best model with full structure on Φ, the 2000 and 2006 were a result of temporarily reduced sampling effort.

Table 2. Model selection results for the set Model Model of 10 Cormack-Jolly-Seber models fit to the Model k QAICc ∆QAICc probability -2log L number e data set including only physical recaptures (w ) of PIT-tagged Lost River suckers. The i number of parameters estimated in each 1 Φ(sex+time),p(sex+time) 19 16,571.3 0.0 0.935 19,327.3 model is indicated by k, and the estimate 2 Φ(sex*time),p(sex+time) 27 16,577.1 5.8 0.051 19,315.3 of the overdispersion parameter ( c ^ ) used in 3 Φ(sex),p(sex*time) 20 16,581.8 10.5 0.005 19,337.3 calculating adjusted model selection criteria (QAICc) was 1.17. 4 Φ(sex+time),p(sex*time) 27 16,582.1 10.8 0.004 19,321.1 5 Φ(.),p(sex*time) 19 16,582.6 11.3 0.003 19,340.5 6 Φ(time),p(sex*time) 26 16,586.7 15.4 0.000 19,328.9 7 Φ(time),p(sex+time) 18 16,587.6 16.3 0.000 19,348.7 8 Φ(sex),p(sex+time) 12 16,587.7 16.4 0.000 19,362.9 9 Φ(sex*time),p(sex*time) 34 16,588.5 17.2 0.000 19,312.2 10 Φ(.),p(sex+time) 11 16,600.0 28.7 0.000 19,379.6

Figure 4. Estimates of survival probability ( ± 95% confidence intervals) Figure 5. Model-averaged estimates of encounter probability ( ± 95% from the top model [Φ(sex+time),p(sex+time)] in the set of models fit confidence intervals) from the set of models fit to the data set including to the data set including only physical recaptures of PIT-tagged Lost only physical recaptures of PIT-tagged Lost River suckers. The estimate of River suckers. Estimates for both sexes in 2000 and 2001 are not shown the overdispersion parameter ( cˆ ) used in calculating adjusted standard because they were on a boundary (1.0), indicating estimability problems. errors was 1.17. Estimates were not model-averaged because of problems with boundary estimates, particularly in the second best model, but the top model received nearly all of the support (model probability = 0.935).

1.0 0.20 Female ●

● ● Male ● ● 0.8 ● ● Male 0.15 ● ● 0.6 ● ● 0.10 ● ● 0.4

● 0.05 0.2 Survival probability

Encounter probability Female 0.0 0.00 1999 2000 2001 2002 2003 2004 2005 2006 2000 2001 2002 2003 2004 2005 2006 2007

224 Fisheries • v o l 35 n o 5 • m a y 2010 • w w w .f i s h e r i e s .o r g Analysis including all encounters strongly favored the inclusion of many effects on p. The top six models all included separate estimates of p for each sex in Model selection results and parameter estimates for the each year and at least one effect for tag type in 2006–2008. data set that included both physical recaptures and remote None of the models without some form of tag type effect had detections supported much stronger inferences about the any support. dynamics of annual survival of PIT-tagged Lost River suck- The additional encounters from the remote antennas pro- ers. Similar to the model selection results for the data set vided substantial improvements in precision for the estimates that included only physical recaptures, the top two models of survival probabilities. Model-averaged estimates of Φ were in the set included additive effects of sex and time on Φ highly precise for the expanded data set (Figure 6), with coef- (Table 3). However, evidence ratios comparing models with ficients of variation never more than 4%. Estimates were so additive versus full structure on Φ with the same structure precise in the most recent years that survival probabilities on p were much more supportive of the full structure on Φ were essentially known for the tagged set of fish. In contrast (model 1 vs. model 3 = 7.0; model 2 vs. model 4 = 2.5). As to the data set including only physical recaptures, survival expected based on the increases in re-encounters provided probabilities for both sexes in 2000 were estimable, and both by the remote antennas during 2006–2008, model selection point estimates were near 1.0. However, point estimates for

Table 3. Model selection results for the top 10 Cormack-Jolly-Seber models fit to the data set including both physical recaptures and remote detections of PIT-tagged Lost River suckers. Twenty-five other reduced models were considered, but all had ∆QAICc > 33 and are not shown. The “tagtype” effect on p in the model names refers to the difference between 125.0 kHz and 134.2 kHz tags, which is only included for the years 2006, 2007, and 2008 (see text). Model Model The tagtype effect is either Model k QAICc ∆QAICc probability -2log L number e constrained to be the same (w ) across years (“tagtype” i alone) or allowed to vary by 1 Φ(sex+time),p(sex*time+[tagtype*time]) 30 22,695.4 0.0 0.606 30,557.7 year (“tagtype*time”). Both 2 Φ(sex+time),p(sex*time+tagtype) 28 22,697.8 2.4 0.179 30,566.4 structures were combined additively (+ precedes 3 Φ(sex*time),p(sex*time+[tagtype*time]) 37 22,699.3 3.9 0.088 30,543.9 tagtype) and interactively 4 Φ(sex*time),p(sex*time+tagtype) 35 22,699.6 4.2 0.076 30,549.7 (* precedes tagtype) with the other effects in the 5 Φ(sex+time),p(sex*time*tagtype) 33 22,700.6 5.2 0.044 30,556.6 models. The number of parameters estimated in 6 Φ(sex*time),p(sex*time*tagtype) 40 22,704.6 9.2 0.006 30,543.0 each model is indicated by 7 Φ(sex+time),p(sex+time+[tagtype*time]) 22 22,709.0 13.6 0.001 30,597.6 k, and the estimate of the overdispersion parameter 8 Φ(sex+time),p(sex+time+tagtype) 20 22,709.8 14.4 0.000 30,604.2 ( cˆ ) used in calculating 9 Φ(sex*time),p(sex+time+[tagtype*time]) 29 22,712.6 17.2 0.000 30,583.5 adjusted model selection criteria (QAICc) was 1.35. 10 Φ(sex*time),p(sex+time+tagtype) 28 22,715.4 20.0 0.000 30,590.1

Figure 6. Model-averaged estimates of survival probability ( ± 95% Figure 7. Model-averaged estimates of encounter probability (± 95% confidence intervals) from the set of models fit to the data set including confidence intervals) from the set of models fit to the data set both physical recaptures and remote detections of PIT-tagged Lost River including both physical recaptures and remote detections of PIT-tagged suckers. Estimates for both sexes in 2001 are not shown because they Lost River suckers. The estimate of the overdispersion parameter were on a boundary (1.0), indicating estimability problems. The estimate ( cˆ ) used in calculating adjusted standard errors was 1.35. In 2006 and of the overdispersion parameter ( cˆ ) used in calculating adjusted standard 2007, confidence intervals are only plotted for females with 125.0 kHz errors was 1.35. tags; for other estimates the confidence intervals are too narrow to ^ distinguish them from the estimates [se(p) <– 0.006]. Note change in y-axis scale compared to Figure 5.

1.0 ● 1.0 ● ● ● ● ● ● ● ● Male 125 kHz tags 0.8 ● 0.8 Male 134 kHz tags ● Female 125 kHz tags 0.6 0.6 Female 134 kHz tags ● Male ● Female 0.4 0.4

0.2 0.2 ● ● ● ●

Survival probability ● ● Encounter probability 0.0 0.0 2000 2001 2002 2003 2004 2005 2006 2007 1999 2000 2001 2002 2003 2004 2005 2006

Fisheries • v o l 35 n o 5 • m a y 2010 • w w w .f i s h e r i e s .o r g 225 both sexes were again on a boundary (1.0) in 2001. Survival ous factors potentially affecting the population dynamics and of PIT-tagged Lost River suckers in 2001 was either very recovery of this species. nearly 100% or there was simply not enough information in In 2006, 2007, and 2008, remote detections of PIT-tagged the data about individuals at large during 2001; either expla- suckers that had never been previously re-encountered con- nation is plausible. verted thousands of encounter histories from relatively unin- Compared to the estimates from the data set including formative strings of nondetections into information about only physical recaptures (Figure 4), point estimates of Φ based encounter probabilities because uncertainty about survival on the expanded data set changed by nontrivial amounts of those fish in earlier years was eliminated. As a result, the extending all the way back to 1999. Except for 2003, esti- data including the remote detections were highly informative mates for males increased by 1–11% when estimated with the for determining whether fish were alive and went undetected expanded data set. Of particular note are the estimates for the or had died. Estimates of both survival and encounter prob- three most recent years, which were all apparently underesti- abilities in the most recent years were so precise as to be mated by more than 7% using the data set with only physical essentially known. recaptures. In contrast, Φ estimates for females decreased in Janney et al. (2008) showed that annual survival of adult each year by 1–9%, and the data set including only physi- suckers in Upper Klamath Lake was reduced in years when cal recaptures produced the largest apparent overestimations conspicuous fish die-offs occurred, but was also low in years in 2002 and 2003. Estimates from both model sets indicated without conspicuous fish die-offs. Those results add to ongo- that female survival was greater than male survival in each ing concerns about the chronic effects of poor water quality year, in contrast to our expectations based on reproductive and other sources of mortality on the recovery of endangered investment. Between 2002 and 2006, female survival ranged sucker populations (NRC 2004). The data provided by the from 85 to 94% and male survival ranged from 79 to 92%. remote detection systems has made it possible for us to pur- The increased precision of the survival probabilities sue modeling that investigates hypotheses about the effects for the expanded data set derives almost entirely from the of various factors on survival. Results of such analyses are of pressing concern to management agencies tasked with con- increased encounter probabilities in 2006 and 2007 (Figure serving and recovering these species. 7). Although precision was better for encounter probabilities in all years using the expanded data set, point estimates were Remote detections and the “big law” of capture-recapture mostly unchanged through 2005, as expected. In contrast, encounter probabilities in 2006 and 2007 increased from Miranda and Bettoli (2007:251) noted that “tagging is 2–15% based on the data set with only physical recaptures to not extensively used to assess mortality of fish populations, over 90% after including the remote detections. Encounter mostly due to cost and practical difficulties.” Results from our probabilities for fish with 134.2 kHz tags were 98% or greater, PIT-tagging program have shown that it is possible to over- whereas encounter probabilities for fish with 125.0 kHz tags come one critical difficulty represented by the “big law”—low ranged from 91% to 96%. These results were consistent encounter probabilities—by use of remote detection systems. with our expectations that encounter probabilities would be Capture probabilities are often very low for traditional fisher- higher for fish with the higher frequency tags. Put simply, ies gears, so the problem of low encounter probabilities will when remote detections were included, tagged fish at large remain a hindrance to fishery-independent capture-recapture in the population were almost certain to be detected at some studies that rely on traditional gears. Capture probabilities point during the spawning season. can be increased by more intensive sampling or alternative sampling strategies, but costs can become prohibitive and Discussion some active gears may not be permissible for some species (e.g., electrofishing). Summary of findings In an important sense, the encounter probability “problem” is also an advantage for fisheries CR studies, in that Remote detections of PIT-tagged Lost River suckers have it requires investigators to confront the realities of sampling improved the utility of capture-recapture methods for mak- in the system. Capture-recapture studies will only provide ing inferences about population dynamics for this endan- robust inferences when the data provide substantial infor- gered species. Despite intensive sampling at relatively small, mation about the animals under study, and low encounter discrete spawning areas, probabilities of recapturing tagged probabilities and imprecise estimates make it clear that much suckers in trammel nets were low. Thus, in the absence of remains unknown. Although this issue is raised most com- remote detections, the vast majority of individual encounter monly for CR studies, it applies equally to any method that histories for tagged fish contained little information about depends on capture data. Inferences based on samples that survival because those fish were never re-encountered. The represent only a small percentage of the individuals in the large proportion of uninformative encounter histories led to population cannot be considered strong inferences. model selection results that favored simple structure on sur- Our results with remote detection systems for PIT tags vival parameters. In addition, when compared to estimates should be encouraging to fisheries scientists considering CR based on all encounters, survival estimates were imprecise and methods and wrestling with the difficulties of low encounter apparently over- or underestimated by nontrivial amounts in probabilities. Although CR methods are often considered to a number of years. As a result, our capture-recapture data be too expensive and too difficult to implement in fisher- were not adequate to address important concerns about vari- ies studies, remote detection systems can shift the balance

226 Fisheries • v o l 35 n o 5 • m a y 2010 • w w w .f i s h e r i e s .o r g in their favor. Advances in technology have made it possible to meet or assess in order to avoid biased estimates (e.g., to construct antennas or arrays of antennas that can span constancy of recruitment, survival, catchability, or growth). entire rivers, up to 50 meters wide or wider. Antenna sys- Indeed, many of the assumptions are known to be violated tems are now primarily restricted only by flow conditions in a priori, but are commonly ignored. Improvements to tradi- the study system (high flows or debris can damage antennas), tional methods, such as year-class curves (Cotter et al. 2007) and antennas often require only infrequent maintenance. In and nonequilibrium length-based estimators (Gedamke and addition, the range at which tags can be read by the anten- Hoenig 2006), can improve the accuracy and precision of nas continues to increase. Finally, data can be acquired and survival estimates, but still require strong assumptions that transmitted in a fully electronic and remote process, reducing are difficult to meet or assess. Furthermore, because studies or eliminating some sources of error. of population dynamics are concerned with understanding The initial investment in establishing a capture-recapture variation in survival and the factors that influence it, the program based on PIT tags and remote detections may be high, utility of many of the traditional methods is limited because but costs decline substantially in subsequent years. The total they assume that survival is constant during the study. equipment costs for full remote antenna implementation at In general, CR methods are designed to estimate many our sites in 2006 was around $54,000 USD, but maintenance of the parameters that are “assumed away” in the traditional and supplemental costs in subsequent years have typically methods. For example, rather than make strong and untest- been less than $1,000. Under budget constraints, PIT-tagging able assumptions about the time-invariance of survival and efforts and remote detection systems can be developed incre- encounter (recapture) probabilities to estimate a single sur- mentally by adding antennas and readers as funding allows. vival rate, CR methods are designed to jointly estimate both Encounter probabilities are estimated as part of the analysis, parameters, and assumptions and temporal dynamics can be so these changes in sampling design can be accommodated in evaluated. As noted by D. G. Chapman (quoted in Cormack models and the precision of survival estimates will progres- 1968:456): sively improve. In most cases, once remote detection systems If far reaching assumptions are made, then strong con- are in place and encounter probabilities increase, the number clusions are reached. But if these assumptions are not of fish that need to be tagged and released each year can accepted then the whole structure built upon sand be reduced, potentially reducing costs associated with field collapses. efforts. In addition, some equipment, including antennas, can be designed and assembled independently to save costs in Modern methods of CR analysis have been developed to comparison to purchasing proprietary equipment from com- accommodate the realities of fisheries and wildlife field stud- mercial manufacturers. ies, including heterogeneity in capture probabilities and fail- ures of the closure assumption (Pollock 1982, 1991; Kendall Capture-recapture as an essential tool in fisheries et al. 1995; Williams et al. 2002; Pledger et al. 2003; Cowen and Schwarz 2006). These methods should be integrated into Freshwater fisheries scientists have been slow to appreci- the standard toolbox of freshwater fisheries scientists. ate the utility and advantages of capture-recapture methods for estimating survival and other demographic parameters, Remaining hurdles and also to adopt modern methods for analyzing CR data. On this count, we reiterate the advice of Pine et al. (2003) that The advantages of remote detection systems to inference CR methods have much to offer fisheries research and man- and estimation are numerous, but there are also limitations. agement. In turn, we concur with the recent synthesis con- Two of the thorniest issues in capture-recapture analyses are cerning mortality estimation by Miranda and Bettoli (2007) goodness-of-fit assessment and heterogeneity in encounter that freshwater fisheries scientists should incorporate into probabilities. Although the primary parameters of interest their toolbox numerous methods that appear to be largely in CJS and related models (e.g., survival probabilities) are restricted to marine and coastal fisheries. We believe that robust to heterogeneity in encounter probabilities (Carothers modern CR methods are one such tool. The use of tagging 1979; Pollock et al. 1990), model selection and inferential studies and tag return data is increasing in analyses of marine procedures become suspect when too much heterogeneity and coastal fisheries. The challenges of implementing tag- exists in the data. Such heterogeneity must be accounted for ging studies in these large and open systems are greater than either by inflating variance estimates using a correction factor those for most freshwater systems (with the possible excep- ( cˆ ) or by adding additional justifiable structure to the model. tion of some commercially exploited stocks), so the ground- Given the nonrandom patterns of association among indi- work has been laid for implementing modern CR methods in vidual fish in many populations, heterogeneity in encounter freshwater fisheries. probabilities is likely to be worse than in many wildlife stud- Capture-recapture methods offer a number of advantages ies (Pollock 1991; Seber and Schwarz 2002). No hard and over traditional methods for estimating survival based on fast rules exist, but cˆ estimates above about 3.0 are suspicious fishery-dependent catch and effort data or fisheries-indepen- and probably indicate that model structure is inappropriate dent survey data (e.g., catch curves, change-in-ratio meth- (Choquet et al. 2009). ods, and length- or age-based methods; Ricker 1975; Hilborn In our study, heterogeneity in encounter probabilities for and Walters 1992; Quinn and Deriso 1999). Most traditional Lost River suckers increased after the remote detection sys- methods provide estimates of survival that are notoriously tems were put in place, and even more so after the change in imprecise and require strong assumptions that are difficult the type of PIT tag (from 125.0 kHz to 134.2 kHz). Ironically,

