Northern Mexican Gartersnake (Thamnophis eques megalops) (NMGS) Basic Conceptual Ecological Model for the Lower Colorado River

LCR MSCP website photo gallery (Jeneal Smith, Bureau of Reclamation)

April 2021

Work conducted under LCR MSCP Work Task G6 Lower Colorado River Multi-Species Conservation Program Steering Committee Members

Federal Participant Group California Participant Group

Bureau of Reclamation California Department of Fish and Wildlife U.S. Fish and Wildlife Service City of Needles National Park Service Coachella Valley Water District Bureau of Land Management Colorado River Board of California Bureau of Indian Affairs Bard Water District Western Area Power Administration Imperial Irrigation District Los Angeles Department of Water and Power Palo Verde Irrigation District Arizona Participant Group San Diego County Water Authority Southern California Edison Company Arizona Department of Water Resources Southern California Public Power Authority Arizona Electric Power Cooperative, Inc. The Metropolitan Water District of Southern Arizona Game and Fish Department California Arizona Power Authority Central Arizona Water Conservation District Cibola Valley Irrigation and Drainage District Nevada Participant Group City of Bullhead City City of Lake Havasu City Colorado River Commission of Nevada City of Mesa Nevada Department of Wildlife City of Somerton Southern Nevada Water Authority City of Yuma Colorado River Commission Power Users Electrical District No. 3, Pinal County, Arizona Basic Water Company Golden Shores Water Conservation District Mohave County Water Authority Mohave Valley Irrigation and Drainage District Native American Participant Group Mohave Water Conservation District North Gila Valley Irrigation and Drainage District Hualapai Tribe Town of Fredonia Colorado River Indian Tribes Town of Thatcher Chemehuevi Indian Tribe Town of Wickenburg Salt River Project Agricultural Improvement and Power District Unit “B” Irrigation and Drainage District Conservation Participant Group Wellton-Mohawk Irrigation and Drainage District Yuma County Water Users’ Association Ducks Unlimited Yuma Irrigation District Lower Colorado River RC&D Area, Inc. Yuma Mesa Irrigation and Drainage District The Nature Conservancy

Other Interested Parties Participant Group

QuadState Local Governments Authority Desert Wildlife Unlimited

Lower Colorado River Multi-Species Conservation Program

Northern Mexican Gartersnake (Thamnophis eques megalops) (NMGS) Basic Conceptual Ecological Model for the Lower Colorado River

Prepared by: Elizabeth A. Johnson and Robert Unnasch, Ph.D. Sound Science, LLC

Lower Colorado River Multi-Species Conservation Program Bureau of Reclamation Lower Colorado Basin Boulder City, Nevada http://www.lcrmscp.gov April 2021

Johnson, E.A. and R. Unnasch. 2021. Northern Mexican Gartersnake (Thamnophis eques megalops) (NMGS) Basic Conceptual Ecological Model for the Lower Colorado River. Submitted to the Lower Colorado River Multi-Species Conservation Program, Boulder City, Nevada, by Sound Science, LLC, Boise, Idaho, under contract No. R16PC00028.

ACRONYMS AND ABBREVIATIONS

AZGFD Arizona Game and Fish Department

CEM conceptual ecological model cm centimeter(s) ft foot/feet

Havasu NWR Havasu National Wildlife Refuge in inch(es)

LCR lower Colorado River

LCR MSCP Lower Colorado River Multi-Species Conservation Program

LLFR lowland leopard frog (Rana = Lithobates yavapaiensis)

m meter(s) mm millimeter(s)

NAU Northern