Fisheries • v o l 35 n o 5 • m a y 2010 • w w w .f i s h e r i e s .o r g 227 because encounter probabilities were so high, goodness-of-fit detection systems located at constriction points in migration, tests were very sensitive to small deviations from expected such as dam bypasses (see the PIT Tag Information System values. To address heterogeneity, we had to adjust model [PTAGIS]; www.ptagis.org). structures to account for the two different tag types and still Tag return data from commercial fisheries can also be had to account for overdispersion because some heteroge- enhanced by using PIT tags and remote detection systems neity remained ( cˆ for data set with all encounters = 1.35). placed at bottlenecks during harvest and processing. By plac- Encounter probabilities are typically not of primary interest ing remote detection systems on selected vessels and com- (“nuisance” parameters), so effort should be directed at reduc- bining PIT tags with conventional T-bar tags, researchers ing heterogeneity to avoid using information in the data to were able to precisely estimate tag reporting rates, exploita- estimate additional encounter probabilities. tion rates, and natural mortality for southern rock lobsters An additional concern with remote detection systems is (Jasus edwardsii) in Australia (Burch et al. 2009; Frusher et that they only detect the marked portion of the population. al. 2009). For fisheries with larger bottlenecks at processing Information on the unmarked portion of the population is facilities, antennas or arrays of antennas could be incorpo- needed to estimate population size, recruitment, and popu- rated into the processing line. For example, antennas could lation growth rate. Also, for models addressing hypotheses be constructed to fit the brailers at processing facilities for about individual covariates (e.g., length, weight, condi- Bristol Bay-Bering Sea crab fisheries to detect tagged crabs tion), individual fish must be captured for the covariate to as they are offloaded from the boats. Bottlenecks in process- be measured. Although modeling strategies are available to accommodate remote detections in some situations (e.g., ing occur in many other fisheries, such as whitefish stocks Al-Chokhachy and Budy 2008), careful attention must be in the Great Lakes and shrimp stocks in the Gulf of Mexico paid to study design and substantial effort will need to be and the Pacific Ocean, and we suspect that remote detection directed at capture sampling when these additional param- systems could be integrated into these facilities. The benefits eters are of primary interest. In addition, new modeling strat- to inference and estimation should be worth the initial costs egies will be needed to accommodate some study situations. and public relations work that would be required. Finally, encounter data from remote detection systems can Future possibilities benefit analyses that integrate multiple sources of information through a joint likelihood approach. For example, tag The technology of PIT tags has made it possible for returns from fishermen can be combined with telemetry researchers to routinely pursue biological questions that were studies (Pollock et al. 2004) as well as various sources of nearly impossible to answer prior to the 1980s (Gibbons and fishery-dependent data (Coggins et al. 2006; Eveson et al. Andrews 2004). Technological advances have continued to 2007, 2009; Conn et al. 2009). An integrated model that reduce the limitations of PIT tags for biological studies, and incorporates information from tag re-encounters can often remote detection systems have further enhanced their util- disentangle parameters of interest that are otherwise not ity (Prentice et al. 1990; Zydlewski et al. 2006). The data estimable, such as tag reporting rate. The future of assess- provided by remote detection systems opens up numerous ments for large commercial fisheries appears to lie in such possibilities for modeling and inference, allowing researchers integrated modeling (Maunder 2003), but other population to ask questions that address more complex hypotheses. For dynamics investigations can benefit from such an approach example, data from studies that previously were aimed at sim- as well. Remote detection systems for PIT tags offer potential ply trying to obtain one estimate of survival can now be used for dramatically increasing encounter probabilities and the to address hypotheses about factors affecting the dynamics of density of data provided by tagged individuals, which would survival. The development of models to pursue such infer- benefit a wide range of fisheries studies. ences has been the focus of much of the theoretical work in capture-recapture over the last decade (e.g., Thomson et al. Acknowledgments 2009). Analyses will be more complex and challenging, but also more interesting from a biological standpoint. Earl Prentice provided valuable advice and assistance Encounter data from remote detection systems should also related to implementation of the remote detection systems. allow capture-recapture methods to be used in systems where Alta Scott and Greta Blackwood oversaw data quality con- they were previously considered impracticable, from studies of rare or elusive species in small streams to assessments of trol and database management. Jeff Laake provided assistance fish populations in large systems. For rare or elusive fishes in with RMark code during model development, specification, small streams, physically recapturing tagged individuals with and analysis. David Anderson, Summer Burdick, Mike Quist, some reasonable probability may be impossible. However, a Robert Al-Chokhachy, and Doug Markle provided suggestions combination of surveys with portable detection systems (e.g., that improved the manuscript. The Klamath Basin office of The wands) and stationary antenna arrays operating continuously Nature Conservancy provided GIS data that was used to cor- can be sufficient (Berger and Gresswell 2009). In large sys- rect the area of the map around the mouth of the Williamson tems, such as Upper Klamath Lake, remote detection systems River to reflect recently restored wetland habitat. This work will need to target areas where individuals aggregate, such was funded in part by an interagency agreement with the U.S. as spawning areas or constriction points in migration routes. Bureau of Reclamation. The use of trade, product, or firm For example, millions of fish are PIT-tagged in the Columbia names in this publication is for descriptive purposes only and River Basin each year and monitored with dozens of remote does not imply endorsement by the U.S. Government.

228 Fisheries • v o l 35 n o 5 • m a y 2010 • w w w .f i s h e r i e s .o r g References Choquet, R., J.-D. Lebreton, O. Gimenez, A.-M. Reboulet, and R. Pradel. 2009. U-CARE: utilities for performing goodness of fit tests Adams, P. B. 1980. Life history patterns in marine fishes and their con- and manipulating CApture-REcapture data. Ecography 32:1071- sequences for fisheries management. Fishery Bulletin 78:1-12. 1074. Al-Chokhachy, R., and P. Budy. 2008. Demographic characteristics, Coggins, L. G., Jr., W. E. Pine III, C. J. Walters, and S. J. D. Martell. population structure, and vital rates of a fluvial population of bull 2006. Age-structured mark-recapture analysis: a virtual-population- trout in Oregon. Transactions of the American Fisheries Society analysis-based model for analyzing age-structured capture-recapture 137:1709-1722. data. North American Journal of Fisheries Management 26:201- Anderson, D. R. 2008. Model based inference in the life sciences. 205. Springer, New York. Conn, P. B., G. C. White, and J. L. Laake. 2009. Simulation perfor- Anderson, D. R., W. A. Link, D. H. Johnson, and K. P. Burnham. mance of Bayesian estimators of abundance employing age-at-har- 2001. Suggestions for presenting the results of data analyses. Journal vest and mark-recovery data. Pages 965-986 in D. L. Thomson, E. of Wildlife Management 65:373-378. G. Cooch, and M. J. Conroy, eds. Modeling demographic processes Bailey, L. L., W. L. Kendall, and D. R. Church. 2009. Exploring in marked populations. Springer, New York. extensions to multistate models with multiple unobservable states. Conroy, M. J. 2009. Application of capture-recapture to addressing Pages 693-709 in D. L. Thomson, E. G. Cooch, and M. J. Conroy, questions in evolutionary ecology. Pages 131-156 in D. L. Thomson, eds. Modeling demographic processes in marked populations. E. G. Cooch, and M. J. Conroy, eds. Modeling demographic pro- Springer, New York. cesses in marked populations. Springer, New York. Banish, N. P., B. J. Adams, R. S. Shively, M. M. Mazur, D. A. Cormack, R. M. 1968. The statistics of capture-recapture methods. Beauchamp, and T. M. Wood. 2009. Distribution and habitat Oceanography and Marine Biology, An Annual Review 6:455- associations of radio-tagged adult Lost River suckers and short- 506. nose suckers in Upper Klamath Lake, Oregon. Transactions of the Cotter, A. J. R., B. Mesnil, and G. J. Piet. 2007. Estimating stock American Fisheries Society 138:153-168. parameters from trawl cpue-at-age series using year-class curves. Berger, A. M., and R. E. Gresswell. 2009. Factors influencing coastal ICES Journal of Marine Science 64:234-247. cutthroat trout (Oncorhynchus clarkii clarkii) seasonal survival rates: Cowen, L., and C. J. Schwarz. 2006. The Jolly-Seber model with tag a spatially continuous approach within stream networks. Canadian loss. Biometrics 62:699-705. Journal of Fisheries and Aquatic Sciences 66:613-632. Doherty, P. F., Jr., E. A. Schreiber, J. D. Nichols, J. E. Hines, W. A. Bonner, S. J., and C. J. Schwarz. 2004. Continuous time-dependent Link, G. A. Schenk, and R. W. Schreiber. 2004. Testing life history individual covariates and the Cormack-Jolly-Seber model. Animal predictions in a long-lived seabird: a population matrix approach Biodiversity and Conservation 27.1:149-155. with improved parameter estimation. Oikos 105:606-618. Bradbury, J. P., S. M. Colman, and R. L. Reynolds. 2004. The his- Eveson, J. P., T. Polacheck, and G. M. Laslett. 2007. Incorporating tory of recent limnological changes and human impact on Upper fishery observer data into an integrated catch-at-age and multiyear Klamath Lake, Oregon. Journal of Paleolimnology 31:151-165. tagging model for estimating mortality rates and abundance. Fishery Buckland, S. T., K. P. Burnham, and N. H. Augustin. 1997. Model Bulletin 105:493-508. selection: an integral part of inference. Biometrics 53:603-618. Eveson, J. P., G. M. Laslett, and T. Polacheck. 2009. A spatial model Burch, P., S. D. Frusher, S. Wotherspoon, and T. Polacheck. 2009. for estimating mortality rates, abundance and movement prob- A modeled cost-benefit analysis of hybrid PIT and conventional abilities from fishery tag-recovery data. Pages 987-1010 in D. L. tagging scenarios. New Zealand Journal of Marine and Freshwater Thomson, E. G. Cooch, and M. J. Conroy, eds. Modeling demo- Research 43:339-346. graphic processes in marked populations. Springer, New York. Burnham, K. P., and D. R. Anderson. 2002. Model selection and Franklin, A. B. 2001. Exploring ecological relationships in survival and multimodel inference: a practical information-theoretic approach. estimating rates of population change using Program MARK. Pages Second edition. Springer, New York. 350-356 in R. Field, R. J. Warren, H. Okarma, and P. R. Sievert, Burnham, K. P., D. R. Anderson, G. C. White, C. Brownie, and eds. Proceedings of the Second International Wildlife Management K. H. Pollock. 1987. Design and analysis methods for fish survival Congress. The Wildlife Society, Bethesda, Maryland. experiments based on release-recapture. American Fisheries Society Frusher, S. D., D. Hall, P. Burch, and C. Gardner. 2009. Combining Monograph 5, Bethesda, Maryland. passive integrated transponder tags with conventional T-bar tags Cam, E. 2009. Contribution of capture-mark-recapture modeling to to improve tag reporting rates in a rock lobster trap fishery. New studies of evolution by natural selection. Pages 83-129 in D. L. Zealand Journal of Marine and Freshwater Research 43:347-353. Thomson, E. G. Cooch, and M. J. Conroy, eds. Modeling demo- Fujiwara, M., and H. Caswell. 2001. Demography of the North graphic processes in marked populations. Springer, New York. Atlantic right whale. Nature 414:537-541. Carothers, A. D. 1973. The effects of unequal catchability on Jolly- Gedamke, T., and J. M. Hoenig. 2006. Estimating mortality from Seber estimates. Biometrics 29:79-100. mean length data in nonequilibrium situations, with application to _____. 1979. Quantifying unequal catchability and its effect on sur- the assessment of goosefish. Transactions of the American Fisheries vival estimates in an actual population. Journal of Animal Ecology Society 135:476-487. 48:863-869. Gibbons, J. W., and K. M. Andrews. 2004. PIT tagging: simple tech- Chatfield, C. 1995. Model uncertainty, data mining and statisti- nology at its best. BioScience 54:447-454. cal inference. Journal of the Royal Statistical Society, Series A Gilbert, R. O. 1973. Approximations of the bias in the Jolly-Seber 158(Part 3):419-466. capture-recapture model. Biometrics 29:501-526. Choquet, R., A.-M. Reboulet, R. Pradel, O. Gimenez, and J.-D. Guy, C. S., H. L. Blankenship, and L. A. Nielsen. 1996. Tagging Lebreton. 2004. M-SURGE: new software specifically designed and marking. Pages 353-383 in B. R. Murphy and D. W. Willis, eds. for multistate capture-recapture models. Animal Biodiversity and Fisheries techniques, second edition. American Fisheries Society, Conservation 27.1:207-215. Bethesda, Maryland.

Fisheries • v o l 35 n o 5 • m a y 2010 • w w w .f i s h e r i e s .o r g 229 Hayes, D. B., J. R. Bence, T. J. Kwak, and B. E. Thompson. 2007. Nichols, J. D., and J. E. Hines. 2002. Approaches for the direct esti- Abundance, biomass, and production. Pages 327-374 in C. S. Guy mation of λ, and demographic contributions to λ, using capture- and M. L. Brown, eds. Analysis and interpretation of freshwater recapture data. Journal of Applied Statistics 29:539-568. fisheries data. American Fisheries Society, Bethesda, Maryland. Nichols, J. D., J. E. Hines, K. H. Pollock, R. L. Hinz, and W. A. Hilborn, R., and C. J. Walters. 1992. Quantitative fisheries stock Link. 1994. Estimating breeding proportions and testing hypothe- assessment: choice, dynamics and uncertainty. Chapman and Hall, ses about costs of reproduction with capture-recapture data. Ecology New York. 75:2052-2065. Janney, E. C., R. S. Shively, B. S. Hayes, and P. M. Barry. 2008. Nichols, J. D., J. E. Hines, J.-D. Lebreton, and R. Pradel. 2000. Demographic analysis of Lost River sucker and shortnose sucker Estimation of contributions to population growth: a reverse-time populations in Upper Klamath Lake, Oregon. Transactions of the capture-recapture approach. Ecology 81:3362-3376. American Fisheries Society 137:1812-1825. Nielsen, L. A. 1992. Methods of marking fish and shellfish. American Jiang, H., K. H. Pollock, C. Brownie, J. M. Hoenig, R. J. Latour, B. Fisheries Society Special Publication 23, Bethesda, Maryland. K. Wells, and J. E. Hightower. 2007. Tag return models allowing Parker, N. C., A. E. Giorgi, R. C. Heidinger, D. B. Jester Jr., E. for harvest and catch and release: evidence of environmental and D. Prince, and G. A. Winans (editors). 1990. Fish-marking management impacts on striped bass fishing and natural mortality techniques. American Fisheries Society Symposium 7, Bethesda, rates. North American Journal of Fisheries Management 27:387- Maryland. 396. Perkins, D. L., J. Kann, and G. G. Scoppettone. 2000. The role of Kann, J., and E. B. Welch. 2005. Wind control on water quality in poor water quality and fish kills in the decline of endangered Lost shallow, hypereutrophic Upper Klamath Lake, Oregon. Lake and River and shortnose suckers in Upper Klamath Lake. Technical Reservoir Management 21:149-158. report, U.S. Geological Survey Biological Resources Division to Kendall, W. L. 2009. One size does not fit all: adapting mark-recapture U.S. Bureau of Reclamation, Contract 4-AA-29-12160. and occupancy models for state uncertainty. Pages 765-780 in D. L. Pfister, C. A. 1998. Patterns of variance in stage-structured populations: Thomson, E. G. Cooch, and M. J. Conroy, eds. Modeling demo- evolutionary predictions and ecological implications. Proceedings graphic processes in marked populations. Springer, New York. of the National Academy of Sciences of the USA 95:213-218. Kendall, W. L., K. H. Pollock, and C. Brownie. 1995. A likelihood- Pine, W. E., K. H. Pollock, J. E. Hightower, T. J. Kwak, and J. A. based approach to capture-recapture estimation of demographic Rice. 2003. A review of tagging methods for estimating fish popula- parameters under the robust design. Biometrics 51:293-308. tion size and components of mortality. Fisheries 28(10):10-23. Laake, J. 2009. RMark: R code for MARK analysis. R package version Pledger, S., K. H. Pollock, and J. L. Norris. 2003. Open capture- 1.9.5. Available at: www.phidot.org/software/mark/rmark. recapture models with heterogeneity: I. Cormack-Jolly-Seber Laake, J., and E. Rexstad. 2009. RMark—an alternative approach to model. Biometrics 59:786-794. building linear models in MARK. Appendix C in E. Cooch and G. Pollock, K. H. 1982. A capture-recapture design robust to unequal White, eds. Program MARK: a gentle introduction. Available at: probability of capture. Journal of Wildlife Management 46:752- www.phidot.org/software/mark/docs/book. 757. Lebreton, J. D., and R. Pradel. 2002. Multistate recapture models: _____. 1991. Modeling capture, recapture, and removal statistics for modelling incomplete individual histories. Journal of Applied estimation of demographic parameters for fish and wildlife popula- Statistics 29:353-369. tions: past, present, and future. Journal of the American Statistical Lebreton, J.-D., K. P. Burnham, J. Clobert, and D. R. Anderson. Association 86:225-238. 1992. Modeling survival and testing biological hypotheses using Pollock, K. H., and R. Alpizar-Jara. 2005. Classical open-population marked animals: a unified approach with case studies. Ecological capture-recapture models. Pages 36-57 in S. C. Amstrup, T. L. Monographs 62:67-118. McDonald, and B. F. J. Manly, eds. Handbook of capture-recapture Lindenberg, M. K., G. Hoilman, and T. M. Wood. 2009. Water qual- analysis. Princeton University Press. ity conditions in Upper Klamath and Agency Lakes, Oregon, 2006. Pollock, K. H., J. D. Nichols, C. Brownie, and J. E. Hines. 1990. U.S. Geological Survey Scientific Investigations Report 2008- Statistical inference for capture-recapture experiments. Wildlife 5201. Monographs 107. Maunder, M. N. 2003. Paradigm shifts in fisheries stock assessment: Pollock, K. H., H. Jiang, and J. E. Hightower. 2004. Combining from integrated analysis to Bayesian analysis and back again. telemetry and fisheries tagging models to estimate fishing and natu- Natural Resource Modeling 16:465-475. ral mortality rates. Transactions of the American Fisheries Society Miller, R. R., and G. R. Smith. 1981. Distribution and evolution 133:639-648. of Chasmistes (Pisces:Catostomidae) in western North America. Pradel, R. 1996. Utilization of capture-mark-recapture for the study of Occasional Papers of the Museum of Zoology, University of recruitment and population growth rate. Biometrics 52:703-709. Michigan 696:1-46. Prentice, E. F., T. A. Flagg, C. S. McCutcheon, and D. F. Brastow. Miranda, L. E., and P. W. Bettoli. 2007. Mortality. Pages 229-277 1990. PIT-tag monitoring systems for hydroelectric dams and fish in C. S. Guy and M. L. Brown, eds. Analysis and interpretation hatcheries. American Fisheries Society Symposium 7:323-334. of freshwater fisheries data. American Fisheries Society, Bethesda, Quinn, T. J., II, and R. B. Deriso. 1999. Quantitative fish dynamics. Maryland. Oxford University Press, New York. NRC (National Research Council). 2004. Endangered and threatened R Development Core Team. 2009. R: a language and environment fishes in the Klamath River Basin: causes of decline and strategies for statistical computing. R Foundation for Statistical Computing, for recovery. The National Academies Press, Washington, DC. Vienna, Austria. Available at: www.R-project.org. Nichols, J. D. 2005. Modern open-population capture-recapture mod- Ricker, W. E. 1975. Computation and interpretation of biological sta- els. Pages 88-122 in S. C. Amstrup, T. L. McDonald, and B. F. J. tistics of fish populations. Bulletin of the Fisheries Research Board Manly, eds. Handbook of capture-recapture analysis. Princeton of Canada 191. Department of the Environment, Fisheries and University Press, Princeton, New Jersey. Marine Service, Ottawa.

230 Fisheries • v o l 35 n o 5 • m a y 2010 • w w w .f i s h e r i e s .o r g Rotella, J. 2009. Estimating reproductive costs with multi-state mark- Seber, G. A. F. 1982. The estimation of animal abundance and related recapture models, multiple observable states, and temporary emi- parameters. Second edition. Macmillan, New York. gration. Pages 157-172 in D. L. Thomson, E. G. Cooch, and M. J. Seber, G. A. F., and C. J. Schwarz. 2002. Capture-recapture: before Conroy, eds. Modeling demographic processes in marked popula- and after EURING 2000. Journal of Applied Statistics 29:5-18. tions. Springer, New York. Senar, J. C., A. A. Dhondt, and M. J. Conroy. 2004. The quantita- Royall, R. M. 1997. Statistical evidence: a likelihood paradigm. tive study of marked individuals in ecology, evolution and conser- Chapman and Hall, New York. Schmutz, J. A. 2009. Stochastic variation in avian survival rates: vation biology: a foreword to the Euring 2003 conference. Animal life-history predictions, population consequences, and the poten- Biodiversity and Conservation 27:1-2. tial responses to human perturbations and climate change. Pages Thomson, D. L., E. G. Cooch, and M. J. Conroy (editors). 2009. 441-461 in D. L. Thomson, E. G. Cooch, and M. J. Conroy, eds. Modeling demographic processes in marked populations. Springer, Modeling demographic processes in marked populations. Springer, New York. New York. USFWS (U.S. Fish and Wildlife Service). 1988. Determination of Schwarz, C. J. 2009. Migration and movement—the next stage. Pages endangered status for the shortnose sucker and Lost River sucker. 323-347 in D. L. Thomson, E. G. Cooch, and M. J. Conroy, eds. Federal Register 53:27130-27134. Modeling demographic processes in marked populations. Springer, New York. _____. 1993. Lost River (Deltistes luxatus) and shortnose (Chasmistes Schwarz, C. J., and A. N. Arnason. Use of tag-recovery information brevirostris) sucker recovery plan. Portland, Oregon. in migration and movement studies. American Fisheries Society White, G. C., and K. P. Burnham. 1999. Program MARK: survival Symposium 7:588-603. rate estimation from both live and dead encounters. Bird Study Schwarz, C. J., J. F. Schweigert, and A. N. Arnason. 1993. Estimating 46(Supplement):S120-S139. migration rates using tag-recovery data. Biometrics 49:177-193. White, G. C., W. L. Kendall, and R. J. Barker. 2006. Multistate sur- Scoppettone, G. G. 1988. Growth and longevity of cui-ui and longev- vival models and their extensions in Program MARK. Journal of ity of other catostomids and cyprinids in western North America. Wildlife Management 70:1521-1529. Transactions of the American Fisheries Society 117:301-307. Williams, B. K., J. D. Nichols, and M. J. Conroy. 2002. Analysis and Scoppettone, G. G., and G. Vinyard. 1991. Life history and management of four endangered lacustrine suckers. Pages 359-377 in W. L. management of animal populations. Academic Press, New York. Minckley and J. E. Deacon, eds. Battle against extinction: native Zydlewski, G. B., G. Horton, T. Dubreuil, B. Letcher, S. Casey, and fish management in the American West. University of Arizona J. Zydlewski. 2006. Remote monitoring of fish in small streams. Press, Tucson. Fisheries 31(10):492-502.

YSI

Fisheries • v o l 35 n o 5 • m a y 2010 • w w w .f i s h e r i e s .o r g 231 Feature: FISH HABITAT The Challenges of Tracking Habitat Restoration at Various Spatial Scales

ABSTRACT: Evaluating the ecological effectiveness of the hundreds of millions of Katie Barnas and dollars invested to recover Pacific Northwest salmon has a number of prerequisites— Stephen L. Katz principally, detailed knowledge of management actions. In the absence of coordinated ecological monitoring, basic restoration project metadata (type, location, timing, etc.) is the primary source of information guiding restoration planning and management Barnas is a research biologist at decisions. There are surprisingly few sources of habitat restoration information at NOAA’s Northwest Fisheries Science scales appropriate to salmon recovery, ranging from small sub-watersheds to multi- Center in Seattle, Washington, and state evolutionarily significant units. We evaluate the consistency of two Columbia may be contacted at Katie.Barnas@ River basin restoration inventories developed in different ways but similar in goal: to noaa.gov. Katz is currently the inform future project placement. We compared the Pacific Northwest Habitat Project research coordinator for the Channel Database, compiled by researchers at NOAA Fisheries Service, and the Subbasin Islands National Marine Sanctuary Inventories, compiled by local entities within each of 62 subbasins. Confederating and was previously the monitoring these two data holdings, we found 11,805 projects in the Columbia River Basin, with and evaluation coordinator for only 13% (1,549/11,805) in common. Subbasin-scale inventories capture more locally sponsored projects than do large-scale data compilations. Attributes (date, cost) of salmon recovery at the Northwest projects in common to both datasets matched less than half the time, suggesting Fisheries Science Center. organizations working on the same project are reporting independently and differently to either the database or inventories. We found that the lack of common data standards and lack of knowledge about accessible data sources are major hurdles to making data INTRODUCTION available for decision making. The magnitude and scale of efforts to manage habitat and revive salmon runs in Los retos de dar seguimiento a la restauración the Pacific Northwest are unprecedented in U.S. history and possibly world history de hábitat en varias escalas espaciales (Lichatowitch 1999; Ruckelshaus et al. 2003). The freshwater range of the five Resumen: La evaluación de la efectividad ecológica de los cientos de millones de dólares Endangered Species Act (ESA) listed que se han invertido en la recuperación del salmón del Pacífico noroeste, tiene varios Pacific salmon speciesOncorhynchus ( pre-requisitos –principalmente, un conocimiento detallado de las acciones de manejo. A tshawytscha, O. keta, O. kisutch, O. falta de un monitoreo ecológico coordinado, los meta-datos (tipo, locación, tiempo, etc.) nerka, O. mykiss) span the majority of derivados de proyectos básicos de restauración, son la principal fuente de información four states: Washington, Oregon, Idaho, para ordenar los planes de restauración y decisiones de manejo. Sorprendentemente and California, and historically parts of existen muy pocas fuentes de información sobre restauración de hábitat en escalas que two others (Montana and Nevada). The sean útiles para la recuperación del salmón; dichas escalas van desde las pequeñas sub- Columbia River Basin, the major river cuencas hasta unidades (multi-estado) evolutivas significativas. Se evaluó la consistencia basin in the Northwest and the second de dos inventarios de restauración de la cuenca del Río Columbia, desarrollados de forma largest in the U.S., was once home to the diferente pero con un objetivo común: informar sobre la asignación de proyectos futuros. most abundant salmon runs in the world, Se comparó la base de datos del proyecto “Hábitat del Pacífico Noroeste”, compilada but now many of these runs require exten- por investigadores del Servicio de Pesquerías de la NOAA, y los “Inventarios de Sub- cuenca”, compilados por las entidades locales dentro de cada una de las 62 sub-cuencas. sive protective measures for persistence Conjuntando estos dos conjuntos de datos, se encontraron 11,805 proyectos en la cuenca (Groot and Margolis 1991; Lichatowitch del Río Columbia, con tan sólo 13% (1,549/11,805) en común. Los Inventarios a nivel 1999). The role of salmon as a cultural de Sub-cuenca captaron más proyectos con patrocinio local que los que captaron las icon of the Pacific Northwest creates a bases de datos de gran escala. Los atributos comunes de los proyectos entre ambos grupos broad range of stakeholders that includes de datos coincidieron en menos de la mitad de las ocasiones, lo que sugiere que aquellas local, state, federal, and tribal agencies; organizaciones que trabajan en un mismo proyecto realizan sus reportes de manera private groups; fishing interests; and pri- independiente y diferente ya sea a la base de datos o a los inventarios. Se encontró vate citizens; resulting in unprecedented que la falta de estándares en los datos comunes y de conocimiento acerca de fuentes de support for habitat restoration. información accesibles son las principales dificultades en cuanto a la disponibilidad de Testing any hypothesis related to the los datos y la toma de decisiones. effectiveness of this restoration on a

232 Fisheries • v o l 35 n o 5 • m a y 2010 • w w w .f i s h e r i e s .o r g managed resource, in this case salmon, has two prerequisites: information about the resource and information about past and present restoration actions (Katz et al. 2007). However, despite hundreds of millions of dollars spent annually on Columbia River Basin management actions (GAO 2002), specific data on the implementation and monitoring of these actions— including habitat restoration, and the links between restoration and the benefits to salmon—remain largely undocumented (McDonald et al. 2007). As highlighted by a few recent studies (Paulsen and Fisher 2005; Roni 2008; Bernhardt et al. 2005), habitat restoration metadata (type, location, timing, etc.) are not readily available to inform restoration strategies, project placement, or monitoring plans for salmon recovery (Rumps et al. 2007). There are historical reasons why until recently restoration A bank stabilization project on the Cowlitz River. data had not been compiled across the Columbia River basin’s large scale. Principally, projects were conducted by local or state groups to improve local conditions—as many projects dam operation on fish and wildlife (Pacific Northwest Electric continue to be. However, listing of salmonids under the ESA Power Planning and Conservation Act 1980 [Northwest Power occurred at a different geographic scale determined by the evo- Act]). By 2005, all subbasins were required by the Northwest lution and demography of the species called an evolutionarily Power and Conservation Council (created by the Northwest significant unit (ESU; Waples 1991; McElhany et al. 2000). Power Act), to complete a plan (NWPPC 2001). The plans Salmon ESUs cover part of one or more states, with the migra- consist of three parts: an inventory of existing programs, tion corridors of these anadromous fishes crossing as many as including restoration actions and management plans (SB three states. Consequently, restoration data likely intended inventories); an assessment of the potential for fish and wildlife only for local accounting purposes must now be assessed at an improvement; and a management plan to guide the next 10-15 ESU and regional level. years of actions. Together these components were expected to The Pacific Northwest is unique in having two recently be the foundation documents for management planning into compiled large restoration datasets available to compare: the the future. Importantly, by focusing on individual subbasins, multi-state Pacific Northwest Salmon Habitat Project Tracking these reports are inherently more local in origin and scale than Database (PNW database) and the watershed-level Columbia the PNW database. River subbasin plans. Both databases were similar in goal to Here we compare the PNW database project information to guide future project planning; however, they differ in the geo- the completed SB inventories. The comparison allows a num- graphical scale of compilation. This allows for an evaluation of ber of assessments including: the role of scale in tracking restoration data, including lessons learned about reconciling data at different scales and recom- 1. estimates of the total number of projects on the ground in mendations for streamlining reporting. the Columbia River Basin; The PNW database (Katz et al. 2007) is a spatially-explicit 2. estimates of project reporting consistency provided by cross restoration database containing over 29,000 projects at over checking multiply-reported projects; 46,000 locations across all of Washington, Oregon, Idaho, and 3. evaluation of how the scale at which data is collected leads Montana, the states that overlap the Columbia River Basin. to similar or different data compilations; and It was created by researchers at NOAA Fisheries to inform 4. Indications of what makes some subbasins more successful future project placement and effectiveness monitoring. The than others at compiling and recording project data. PNW database confederated data from many sources (federal, state, watershed, and private), though the majority of the Finally, we suggest ways to streamline reporting to benefit res- projects were federally or state funded. A potential weakness toration and monitoring and evaluate the effectiveness of res- of the PNW database is that by targeting large data holders toration actions. and funders (federal and state), it may under-represent county, municipality, and other locally-sponsored projects. Local-scale METHODS management is increasingly assumed to be intrinsically more knowledgeable and more effective than management con- Data Acquisition ducted at larger scales (e.g., United Nations 1992; Tsing et al. 1999). This idea is parsimonious since scaling up of the norms We found 40 plans available on the Columbia River sub- and rules that govern environmental management are increas- basin planning website, www.subbasin.org/fw/subbasinplan- ingly challenging as the management community becomes ning/Default.htm. Together, these plans cover 58 of the 62 more diverse at larger scales (Ostrom et al. 1999). Columbia River subbasins; some of the 40 plans cover more Within the Columbia River Basin, a portion of the area than one of the 58 subbasins. Plans were not available for covered by the PNW database, a parallel inventory of restora- the following subbasins: Bitterroot, Blackfoot, or Clark Fork tion project data was assembled at the subbasin level. Sixty-two (Montana), Sandy (Oregon). We found restoration invento- subbasins were designated to manage the effects of hydropower ries in 50 of the 58 subbasins with plans (Figure 1).

Fisheries • v o l 35 n o 5 • m a y 2010 • w w w .f i s h e r i e s .o r g 233 Figure 1. Subbasin plan and restoration inventory data availability for Columbia River subbasins. The majority of subbasins contained complete plans (light gray). However, eight subbasins lacked an inventory of restoration actions in their plan (medium gray), and five subbasins lacked a plan altogether.

234 Fisheries • v o l 35 n o 5 • m a y 2010 • w w w .f i s h e r i e s .o r g Semantics: records for a subbasin. These include: (1) size of subbasin, Defining a project record and estimating project record abundance (2) total records in the subbasin, (3) the contributions to the total from the SB inventory and PNW database, and (4) who We defined a restoration project as a completed action or set of compiled the SB inventory. We found five classes of inventory actions that actively changed the physical characteristics of a site, compilers: consultant, local, state, local and state, and local thus management plans, assessments, monitoring activities, and and consultant. Local authors included soil and water conser- land acquisitions, though sometimes included in the subbasin res- vation districts, counties, Indian Nations, fish recovery boards, toration inventories, were excluded from our analysis. There was and local watershed groups/councils, among others. All inven- no universal identifier or common data dictionary that allowed tories authored by a state agency (e.g., Oregon Department us to easily find projects in common to both the SB inventory of Fish and Wildlife, Washington Department of Fish and and the PNW database (http://webapps.nwfsc.noaa.gov/pnshp). Wildlife, Idaho Department of Fish and Game [IDFG]) were Therefore, we established a series of criteria to determine if the SB considered “state.” Inventories compiled by multiple co- inventory and PNW database overlapped in projects captured. authors were given multiple categories. We further looked to We first looked at co-location; subbasin inventories report see if the compiler influenced the number of projects collected location at the subbasin (or in some cases a watershed within in the SB inventory as a percent of the total, regardless of in- the subbasin) level, while the PNW database maintains location common projects. as latitude and longitude. Thus we mapped the PNW database projects to a subbasin to allow comparison of all projects in one Comparing Metrics: subbasin at a time. Projects in the PNW database with start dates Project Attributes (Cost, Year, Location) after 2004 were excluded from this analysis as the SB inventories were completed in 2005. Next, projects presumably occurring in The in-common project records provided the opportunity the same subbasin were compared on the basis of title, descrip- to evaluate the consistency of reporting by various compil- tion, funding source, year, cost, and location details (e.g., stream ers. In-common projects were examined for cost (within $100 name). Finally, projects having the same title and/or description was considered the same), start year, and end year. Cost was and occurring over the same time frame were then considered to compared only when cost for both records was available. Costs be the same project or “in common.” that differed but by less than $5,000 were recorded separately In many cases we did not find a one-to-one relationship as a measure of similarly-reported projects (e.g., in-kind funds between projects captured by both datasets. For example, a single could be included in one cost and not another, rounding of project in the PNW database consisting of multiple installations costs could account for the differences, etc.). (e.g., riparian fencing and large woody debris addition along the Similar to cost, we compared project metadata for year and same stream segment and on the same contract) may be reported compiler. Dates that differed but by only one year, we catego- as two projects in the SB inventory data. Project records in the rized separately to accommodate possible differences in how PNW database that refer to multiple sites were maintained in parties defined when a project began (e.g., planning time vs. the format received from the data provider, which rarely allowed mapping of restoration activity to a single specific location among permitting, recording fiscal year rather than calendar year). several possibilities (Katz et al. 2007). We counted in-common We then compared the consistency of compiler metadata sub- projects as the number of projects in the PNW database that basin by subbasin. matched projects in the SB inventory. We calculated the total We also looked at differences in location reporting between number of unique records from both databases to estimate a mini- in-common records. Five subbasin plans included project mum “total” number of projects in the Columbia River Basin. data for more than one subbasin: the Lower Columbia Plan (Estuary, Cowlitz, Elochman, Grays, Kalama, Lewis, Little Comparing Metrics: Data Source White Salmon, Columbia Lower, Washougal, and Wind); Upper Snake Plan (Snake Headwaters, Snake Upper and Snake To evaluate the expectation that SB inventories better cap- Upper Closed); Boise, Weiser, and Payette Plan; Intermountain ture locally sponsored projects, we compared the percentage of Plan (Columbia Upper, Couer d’Alene, Pend Oreille, Sanpoil, federal and state projects in common to both datasets (see Katz Spokane); and the Lower and Upper Middle Snake Plan. We et al. 2007 for more information on PNW database sources). We compared each of these plan inventories to the PNW database defined local projects as those not funded by or reported in fed- both as a group as well as by individual subbasin. For example, eral or state databases. Not all subbasin inventories provided data data was queried out of the PNW database for all the subba- source or funder, but all projects in the PNW database have a sins covered by the Lower Columbia Inventory. This group of source, therefore dually reported projects were assigned to the projects was then compared to the Lower Columbia Inventory source reported in the PNW database. without regard to individual subbasin. We compared these Further, we calculated the fraction of the combined total proj- matches to the already completed subbasin-by-subbasin com- ects per subbasin in-common (C) and used this as one measure of parison to see if more matches were found, providing an addi- overlap between the two datasets: tional check on the consistency of project location reporting.

Cbasin  Obasin =   Data Analysis  Totalbasin  We then evaluated predictive factors, or covariates, that might Correlations between measures of data quality and predic- explain some of the variability in the number of in-common tive factors were performed with Kendall’s rank correlations

Fisheries • v o l 35 n o 5 • m a y 2010 • w w w .f i s h e r i e s .o r g 235 using SYSTAT 12 software (Systat Software, Inc. San Jose, basin inventories; Figure 2). Forty of the 50 subbasins with CA). inventories had projects in common with the PNW database. The SB inventory contained 2,700 records of projects that RESULTS AND DISCUSSION the PNW database lacked. Thus, there are now 11,805 total known projects in the Columbia River Basin for the 50/62 Project Overlap subbasins with inventory data (the total increases to 16,301 when including projects from the PNW database for the 12 The 50 Columbia River subbasins with restoration data subbasins without SB inventories). Of the 11,805 projects, reported a total of 4,249 projects while the PNW database 64% came from the PNW database exclusively, 23% came reported 9,105 projects for these same subbasins. Fourteen from the SB inventories exclusively, and 13% were in com- out of the 50 SB inventories provided project descriptions for mon (Figure 2). The percent of the total contributed by a all projects while 15 subbasins provided descriptions for some single SB inventory ranged from 0-80%, suggesting that projects (ranging from 25-95%). Twenty-one subbasins pro- collection methods, effort, and engagement in the subbasin vided no project description. Using this project information planning process varied widely by subbasin (subbasins that (title, description, and subbasin) we found 1,549 projects contributed 0% completed a full inventory that reported reported in both the SB inventories and the PNW database projects of many types: wildlife management, riparian assess- (representing 17% of the PNW database and 36% of the sub- ment, fisheries research, monitoring, etc. but none of the reported projects fit our definition of a restoration project). Figure 2. Relative contribution of the PNW database and subbasin inventories to the Columbia River The limited overlap between the Basin. Total broken down by data source based on source in PNW database. PNW database and the SB inventories is notable. Only 13% of the total records were common to both inventories, suggesting that the SB inventories and PNW database focused on data sources with different spatial extents and report very different snapshots of what restoration actions are taking place in the subbasin. Further, SB inventories captured a median of 27% of the total records per subbasin, missing the majority of known projects. This lack of overlap also limits our ability to test the consistency of specific metrics by allowing comparison with only a small portion of either dataset. Using the overlap metric equation, we evaluated factors that might explain some of the variability in the number of in-common records across subbasins. The fraction of total records contributed by a SB inventory correlated with the number of records in common to both data sources (τ = 0.43, P < 0.01). The fraction of the total number of records in a subbasin in common to both data sources was also correlated with the size of the subbasin (τ = 0.23, P = 0.02), but not to the total number of records in the subbasin (τ = 0.04, P = 0.67). Thus, SB inventories captured a greater fraction of completed restoration in larger basins, but not necessarily in basins with more restoration data. In contrast, the relative performance of the PNW database in capturing records did not correlate with basin size or data abundance.

236 Fisheries • v o l 35 n o 5 • m a y 2010 • w w w .f i s h e r i e s .o r g Project Source Subbasin restoration inventories were compiled by five different entity types: consultant, local, state, local and state, and Two-thirds of the PNW database (updated since the publi- local and consultant. Different areas used different entities that cation of Katz et al. 2007) came from three federal and three were spatially clustered in the Columbia River basin (Figure 3). state sources: the Regional Ecosystem Office of the U.S. Forest For example, most inventories in Idaho were compiled solely by Service, the Bureau of Land Management, the Columbia Basin state agencies and all inventories in Oregon were compiled by Fish and Wildlife Authority, the Oregon Watershed Restoration local or state entities. Local entities were responsible for col- Inventory (OWRI), the Washington Salmon Recovery Funding lecting close to half (20) of the inventories basinwide followed Board (WA SRFB), and the Idaho Department of Fish and by consultants (14; Figure 4). Inventories compiled jointly by Game. Of the 4,249 records contributed to the total by the SB state and local, or only by state representatives, most thoroughly inventories, 83% were from sources other than the six major captured restoration projects; however, both of these categories federal and state datasets (Figure 2). Thus, the SB inventories also had the largest variance. The variability in the overlap met- did not rely on large federal and state data sources to compile ric, the fraction of the total number of records in a subbasin in their inventories. Since the PNW database targeted large data common to both data sources, is in some cases limited by the holders, including states, one might expect greater agreement available data in the subbasin. Coefficients of variation in the between the PNW data and SB inventories compiled by state overlap metric by compiler indicated a difference (0.71-0.85) agencies. One would then expect less overlap with locally- across subbasins. Inventories compiled solely by local groups sourced inventories which would potentially report different included the least coverage of state-sponsored restoration proj- details on a local level project (e.g., when there are many ect data. This may be due to some combination of causes includ- contributors to a single project, local groups may only report ing limited funding to support the subbasin process and the level their contribution as the project attributes). Bernhardt et al. of awareness of large-scale data availability on the part of inven- found that on average a single river restoration project has tory compliers (e.g., not looking beyond local sources). Further, more than three funding sources and seven to eight different inventories compiled by both local and state representatives entities or partners (Bernhardt et al. 2007). Project partners or found a greater percentage of project data than local groups funders reporting results of a project separately or differently working alone. State representatives may have contributed may explain some of why we found limited overlap between knowledge about state-level restoration datasets or have access the PNW database and SB inventories. to additional resources that local groups alone lacked.

Figure 3. Spatial distribution of restoration inventory compiler by subbasin.

Fisheries • v o l 35 n o 5 • m a y 2010 • w w w .f i s h e r i e s .o r g 237 and PNW database. Our expectation was Figure 4. Box and whisker plots summarizing the percent of the total (T) contributed by the subbasin inventory in a given subbasin by compiler type (#SBInventory basin/ Totalbasin). Hatched that records for the same project, no mat- line indicats global mean. Width of the boxes indicates the number of plans contributing. ter who reported it, would have the same attributes. Ten of the 40 subbasins with in-common records reported none of the three metrics, so only 1,449 projects could be compared.

Cost

Only 6 of the 40 SB inventories with in-common records reported cost. Further, because not every record in the PNW database includes cost (Katz et al. 2007), only 104/214 records had cost reported by both sources, allowing a direct comparison. When cost information was available, 71 (33%) of the dually-reported projects had the same cost, and 33 (15%) had different costs (Figure 5). The median and average cost per project for the six SB inventories were $35,000 and $139,000 respectively (n = 355),while the median and average cost per project in the PNW database for these same six subbasins were $16,000 and $116,000, respectively, n = 874 (for all 62 subbasins $14,000 median, and $158,000 Inventories compiled by the state in Idaho (7) reported a median of 96% of the average, n = 10,501). state-sponsored (IDFG) projects in the PNW database while SB inventories compiled by Thus, the SB inventories had both a consultants (7) in all states found only 25% of state (IDFG, WA SRFB, OWRI) projects. higher median and average cost than the Surprisingly, Idaho Department of Fish and Game was also the only data source in the PNW database. This is surprising given PNW database for which the SB inventories found more than half of the known proj- our expectation that the SB invento- ects. This may result from a combination of factors, including state recording practices ries collected smaller-scale, more local that include more complete metrics allowing for easier matching, facilitated data sharing restoration project data. More impor- agreements, and significant awareness of state-sponsored projects among state staff. tantly, this suggests that there are large, Interviews with compilers would be required to identify the reasons for the expensive local projects taking place in oversight of regional data sources. SB inventories are expected to guide the next addition to what is reported in state and 10–15 years of management actions; overlooking regional data sources severely federal databases. limits the restoration data available to influence what types of future Figure 5. Comparison of reported metrics (cost n = 214, start date n = 1,431, end date n = 1,430) projects are conducted in the subbasin. for projects reported in both the subbasin inventories and PNW database. Including regional data sources could change both the project density and project type composition of the subbasin, thus changing priorities for future project placement. While access and knowledge of data appear to be the major limitations to using data, subbasin inventory compiler alone was not a good predictor of the amount of in-common projects (τ = 0.10, P = 0.31).

Project Attributes: Cost, Year, Compiler, Location

To evaluate the relative data quality in the different reporting systems, we compared the attributes of the 1,549 projects reported in both the SB inventories

238 Fisheries • v o l 35 n o 5 • m a y 2010 • w w w .f i s h e r i e s .o r g Year Figure 6 A, B, C. Comparison of reported Almost twice as many subbasin inventories metrics (cost, start date, reported start and end year than reported costs end date) for projects reported in both the (29/40 start year, 30/40 end year). About half of subbasin inventories the in-common projects reported the same start and PNW database and and end years. However, 15% of what appear grouped by SB inventory to be in-common records (based on title and compiler. description) had different start or end years, half of which differed only by 1 year (Figure 5). This suggests that in many cases, different organizations involved in the same A. restoration projects are reporting the project independently and differently to either the PNW database or the SB inventory sources. For example, a local watershed group collaborates with a state fish and wildlife agency, and both groups reported the project; however, the attributes reported were not identical. With the available data, it is impossible to tell if projects with vastly differing start and end years may or may not actually be in-common projects. This illustrates the inadequacy of using only a title or description as a project identifier and the need for standardized project reporting metrics to allow in-common projects to be easily identified.

Compiler

We further examined the data to see if inventories by different compilers had different levels of metric reporting consistency. State-compiled inventories provided start/end year matches; however, no state compiled inventories reported cost (Figure 6 a,b,c). Any compiler that included local (local/state, local/consultant, local) had close matches (± 1 year) for the B. majority of non-identical start and end years, suggesting minor differences in reporting of the same project. Only projects in SB inventories compiled by consultants or local groups reported costs. Of these, consultant cost metrics were more often the same as the records in the PNW database. Data agreement aside, among the potential in-common records, half had insuf- ficient documentation of cost to allow a comparison.

Location

Five subbasin plans included project data for more than one subbasin, allowing us to look at the consistency of reporting project records at a larger, multi-subbasin scale. For example, in the Lower Columbia (10 subbasins) we found 22 additional in-common records (18% increase) by comparing the 10 inventories as a group to the PNW database than we did matching subbasin by subbasin. Similarly in the Upper Snake SB inventory we found 11% (6) more matches when comparing the inventory for all three subbasins as a group than individually (54). However, the Boise, Payette, Weiser C. Plan and the PNW database were in agreement on all project assignments to subbasins but one. In the Lower and Upper Middle Snake Inventory, subbasin was not specified for all projects, however most had either a U.S. Geological Service Hydrologic Unit Code (watershed) or county specified. From the location information available, we were able to assign each project to a subbasin and found no discrepancies between datasets. However, in the process of dividing the inventory into Lower Middle and Upper Middle Snake, we found 29 projects that were located outside of both the Upper and Lower Middle Snake subbasin (compared to 29 total projects positively identified as Upper and 14 as Lower). Out of all five multi-subbasin plans covering a combined 24 subbasins (or over one-third of all Columbia River subbasins), only the Intermountain Plan (Columbia Upper, Couer d’Alene, Pend Oreille, Sanpoil, Spokane) assigned all in-common projects to the same subbasin as the PNW database.

Fisheries • v o l 35 n o 5 • m a y 2010 • w w w .f i s h e r i e s .o r g 239 For the remaining 16 subbasin inventories that reported Compiling the PNW database provides numerous lessons individually (i.e., not part of a multi-basin inventory), we on what contributes to a good inventory, but this comparison were not able to identify additional in-common projects where with the SB inventories elucidates other reporting challenges. assignment to subbasin differed. As the fundamental spatial In the Pacific Northwest, projects are being implemented and framework of the SB inventories was the subbasin, we com- reported separately by entities ranging in scale from single Boy pared projects subbasin-by-subbasin and did not have a rea- Scout troops to large federal agencies. Furthermore, reporting soned basis for cross-referencing project records from other standards and requirements vary drastically across the region basins. Consequently, we did not estimate a correction for depending on the project sponsor and funding source. This matching projects assigned to different subbasins in making an lack of a single data repository or at least common standards estimate of the total number of restoration actions across the in distributed databases creates further inconsistencies as the entire Columbia River Basin. project sponsor, contributing agencies/organizations, and project funder all may report the same project and with slightly Lessons Learned and the Need for Common Data Standards different attributes. This can result in apparent bias, such as in the total effort expended and other characteristics (Bernhardt Our findings demonstrate the variation in two attempts to et al. 2007) as we discovered when comparing the cost, year, compile restoration data across a large geographical area. This and location of projects. reveals important opportunities to improve the reporting of As characterized in this study, failure to resolve multiply- management actions and the accountability for the resources reported projects on the basis of project attribute not only may applied to those actions. Methods, effort, and engagement var- result in over-reporting inefficiencies, but results in increased ied widely by SB inventory as demonstrated by the diversity regional cost inefficiencies in the time spent resolving the of compilers and variation in percentage of in-common proj- data. A standardized definition of a project and basic project ects. In general, we found that SB inventory data was focused metrics (identifier, date, description, etc.) that are consistently less on state and federal data sources and thus assume it was reported would go a long way to solving this problem. Perhaps focused more on locally-sponsored projects. The comparison the most crucial reporting metric is spatial location. Absent with the SB inventories revealed that the PNW database is consistent and specific location information (latitude, longi- missing as much as one-fifth of the available restoration data tude), many projects were misassigned to subbasin, creating a and the missing data is likely for local projects. These missing significant hurdle to correlating projects across reporting sys- local projects were not necessarily inexpensive, implying they tems. In some cases, the location of the agency or organization are larger and more significant restoration efforts that are lack- doing the work may have been used rather than the subbasin ing in the PNW database. Subbasin-scale inventories appear to the work was actually located in. capture locally-sponsored projects better than the large-scale Beyond data verification, spatially explicit project data can data inventories (such as the PNW database), but it is still not be used in a geographic information system (GIS) and com- clear whether they have better knowledge of projects taking bined with data such as salmon habitat utilization or water place in their local watershed. Three data sets publicly avail- temperature to evaluate treatment-response effectiveness. The able on the web, the Regional Ecosystem Office Interagency added value of spatial data also comes with added complex- Restoration Database, the Oregon Watershed Restoration ity and data management difficulties since a project often has Inventory, and the Washington Salmon Recovery Funding multiple restoration sites (Katz et al. 2007). Having some con- Board, were not well utilized in the subbasin inventories. Yet, sistent spatial references for projects also allows association small local datasets included in the SB inventory were not with more diverse, and in some cases familiar and useful spa- locatable by the PNW database. This suggests that knowledge tial attributes (e.g., county, state, stream, etc.) that makes spa- of a data source’s existence may be a better predictor of inclu- tial data usable to groups working outside of GIS (e.g., query sion in a database than the ability to access the data. for the number of projects on a given stream or in a county). Though we observed that the total records for the Columbia Creating the full set of spatial data requires time and data River Basin include thousands of projects not reported in SB management capabilities as well as adequate support resources, inventories and concluded that SB inventory compilers knew beyond what is required to store only tabular data. Absent this how to locate more local sources of data—despite missing approach to spatial information, restoration data can only be a lot of large-scale, accessible information about their local handled as it was in the SB inventories with presence/absence areas—there is an alternative hypothesis. Local entities may at a subbasin or even larger geographic scale, which ultimately know both their local data and local areas better, and the large limits its usefulness. amount of additional data available represents an over-report- Significant resources were required to compile the PNW ing of management investments in habitat improvement by database in part as a consequence of the lack of standardized state and federal agencies. The quality of restoration action reporting or a centralized reporting system (Katz et al. 2007). data is an accumulated consequence of the actual amount This resulted in numerous partially complete records and a of projects in the field, the quality of the data reporting and long latency between project initiation and reporting of the record keeping, and the quality of our investigative work in project (Katz et al. 2007). It is very likely that the SB inven- discovering data. However, eliminating the hypothesis that tories lacked information from some of the larger, regional large numbers of these records do not correspond to work done data sources (e.g., REO, OWEB etc) because they lacked the on the ground requires visits to the sites and evaluation of the resources to reconcile the diverse data formats. The lack of quality of the data reporting. We are currently evaluating this consistent sources to find all restoration data is a larger regional and will report those results in a subsequent publication. problem, if not a national restoration problem (Bernhardt et al.

240 Fisheries • v o l 35 n o 5 • m a y 2010 • w w w .f i s h e r i e s .o r g 2005; Palmer and Allen 2006) A national inventory in 2005 Katz, S. L., K. Barnas, R. Hicks, J. Cowen, and R. Jenkinson. found 37,000 projects, of which 62% (23,000) were from the 2007. Freshwater habitat restoration in the Pacific Northwest: PNW database (Bernhardt et al. 2005). Though the PNW may a 10-year census. Restoration Ecology 15:494-505. have a greater density of restoration effort than many parts of Lichatowich, J. 1999. Salmon without rivers: a history of the the country, four states accounting for 62% of all national proj- Pacific salmon crisis. Island Press, New York. ects is not our expectation. It suggests that there may be simi- McDonald, L. L., R. Bilby, P. A. Bisson, C. C. Coutant, J. M. lar data management issues at work in the rest of the country Epifanio, D. Goodman, S. Hanna, N. Huntly, E. Merrill, B. and this disparity reflects data accessibility limitations rather Riddell, W. Liss, E. J. Loudenslager, D. P. Philipp, W. Smoker, than the actual distribution of on-the-ground projects. More R. R. Whitney, and R. N. Williams. 2007. Research, monitor- money and effort are spent on restoration in the Columbia ing, and evaluation of fish and wildlife restoration projects in River Basin than anywhere else in the United States, yet basic expectations for what data should be available (standardized the Columbia River Basin: lessons learned and suggestions for fields, location, monitoring) are still not met. Even with the large-scale monitoring programs. Fisheries 32(12):582-590. number of papers and groups providing guidelines for captur- McElhany, P., M.H. Ruckelshaus, M. J. Ford, T. C. Wainwright, ing restoration data (Palmer et al. 2005), we found in both the and E.P. Bjorkstedt. 2000.Viable salmonid populations and PNW database compilation effort and the SB inventories that the recovery of evolutionarily significant units. U.S. Dept. only the most basic project implementation data is available. Commerce, NOAA Tech Memo NMFS-NWFSC-42. Here we examined the products of two attempts at compiling NWPCC (Northwest Power and Conservation Council). restoration data to inform future project planning and analysis. 2001. Technical guide for subbasin planners. NWPCC, To assess the impact of restoration on habitat quality for sal- Portland, Oregon. Available at: www.nwcouncil.org/ monids requires more specific information on specific habitat library/2001/2001-20.pdf. project-level metrics (e.g., number of culverts removed, spe- Ostrom, E., J. Burger, C. B. Field, R. B. Norgaard, and D. cies and amount (acres) of planting etc.), the level of habitat Policansky. 1999. Revisiting the commons: local lessons, alteration from natural conditions, and the distribution of sal- global challenges. Science 284:278-282. monid population limiting factors. Compiling restoration data Palmer, M.A., and 22 co-authors. 2005. Standards for ecologi- into a database is not an answer to the question of “what kind cally successful river restoration. Journal of Applied Ecology of response are we getting for all the restoration effort?” Yet, 42:208-217. it does inform our understanding of what information people have available to them to make decisions about restoration. Palmer, M. A., and J. D. Allan. 2006. Restoring rivers. Issues In the absence of targeted monitoring of habitat response to in Science & Technology, National Academy of Sciences, restoration, however, project data is one of the only tools man- Winter:40-48. agers have to use in making the best tactical decisions about Paulsen, C. M. and T. R. Fisher. 2005. Do habitat actions affect future restoration activities. juvenile survival? An information-theoretic approach applied to endangered Snake River Chinook salmon. Transactions of ACKNOWLEDGEMENTS the American Fisheries Society 134:68–85. Roni, P., K. Hanson, and T. J. Beechie. 2008. Global review We wish to thank Y. Lucero, T. Beechie, and C. Jordan for of the physical and biological effectiveness of stream habitat insightful comments on earlier drafts of this manuscript. We rehabilitation techniques. North American Journal of Fisheries would also thank two anonymous reviewers whose recommen- Management 28:856-890. dations improved the article greatly. The entire content pre- Ruckelshaus, M. H., P. McElhany, and M. J. Ford. 2003. sented here does not constitute nor reflect the official views Recovering species of conservation concern: are populations of NOAA Fisheries, the National Oceanic and Atmospheric expendable? Pages 305-329 in P. Kareiva, and S. A. Levin, eds. Administration, or the U.S. Department of Commerce. The importance of species: perspectives on expendability and REFERENCES triage. Princeton University Press, Princeton, New Jersey. Rumps J. M., S.L. Katz, K. Barnas, M. D. Morehead, R. Bernhardt, E. S., and 24 co-authors. 2005. Synthesizing U.S. Jenkinson, S. R. Clayton, and P. Goodwin. 2007. Stream res- river restoration efforts. Science 308:636–637. toration in the Pacific Northwest: analysis of interviews with Bernhardt, E. S., E. B. Sudduth, M. A. Palmer, J. D. Allan, J. project managers. Restoration Ecology 15:506-515. L. Meyer, G. Alexander, J. Follastad-Shah, B. Hassett, R. Tsing, A., J. Brosius, and C. Zerner. 1999. Assessing commu- Jenkinson, R. Lave, J. Rumps, and L. Pagano. 2007. Restoring nity-based natural-resource management. Ambio 28:197-199. rivers one reach at a time: results from a survey of U.S. river United Nations Department of Economic and Social Affairs. restoration practitioners. Restoration Ecology 15:482-493. 1992. Report of the United Nations conference on environ- GAO (General Accounting Office). 2002. Columbia River Basin ment and development, Rio de Janeiro, Brazil. United Nations, salmon and steelhead: federal agencies’ recovery responsibilities, expenditures and actions. U.S. General Accounting Office New York. Technical Report GAO-02-612. Waples, R. 1991. Pacific salmon, Oncorhynchus spp., and the Groot, C., and L. Margolis (editors). 1991. Pacific salmon life definition of “species” under the Endangered Species Act. histories. University of British Columbia Press, Vancouver. Marine Fisheries Review 53:11-22.

Fisheries • v o l 35 n o 5 • m a y 2010 • w w w .f i s h e r i e s .o r g 241 Column: Gail Goldberg Guest Director’s line Goldberg is the AFS units coordinator and may be contacted at Certification— [email protected]. Promoting the Importance of Fisheries Professionals

Certification enhances the pro- was an opportunity to provide fur- downturn, did not agree with paying fessional image and formalizes ther comments regarding perceived a fee to renew their certification every accomplishment levels. AFS began benefits. Here most responses were five years; one respondent said it was the Certification Program at the positive and supported the concept of “too expensive to renew with the fur- 93rd Annual Meeting in 1963, when enhanced credentials. However, a fre- lough salary reductions in California.” attendees approved standards devel- quent negative comment was about Moreover, there were comments from oped by the Professional Standards the lack of any financial benefit. state employees noting the difficulty Committee. Certified Fisheries Finally, our survey asked for “gen- of renewing certification with less or Professionals are recognized as those eral comments on the Professional even no funding available to support who achieved specific standards of Certification Program.” Many of the travel to meetings and continuing professional competence. positive comments reinforced the education. notion of personal satisfaction, with We are approaching the 50th Other negative comments also comments such as: anniversary of the AFS Certification cited the absence of financial Program, so we conducted a survey * Very worthwhile program, I benefits. of Certified Fisheries Professionals always encourage my students While I value my certification, to assess the program’s efficacy. to set this as one of their my employer does not. I work for About 680 certified individuals were career goals. asked to provide feedback on the a state agency and there is no * Of greatest value early in my program and over 250 responses incentive for promotion, merit career. were received—a respectable 36% pay, or any other consideration response rate. Here is some of what * Considered important in the for obtaining certification. A we learned. environmental consulting Certified Fisheries Professional We first measured the level of industry. is not recognized in the same satisfaction with the application pro- * Fine program with the right way as a Professional Engineer cess, including the application form, mix of academic education (PE) and that is unfortunate for web site information, and online and professional experi- our profession….Sadly, CFP has submission. In each of these areas, ence while fostering AFS not been considered in the hir- the majority of replies were “very involvement. ing decisions of staff internally satisfied” (50–75%). Second, we * AFS needs to push hard for or externally. This is probably asked how rigorous were the require- participation in a rigorous because rarely is someone in a ments to obtain Certified Fisheries program for the sake of the hiring position also a CFP. AFS and Professional status. Eighty-six percent profession. the profession need to find a way thought the requirements were about to make certification more recog- * As I apply for increasingly right. Respondents were then asked nizable, accepted and a minimum advanced positions within my about the renewal process and how condition for employment. Those agency I have appreciated the they would rate the requirements of us in my agency who are certi- certification opportunity pro- to maintain certification and apply fied have done so because we for renewal. Again, most (79%) said vided by AFS because it gives me a chance to document believe in a higher standard, but these requirements were about right. that is not the norm apparently in The survey also asked what a broad range of education and experience that I believe our profession. benefits were derived from certifi- are important for modern cation as a Fisheries Professional. We appreciated the many thought- scientists working on fisheries Seventy-eight percent of respondents ful suggestions that were received. issues. selected “achievement of a career An often repeated recommendation goal” and “personal satisfaction” as Negative comments focused on was to get state and federal agencies the main benefits. After that, 42% the renewal program. Many, citing a to endorse or provide compensation cited “recognition by peers.” There consequence of the general economic based on professional certification.

242 Fisheries • v o l 35 n o 5 • m a y 2010 • w w w .f i s h e r i e s .o r g I wish more (all) state agen- ship that they should take the echo what we learned here from cer- cies recognized the importance of time to complete the form. tified individuals. In his conclusions, the Certified Fisheries Professional Jubar stated, We as a Society should do status and would use it for hiring Although very few state agen- MUCH MORE to promote the and adjusting salaries. cies require new hires to either importance of certification. meet the requirements of or pos- Geologists did this about 20 We need to make certification sess AFS certification, applicants years ago and now being a PG is more meaningful….for example who have these credentials are important in that field. getting more employers to com- looked favorably upon….Most pensate for certification. We need to promote our states do not offer incentives in profession and increase the the form of promotions or wage AFS and the profession need rigor of certification so it means increases for current employees to find a way to make certifica- something. Consider recognizing that obtain their certification. tion more recognizable, accepted, certified professionals at annual However, a few respondents and a minimum condition for meetings? Perhaps a note on indicated they were encouraged employment. their name badge and/or foot- by supervisors to seek certification. In some cases, agencies A movement in that direction actu- note on abstracts they author. will support employees who ally can be detected, with state agen- It must be noted that some AFS are actively seeking certification cies in Alabama and Kentucky using Units are taking the lead in that by allowing them work time to or requiring professional certification regard. In particular, the effort of the complete application materials, in the employment process. Atlantic International Chapter (AIC) to and possibly reimburse expenses Another commenter suggested encourage professional certification incurred such as the application that AFS leaders demonstrate their is especially worthy of mention. AIC fee or fees associated with train- support of the professional certifica- has 15 active members that are certi- ing or coursework required for tion program by becoming certified fied and current AIC President Scott certification. themselves. Craig has just announced an initiative to double that number before the More information on AFS When it becomes important to Annual Meeting in Pittsburgh. To sup- certification, including guidelines, AFS, then it will start to become port this effort, the Chapter will offer frequently asked questions, important to our members…. financial assistance. application materials, and a list of Leaders lead, and in this case In December 2009, Aaron Jubar, certified professionals, is available at become certified. That demon- with the AFS Texas Chapter, surveyed the AFS website at www.fisheries. strates to the general member- state fisheries agencies. His findings org/afs/certification.html. SRX-DL Data-logging Radio Receiver

Now with dual-antenna capability for upstream/downstream discrimination Low cost translates to more stations and better coverage High sensitivity means greater detection range Weatherproof provides for more deployment options

Fisheries • v o l 35 n o 5 • m a y 2010 • w w w .f i s h e r i e s .o r g 243 NEWS: AFS UNITS

Jim Hawke, Hans Stephenson, and Ev Hoyt, representing the Black Hills Fly Randy Hiltner, Awards Committee Randy Hiltner presents the Best Fishers, present Dennis Unkenholz with a plaque in recognition of his years chair, presents the Best Graduate Student Poster Award to of service to fisheries conservation in South Dakota. Undergraduate Presentation Michael Weber. Award to Jessica Howell.

Dakota Chapter of engagement stressed Location on Survival Meets in Spearfish, by the Chapter leadership, Growth, and Tag Retention South Dakota angler groups, such as Using Passive Integrated The Dakota Chapter’s the Black Hills Fly Fishers Transponders.” Best annual meeting was and the Rapid City and Graduate Student Paper held in Spearfish, South Sioux Empire chapters of went to Justin VanDeHey Dakota, in the heart Walleyes Unlimited, were for his paper titled, of the Black Hills on involved both financially “Effects of Simulated Cold- February 22 through 24. and in the technical ses- fronts on Hatch Success The theme was “New sions. The meeting closed and Survival of Yellow Fisheries Solutions for a with the awards ceremony Perch.” Greg Wanner of Randy Hiltner presents Best Graduate New Decade.” With over and banquet. the U.S. Fish and Wildlife Student Paper to Justin VanDeHey. 125 participants and 50 South Dakota State Service won the award presentations, the meet- University (SDSU) students for Best Professional ing was well attended and dominated the best stu- Presentation, “Spatial full of information. The dent presentation awards and Temporal Patterns in meeting started with a fish ceremony with Jessica the Niobrara River Fish disease continuing educa- Howell winning the Best Community.” tion course taught by Rick Undergraduate Paper with Several scholarships Cordes. A special retiree her presentation “Diet were awarded, includ- social was held, during Overlap Patterns Among ing the Schmulbach which former member Age-0 Common Carp and Scholarship presented Dennis Unkenholz was Four Native Fishes.” The to Bobbi Adams from recognized for his induc- Best Graduate Student SDSU. Two Alven Kreil tion into the Fisheries Poster was awarded to Memorial Scholarships Management Hall of Matt Wipf is recognized for his Michael Weber, for his were presented to Nikki exemplary efforts as assistant editor Excellence. Also, in keep- poster titled, “Effects of Lorenz and Jake Mertes. of the Dakota Chapter Newsletter. ing with the positive spirit Morphology, Size and The Sauger Scholarships

Bobbi Adams receives the The Sauger Scholarships are presented to: Donna Abler, Jessica Howell, Bobbi Adams, Schmulbach Scholarship from Will and Nikki Hegna from SDSU by McLain Johnson (middle), president of the SDSU Student Sayler, Schmulbach Committee Subunit. Chair.

244 Fisheries • v o l 35 n o 5 • m a y 2010 • w w w .f i s h e r i e s .o r g Randy Hilter presents the award Bob Hanten and Greg Simpson Dave Lucchesi presents Norm Ken Edel accepts the Aquatic for Best Professional Presentation are presented with awards for Kopecky with a Special Resource Conservation Award to Greg Wanner of Pierre, South Distinguished Professional Service Recognition Award. from Gene Galinat. Dakota. . were presented to Donna Matt Wipf from Black Hills Abler, Bobbi Adams, Jessica State University was recog- Howell, and Nikki Hegna nized for his work as asso- from SDSU. ciate editor of the Dakota Awards were presented Chapter newsletter. for recognition of profes- The Dakota Chapter also sional service and contribu- held a photography contest tions toward improving the and accepted photographs appreciation and use of from Chapter members, aquatic resources in their who had to be present to communities. Bob Hanten win. The categories and and Greg Simpson won the winners were: Black Hills Chapter Walleyes Unlimited members Distinguished Professional Angling— Denny Bohls and Mike Johnson accept the Aquatic Service Awards. A Special Breanne VanDeHey; Conservation Award from Gene Galinat. Recognition Award went Kids and Fishing— to Norm Kopecky for Justin VanDeHey; Breanna VanDeHey; and his efforts in introduc- Fisheries in Action— Best in Show— ing recreational angling Michael Weber; Michael Weber. to special needs indi- Just Fish—Wes Bouska; New Chapter officers viduals. Aquatic Resource Scenery—Joshua Peters; were elected at the meet- Conservation Awards were Humor—Luke Schultz; ing. Bill Haase (secretary/ presented to the Rapid City Wildlife—Will Schreck; Area Chapter of Walleyes Outdoor Activities— treasurer), Paul Bailey Unlimited and Ken Edel. Matt Wipf; (vice president), and Chris Will Sayler was presented Miscellaneous— Longhenry (president elect) with a “Parliamentarian Tyler Berger; join President Mike Barnes. Extraordinaire” award and People’s Choice— —Kasondra Brooke Will Sayler is presented with a Parliamentarian Award.

Photography winners pose with Chapter member Michelle Bucholz, pictured in the left front.

Fisheries • v o l 35 n o 5 • m a y 2010 • w w w .f i s h e r i e s .o r g 245 COLUMN: PRESIDENT’S HOOK Continued from page 212

game birds), mountain goat, whitetail deer, and tilapia and common carp fisheries attendance but unable to travel to the elk, black bear, wild turkey, striped bass, everywhere? What about brook trout in event, AFS and TWS provided live stream- rainbow trout, brown trout, brook trout, Colorado, channel catfish and flathead cat- ing of the symposium through their respec- smallmouth bass, largemouth bass, walleye, fish along the eastern seaboard, and striped tive websites. Summary papers will be bluegill, channel catfish, tilapia, Nile perch, bass along the California coast? When I developed for Fisheries and for The Wildlife red drum, peacock bass, Pacific salmon… review the above examples I will say “yes” Professional. Abstracts from presentations, and the list goes on and on…because we to some and “no” to others. But my two along with a webcast of the event, will have decided that we like these plants lists very likely differ from yours. be available on AFS and TWS websites so and animals and are willing to cultivate Last month the American Fisheries that members from both societies will have them, generally at the expense of other life Society (AFS) and The Wildlife Society access to the information shared during this forms. We have extirpated native species, (TWS) addressed the issue of “gardening” assemblages, and entire communities, on through a collaborative symposium on TOM, and be able to “attend” the meeting terrestrial landscapes as well as in aquatic species introductions and reintroductions. virtually and at their leisure. systems, and oftentimes have supplanted Held on the campus of Mississippi State The meeting allowed participants oppor- them with less complex (and thus in our University, it was the first such joint venture tunities in both formal and informal settings minds more efficient) assemblages (often of sister organizations from within the (i.e., two offsite evening events and end-of- monoculture). newly-established Consortium of Natural conference field trips) to discuss the topic It is all a matter of value systems and Resources Societies (AFS, TWS, the Society at hand. Good science was presented and, what we see as our professional roles and of American Foresters, and the Society of as the conference evolved, trans-disciplinary responsibilities. We certainly make mistakes Range Management). This was also the common denominators were revealed. along the way. But is it a mistake to build first venture by AFS to conduct a topic- These common denominators revealed a pond or a reservoir, stock it with fish that oriented meeting (TOM) which, by design opportunities for new relationships, new people like, and establish a fishery? Was it and purpose, is a highly-focused event, synergism, and new collaborative ventures a mistake to create rainbow trout fisheries rapidly-organized (< 1 year from idea to between and among participants who, in cold tailwaters throughout the eastern event closure), and is conducted within prior to the meeting, were in many cases United States (I’ll also leave it to you to the framework of a relatively small budget unknown to one another. There was determine if the dams that created the (<$20,000). During this meeting there incredible energy in the group during the tailwaters were a mistake or not); large- were over 50 participants from through- mouth bass fisheries in California, Japan, out North America and beyond, 20 oral conference and a “sparkle in their eyes” as and Kenya; Pacific salmon fisheries in the presentations, and a dozen posters. The they rode the crest of the wave at the end Great Lakes and New Zealand; peacock plenary session included keynote addresses of the conference. I am firmly convinced bass fisheries in Gatun Lake (Panama); Nile by Dan Simberloff (University of Tennessee- that we can, in the future, expect increas- perch fisheries in Lake Victoria (East Africa); Knoxville), and Jeff Hill (University of Florida ingly good guidance from these and other landlocked coho salmon fisheries in central and president of the AFS Introduced Fish “Master Gardeners” within our profes- Alaska; brown trout fisheries in Patagonia; Section). For persons able to schedule sional ranks.

Oregon RFID

246 Fisheries • v o l 35 n o 5 • m a y 2010 • w w w .f i s h e r i e s .o r g Fisheries • v o l 35 n o 5 • m a y 2010 • w w w .f i s h e r i e s .o r g 247 Column: Students’ Angle

Excellence in Fisheries Management Made Possible by the Fenske Fellowship

Heidi Ziegenmeyer dents from the under-served com- Mary Bremigan. My mentors led me munity with a year-long experience through the Fenske Fellowship experi- Ziegenmeyer is a second year designed to assist the recipient in ence, guiding me through several master’s student in the Department learning about the role and activities meetings. I attended two Michigan of Fisheries and Wildlife at Michigan of fisheries management agencies. DNR Fisheries Division Management State University. Her research focuses During the year, recipients interact Team Meetings where I observed on the effects of spring angling on with numerous fisheries professionals, the dynamics of communication and black bass nesting success in inland work on a project of their design to decision-making at the state level by Michigan lakes. Her work is supported inform management decisions, and witnessing discussions among repre- by the Janice Lee Fenske Excellence learn from great mentors. Priority is sentatives from management units, in Fisheries Management Fellowship, given to female applicants with excel- the research section, and the chief of a grant from the Michigan DNR, and lent academic and service records the Michigan DNR Fisheries Division. I an MSU Teaching Assistantship. More and an interest in fisheries manage- also participated in meetings between information on the Fenske Fellowship ment. The fellowship was created to the Michigan DNR Fisheries Division can be found at www.fw.msu.edu/ honor the commitment, integrity, and and other agencies, such as the fellowships.htm or by contacting memory of Jan Fenske (see inset), Huron-Erie Corridor Initiative Meeting Ziegenmeyer at [email protected]. Michigan’s first female fisheries biolo- at the Environmental Protection gist, who passed away in 2005. Agency offices in Detroit, Michigan. As a Fenske Fellow I worked closely There fisheries professionals from When I learned about the Janice with mentors within the former multiple states and Canada interacted Lee Fenske Excellence in Fisheries Michigan Department of Natural to work toward the common goal of Management Fellowship, I jumped Resources (DNR, now Michigan managing the shared resource. I also at the opportunity to apply. That’s Department of Natural Resources and had the opportunity to attend several because the fellowship is not your Environment), Gary Towns (fisheries other meetings, such as the AFS average funding opportunity. The supervisor, Lake Erie Management Michigan Chapter and society-level Fenske Fellowship provides Michigan Unit), and Jim Breck (fisheries research AFS meetings, Great Lakes Fishery State University Department of biologist, Institute for Fisheries Commission Lake Committee meet- Fisheries and Wildlife graduate stu- Research) as well as my major pro- ings, and the Michigan DNR Fisheries fessor at Michigan State University, Research/Biologist Meeting.

Left: Ziegenmeyer poses before snorkeling with the bass on a rainy day at Long Lake in May 2009. Center: Ziegenmeyer leans over the boat to look at a bass nest using a viewscope in 2008. Right: Hands full of equipment, Ziegenmeyer works to survey a bass nest in 2008.

248 Fisheries • v o l 35 n o 5 • m a y 2010 • w w w .f i s h e r i e s .o r g My diverse group of mentors also provided significant input in planning On 10 March 2005 worked untiringly for my fellowship research project, an Janice Lee Fenske the protection and passed away after a wise use of Michigan’s assessment of a change in Michigan courageous battle with natural resources. Her DNR fishing regulations allowing cancer. Born in 1954 in numerous achievements angling of black bass during the Grand Rapids, Michigan, led her to receive spring nesting season. My mentors Fenske received her many awards from the guided me through the process of bachelor’s degree in Fisheries Division, Trout planning a large, collaborative project zoology from Michigan Unlimited, and from involving MSU students as well as State University in 1976 Janice Lee Fenske the American Fisheries and her master’s degree First female fisheries biologist at Society. She was very members of the Michigan DNR from in fisheries management the Michigan Department proud of her association across the state. We worked together of Natural Resources. in 1983 from the with the latter and to plan black bass nest survey meth- University of Michigan. In 1978, particularly enjoyed participating in ods and coordinated electrofishing Fenske began her career in fisheries professional debates and discussions at efforts across four different study management by accepting a position as Society meetings. lakes. From my mentors I learned a fisheries technician with the Michigan Fenske blazed a trail for women more about conducting research, but Department of Natural Resources working in fisheries management in equally invaluable is the example they Fisheries Division. Five years later she Michigan by entering the Fisheries was promoted to fisheries biologist, the Division, then considered a “man’s set for me concerning the importance first female to hold such a position in division.” She succeeded in her of communication and seeking input Michigan. She subsequently became endeavors by facing the toughest of on the project from managers, biolo- the first female district fisheries biologist challenges with a steadfast resolve, gists, and technicians, in addition to in 1985. and these experiences spurred her other researchers. Fenske worked in many capacities desire to be a mentor to others, For me, the Fenske Fellowship was during her 27 years of service to the especially women. It was in honor of a challenge. Transitioning into gradu- Fisheries Division. She was a research her commitment, integrity, and memory fisheries technician, field fisheries that the Janice Lee Fenske Excellence in ate school can be difficult on its own, biologist and district fisheries biologist, Fisheries Management Fellowship was and the fellowship added another natural rivers coordinator, Federal created. For more information about layer of commitment. It pushed me Energy Regulatory Commission Project Jan Fenske, visit the Michigan Chapter to develop myself professionally and licensing coordinator, and Tribal unit of the American Fisheries Society to create connections between myself coordinator. Needless to say, Fenske website at www.fisheries.org/units/ and members of the Michigan DNR loved fisheries management and miafs/jfenske_award.html. Fisheries Division. The fellowship provided me with unique lessons and experiences I never would have without the fellowship, ones that have been extremely valuable in preparing me for a career in fisheries management. I also found inspiration in Jan Fenske’s story since she worked tire- lessly to make a difference in fisheries management despite adversity. After everything, I am extremely grateful for the opportunity that the Janice Lee Fenske Excellence in Fisheries Management Fellowship has provided for me. I encourage anyone with an interest in fisheries management to apply for this opportunity; you won’t regret it.

Fisheries • v o l 35 n o 5 • m a y 2010 • w w w .f i s h e r i e s .o r g 249 News: Fisheries Continued from page 214

Connecting Habitats barriers and improving the water qual- mercial viability. The goal of the plan ity of enclosed waterways. Networks of is to reduce contaminated sediments The plan not only includes ways to tributaries connect rivers and streams throughout the estuary by isolating or create and restore habitats, but also to the estuary. Each year, migratory fish removing the contamination. Cleanup methods to connect the habitats so navigate these connections, swim- of contaminated sediments will reduce that they can benefit from one another. ming many miles upstream to spawn. human health and ecological risks, These methods include rebuilding Man-made barriers such as dams can improve fishery resources, remove shorelines and creating habitats for fish, prevent fish from reaching egg-laying fish consumption advisories, and crabs, and lobsters. sites, threatening the future of these provide the public economic benefits. Over the years, the estuary’s 1,000 fish populations. Increasing and improving public access miles of natural shorelines have been The plan recommends removing throughout the estuary is also a goal replaced with piers, docks, and bulk- unnecessary barriers and reconstructing of the plan, with the hope of enabling heads which have altered the naturally others to include fish passage such as everyone in the region to be able to sloped shorelines which transition from fish ladders that can connect upstream reach the estuary in a 20-minute walk shallow to deep water and are needed habitats with the rest of the estuary. or by public transportation. by fish and sea life to thrive. The plan In some areas of the estuary, bodies of suggests replacing abandoned piers water are isolated or enclosed, such Project Goals with naturally sloped shorelines and as dead-end canals. Often these areas creating new piers and other shore collect pollution discharge and storm- Just some of the plan’s short and structures that will be designed in a water runoff, resulting in water that long-term goals include creating 1,200 way to have less of an impact on the is polluted, stagnant, contains sparse acres of coastal wetlands by the year natural shoreline and foster habitat vegetation, has low species diversity, 2015 and 15,200 by 2050, enhancing complexity. Regional fish, crabs, and and emits noxious odors. four or more islands for waterbirds by lobsters require many different habitats 2015 and all islands by 2050, creating or to breed and for their young to develop Health and Societal Value restoring 250 acres of coastal and mari- into full maturity. time forest by 2015 and 1,000 by 2050, The plan includes ways that millions and creating 500 acres of oyster reefs Habitat Support Structures of people can safely enjoy access to habitats by 2015 and 5,000 by 2050. the estuary. For centuries, the estu- “The team is already achieving the The plan also focuses on the physical ary has been an area for disposal of plan’s restoration goals,” said Weppler. landscape, balancing necessary urban chemical and industrial wastes. These “Salt marshes are being created within infrastructure with environmental resto- contaminated sediments are harmful some of the estuary’s regional areas. ration to include opening up tributaries to the estuary’s wildlife, pose a public This is tangible evidence towards fulfill- that may be obstructed by man-made health risk, and reduce the port’s com- ing our vision of a restored estuary.”

Water Quality Meters Scales Fish Measuring Boards Stealth Boats Data Loggers Nesting Boxes

And have for more than 60 years. Our 720+ page catalog and easy-to- Yeah. use web site features thousands of products, and we still back each one with our 100% Satisfaction Guarantee. Call 800-360-7788 or go online Forestry Suppliers We do that. to order your FREE catalog today.

FORESTRY SUPPLIERS

Fisheries_030110.pdf 1 3/1/10 1:51 PM 250 Fisheries • v o l 35 n o 5 • m a y 2010 • w w w .f i s h e r i e s .o r g You are 1 degree of separation from changing your world. Which 1 will it be?

76 degrees of distinction – delivered 100% online, including undergraduate and graduate degrees in environmental studies with concentrations in:

Environmental Policy Environmental Technology & Mgmt. American Public University Environmental Sustainability Fish & Wildlife Mgmt. Global Environmental Mgmt. Regional & Community Environmental Mgmt.

At APU, your degree is within reach. Emphasizing affordable excellence with in-state tuition rates, wherever you are.

Let us help you get started today. 1.877.777.9081 • www.apu.apus.edu/environmental-studies

Respected. Affordable. Online.

Floy Tag

Fisheries • v o l 35 n o 5 • m a y 2010 • w w w .f i s h e r i e s .o r g 251 12–16 S e p t. www. f isheries.org American Fisheries Society 140th Annual Meeting

Continuing Education Program

Introduction to Instream Instream Structures for Habitat Habitat Modeling using Enhancement MesoHABSIM Wolfgang Wolter, AECOM; Wolfgang. he Continuing Piotr Paraiewicz and Josef N. [email protected]; and John Parish, Education Committee Rogers, Rushing Rivers Institute; piotr@ PARISH Geomorphic Ltd; jparish@ T has put together a rushingrivers.org, [email protected] parishgeomorphic.com) diverse suite of courses for Student $150; Member $250; Student $75; Member $125; the 2010 Annual Meeting. Non-member $300 Non-member $175 The courses cover a variety This two-day course will serve as an Topics include an introduction to of topics, ranging from the introduction to modeling instream habitat stream restoration, benefits of stream using the MesoHABSIM approach (and restoration, adaptive environmental responsibilities of fisheries management, use of instream structures professionals in Endangered associated SimStream software) and how it can be applied to river restoration and for habitat enhancement, and design and Species Act consultations, to management. implementation of instream structures, building leadership skills in including working examples. A case AFS, to complex modeling Basic/Intermediate GIS for study will be presented that will examine of instream habitat for river Fisheries Biologists the physical, biological, and technical management. When you Joanna Whittier, Kansas State University; aspects of channel and structure design register for the meeting, [email protected] with a focus on fish habitat. Material will be presented in lecture format with Student $125; Member $220; please consider taking one encouragement of participation from Non-member $250 or more of these courses or students. technology workshops. All of This course will provide an overview of basic/intermediate GIS skills for fisheries them will help increase your Introduction to Programming in biologists using ArcGIS, including use of R for Fisheries Scientists professionalism, perspective, existing data, creating your own data, and Matt Catalano and Cheryl Murphy, and skill set when you return review of fundamental concepts for GIS. to your job. Michigan State University; mcatalano@ msu.edu, [email protected] Advanced GIS for Fisheries Student $100; Member $150; Biologists Non-member $200 Joanna Whittier, Kansas State University; This course will introduce the basics of [email protected] Program R using a command-line interface Student $150; Member $220; and examples from fisheries research. Non-member $270 Program R is a powerful open-source Building on the “Basic/Intermediate GIS mathematical and statistical software for Fisheries Biologists” course, this course program gaining popularity in the fisheries will focus on geoprocessing, interpolation, and ecological sciences. and spatial analysis methods to aid in fisheries monitoring and research. NEW! Mapping Habitat of Aquatics Systems using Low-Cost Leadership at All Levels in AFS Side-Scan Sonar Dirk Miller, Wyoming Game and Fish Full-Day Course Department; [email protected] Adam J. Kaeser and Thom Litts, Georgia FREE! Department of Natural Resources; Adam. This course will focus on helping AFS [email protected], Thom.litts@dnr. leaders understand how to work effectively state.ga.us within the AFS governance structure at all Student $100; Member $150; levels: Chapter, Section, and Division. Non-member $200

252 Fisheries • v o l 35 n o 5 • m a y 2010 • w w w .f i s h e r i e s .o r g This course is an introduction to using quantitative analysis of fish populations. Plenary Session Brings the inexpensive Humminbird® Side The program requires no experience in Together Various Imaging system to map and quantify modeling, but a basic understanding of fish Perspectives on Fisheries benthic habitats at the landscape scale. mortality and growth (e.g., von Bertalanffy Profession The course includes a practical session growth model) would be helpful. covering techniques for geoprocessing Following the theme, “Merging Our sonar imagery and map development NEW! River Morphology and Deeper Currents,” the Plenary Session will within ArcGIS 9.x. Restoration explore the common denominators and core values of the fisheries profession Jim Gracie, Brightwater, Inc; jgracie@ from science, government, conservation, NEW! Science, Tools and brightwaterine.com and emerging cohort perspectives. Information Resources for Student $150; Member $250; Downstream Fish Passage Non-member $300 Ian Cowx is director of Doug Dixon, Bioengineering Section AFS; the University of Hull This two-day course covers general International Fisheries [email protected] principles of river morphology, classification Institute (HIFI) in Hull, Student $75; Member $125; systems, hydrology and hydraulics, stream UK. In his talk “Fisheries Non-member $175 stability, natural channel restoration and the Science This course will provide attendees approaches, and a description of Imperative,” he will examine the shifting with state-of-the-art knowledge of basic stabilization devices, habitat improvement and applied fisheries bioengineering position and roles of devices, and performance monitoring. science (in its widest information associated with downstream The material is presented with real-life sense), people, and passage for fish at dams, primarily those examples and includes extensive class emerging environmental associated with hydroelectric power participation and problem solving. and sustainability issues in the management generation and navigation. The program of fisheries. will focus on the science of downstream NEW! Endangered Species Act Jane Lubchenco, passage including the reasons for providing Consultations and the Role of a marine ecologist passage, the physics of water flow, fish Fisheries Professionals and environmental biology, fish way design and evaluation, Margaret Murphy, Michael Schiewe scientist, was sworn in and criteria and processes for fishway as the ninth and first and Paul Schlenger, ANCHOR QEA, LLC; implementation—as required in both woman administrator [email protected], MSchiewe@ Canada and the United States. of the National Oceanic anchorqea.com, PSchlenger@anchorqea. and Atmospheric com Administration (NOAA) NEW! Applying FishXing in Fish Student $40; Member $70; in March 2009. Her talk Passage Assessment and Design Non-member $100 is titled, “Fisheries in the at Road-Stream Crossings Current: The Role of Fisheries in the Nation’s With the recent expansion of the Margaret Lang and Mike Love, New Ocean Policy.” ESA listing for the Gulf of Maine Atlantic Bioengineering Section AFS; Margaret. Larry Schweiger is [email protected], mlove@h2odesigns. salmon, the numerous listings of salmonids on the West Coast, and many freshwater president and chief com executive officer of Student $60; Member $100; listings, fisheries professionals have the National Wildlife Non-member $150 important roles in recovery planning, Federation. In his This half-day course will present ESA Section 7 consultations, and Habitat talk he will address opportunities for participants with an introduction to the Conservation Plan (HCP) development. This program will introduce participants synergism among FishXing software and an overview of conservation how to use the software for assessment to the ESA and explain the roles and organizations, scientists, and design of fish and aquatic organism- responsibilities of fisheries professionals in and government friendly road crossings. This course will supporting species recovery. agencies to achieve specific goals. be of particular interest to engineers and Melissa Wuellner is an biologists responsible for assessment, NEW! AFS Technology Workshops: assistant professor and planning, or design of road crossings on VEMCO Acoustic Telemetry the distance education natural waterways. Technology Workshop coordinator in the Nancy Edwards (VEMCO Division and Department of Wildlife NEW! Basic Fish Population AMIRIX Systems Inc.; nancy.edwards@ and Fisheries Sciences amirix.com) at South Dakota State Modeling using Excel University. She was Mike Allen, University of Florida; msal@ FREE! recently recognized with ufl.edu This course will discuss VR2 (VR3 the AFS Distinguished Student $75; Member $125; & VR4) passive and active automated Service Award for her Non-member $175 acoustic receiver technology, designed and efforts at the Chapter, Section, and Society This course is intended to teach basic produced by VEMCO, and its application to levels. Her talk is titled “The Emerging Cohort of Fisheries Professionals: Are We Really So fish population modeling skills to students assess movement patterns, behavior, and Different from the Older Generations?” and professionals who are interested in site fidelity of fishes and invertebrates.

Fisheries • v o l 35 n o 5 • m a y 2010 • w w w .f i s h e r i e s .o r g 253 To submit upcoming events for inclusion on the AFS Web site Calendar, send CALENDAR: event name, dates, city, state/province, web address, and contact information FISHERIES EVENTS to [email protected]. (If space is available, events will also be printed in Fisheries magazine.)

More events listed at www.fisheries.org.

May 23-26 Australasian Aquaculture International Hobart, Tasmania www.australian-aquacultureportal.com Conference and Trade Show

May 30- AFS Early Life History Section’s 34th Annual Santa Fe, New Mexico www.larvalfishcon.org Jun 3 Larval Fish Conference

Jun 16-18 Offshore Mariculture Conference Dubrovnik, Croatia www.mercatormedia.com

Jun 20-22 Second International Catfish Symposium St. Louis, Missouri www.catfish2010.org sponsored by AFS North Central and Southern Divisions

Jun 21-24 International Symposium on Genetic Biocontrol Minneapolis, Minnesota www.seagrant.umn.edu/ais/biocontrol of Invasive Fish

Jul 7-12 Joint Meeting of Ichthyologists and Providence, Rhode Island www.dce.ksu.edu/conf/jointmeeting Herpetologists

Jul 24-28 Second International Sclerochronology Mainz, Germany www.scleroconferences.de Conference

Jul 25-30 Fisheries Society of the British Isles Conference: Belfast, Northern Ireland www.fsbi.org.uk/events.htm. Climate Change and Fish

Aug 1-6 95th Annual Meeting of the Ecological Society Pittsburgh, Pennsylvania www.esa.org/pittsburgh of America

Aug 15-20 Second International Conference on the Effects Cork, Ireland www.aquaticnoise.org of Noise on Aquatic Animals

Aug 31- Third Annual Conference of the North American Chico Spring, Montana www.wscs.info Chapter of World Sturgeon Conservation Society Sep 2

Sep 8-11 Fish Sampling with Active Methods Meeting Ceske Budejovice, Czech www.fsam2010.wz.cz Republic

Sep 12-16 American Fisheries Society 140th Annual Pittsburgh, Pennsylvania www.fisheries.org/afs10/ Meeting

Sep 22 World Ocean Council: Sustainable Ocean Summit Honolulu, Hawaii www.oceancouncil.org

Sep 22-23 Electrofishing Class Vancouver, Washington www.smith-root.com

Ssep 20-24 ICES Annual Science Conference 2010 Cite des Congres, Nantes, www.ices.dk/iceswork/asc/2010/index.asp France

Sep 27-28 Fourth International Natural Channel Systems Mississauga, Ontario, www.naturalchannels.ca Conference: Stream Corridors: Restoring Our Canada Natural Infastructure

Sep 28-30 Wild Trout Symposium West Yellowstone, Montana www.wildtroutsymposium.com

254 Fisheries • v o l 35 n o 5 • m a y 2010 • w w w .f i s h e r i e s .o r g Oct 3-8 Aquatic Resources Education Association Biennial Omaha, Nebraska www.areanet.org Conference

Oct 10-14 Integrating Biogeochemistry and Ecosystems in a crete, Greece http://imbizo-2010.confmanager.com Changing Ocean: Regional Comparisons

Nov 8-11 Alaska Sea Grant Meeting: Ecosystems 2010 Anchorage, Alaska http://seagrant.uaf.edu/conferences/2010/ Lowell Wakefi eld Fisheries Symposium:Global wakefi eld-ecosystemb/index.php Progress on Ecosystem-based Fisheries Management

Dec 1-2 12th Flatfi sh Biology Conference Westbrook, Connecticut www.mi.nmfs.gov/fl atfi shbiologyworkshop. html

Dec 10-13 Fifth Shanghai International Fisheries and shanghai, china www.sifse.com Seafood Exposition—The Best Opportunity fo Explore Chinese Market

Halltech Aquatic Research, Inc.

Fisheries • v o l 35 n o 5 • m a y 2010 • w w w .f i s h e r i e s .o r g 255 EMPLOYERS: To list a job opening on the AFS Online Job Center submit a position Announcements: description, job title, agency/company, city, state, responsibilities, qualifications, salary, closing date, and contact information (maximum 150 words) to jobs@fisheries. Job Center org. Online job announcements will be billed at $350 for 150 word increments. Please send billing information. Listings are free (150 words or less) for organizations with Associate, Official, andS ustaining memberships, and for Individual members, who are faculty members, hiring graduate assistants. If space is available, jobs may also be printed in Fisheries magazine, free of additional charge.

M.S. Graduate Research Assistantship, Illinois Natural Contact: Submit a letter of interest, resume, contact History Survey and the University of Illinois. information for three references, copies of transcripts Salary: approximately $16,000 per year plus tuition and (unofficial are OK) and GRE scores (unofficial are OK) to fee waiver. Greg Sass, Director, Illinois River Biological Station, ggsass@ Closing: 30 May or until filled. illinois.edu; and Cory Suski, at Urbana—Champaign, Dept Responsibilities: Participate in a multidisciplinary study to of Natural Resources and Environmental Sciences, W401-C investigate the impacts of invasive Asian carp on Midwest Turner Hall, MC047, 1102 S. Goodwin Ave. Urbana, Illinois, river ecosystems. Projects will involve investigations 61801; 217/244-2237; fax: 217/244-3219 University of of physiological parameters that influence the spread Illinois, [email protected]. See http://fishlab.nres.uiuc.edu. and abundance of invasive Asian carp, and interactions between invasive Asian carp and native fishes. M.S. or Ph.D. Graduate Assistantship, State University of New York, College of Environmental Science and Forestry, Qualifications: B.S. in fisheries, ecology, biology, or a Department of Environmental and Forest Biology. related field and 3.0 GPA minimum 1100 V Q GRE. Previous Salary: $18,000 per year, tuition waiver, housing available experience with fish collection and fisheries field work is during field season. preferred but not essential. Closing: 31 May 2010 or until filled. Responsibilities: Perform independent research on coolwater fish habitat in the St. Lawrence River. Research will involve an intense field and analytical effort, reporting results in written reports, peer-reviewed publications, and oral presentations at professional meetings. Qualifications: B.S. or M.S. in fisheries, aquatic sciences, or related discipline. Quantitative fisheries or related field and spatial database and survey skills are preferred. Highly qualified individuals with strong GPA and GRE and experience are encouraged to apply. Highly motivated with demonstrated ability to work well in a team environment. Contact: Send application to John M. Farrell, Office of Instruction Graduate Studies, SUNY-ESF One Forestry Drive, 227 Bray Hall, Syracuse, NY 13210. See www.esf.edu/ graduate/admission.htm.

Quantitative Ecologist/Data Analyst, Normandeau Associates, Inc., New Hampshire. Salary: Depends on experience. Closing: 1 June 2010. Responsibilities: Work independently and with project scientists to develop and manage project-specific databases, facilitate quality control/quality assurance of these data, and perform statistical and spatial analysis. Qualifications: M.S. in applied statistics or biostatistics, with supporting formal training in ecology, fisheries, and Federal Energy Regulatory Commission aquatic ecology, or a M.S. in ecology, fisheries, aquatic ecology with supporting formal training in applied statistics

256 Fisheries • v o l 35 n o 5 • m a y 2010 • w w w .f i s h e r i e s .o r g and biostatistics, and a minimum of three years experience, models, mark-recapture techniques, and SAS programming. preferably in quantitative ecology. Demonstrated Ability to work in U.S. without company sponsorship. knowledge of analysis software, including but not limited Ability to design and conduct successful field fisheries to SAS, MATLAB, GIS and the application of this software programs. Excellent report writing and communication to describe patterns in biological and environmental data skills. and test hypotheses concerning these data. Contact: Submit cover letter and resume to HR@ Contact: Submit cover letter and resume to HR@ normandeau.com. normandeau.com, [email protected]. ZebrafishR esearch Technician—Fish Facility Fisheries Scientist, Normandeau Associates, Inc., Salary: Maintanance, New York University, New York, New York. Depends on experience. Salary: $37,500 per year. Closing: 1 June 2010. Closing: 1 June 2010. Responsibilities: Support senior staff in population Start date: 1 March to 1 June 2010. monitoring and impact assessments responsible for Responsibilities: Two-year commitment. Maintain a program implementation, quality control, data analysis and interpretation. Travel to, and participate in, field fisheries zebrafish aquaculture facility of 3,500 tanks. Breed and rear programs. Occasional overtime, weekend, or evening work zebrafish; maintain, and repair the water system; and assist may be required to meet deadlines. with experiments investigating the genetic regulation of Qualifications: M.S. in biology, zoology, fisheries, or zebrafish development. related discipline with 1–3 years of experience and a Qualifications: B.S. or M.S. in biology, fisheries, or broad background in freshwater or marine fish population related fields obtained in the last five years. Aquaculture ecology, with emphasis on quantitative field applications. experience. Excellent time management, leadership, Specific experience is desired in one or more of the verbal, and written communications skills. Plumbing and/or following areas: sampling design statistics, population electrical hands on knowledge and/or research experience. Sonotronics

Emperor Aquatiacs, Inc.

Fisheries • v o l 35 n o 5 • m a y 2010 • w w w .f i s h e r i e s .o r g 257 Contact: Send to [email protected] using the Starting date: 1 January 2011. subject line: Zebrafish technician: Your name, include Responsibilities: Participate in Oklahoma Biological Survey your name, address, degree date awarded, and GPA. Also Monitoring and research on zebra mussels, Dreissena include name, e-mail, and phone number of 3 references polymorpha on Lake Texoma, Oklahoma. Research topic and state in which capacity these references know you. will likely be directed toward local life history, expansion, NYU School of Medicine offers an excellent salary and and management of zebra mussels. benefits package including four weeks paid vacation and Qualifications: B.S. in zoology or biology with preference tuition remission at NYU. for emphasis on aquatic ecology or related discipline. Contact: Send cover letter, resume, copies of transcripts M.S. Graduate Assistantship in Fisheries and Aquatic (unofficial OK), GRE scores (mandatory) and three letters Ecology, Oklahoma Deparatment of Wildlife Conservation, of reference to Greg Summers, Oklahoma Fishery Research University of Oklahoma. Laboratory, 500 E. Constellation, Norman, Oklahoma; Salary: $1,545 per month plus tuition waiver. 73072; [email protected]; 405/325-7288. See Closing: 31 July 2010. www.ou.edu/cas/zoology/graduate_degrees.htm.

5 May

258 Fisheries • v o l 35 n o 5 • m a y 2010 • w w w .f i s h e r i e s .o r g Hydroacoustic Technology, Inc.

Fisheries • v o l 35 n o 5 • m a y 2010 • w w w .f i s h e r i e s .o r g 259 HTI 260 Fisheries • v o l 35 n o 5 • m a y 2010 • w w w .f i s h e r i e s .o r g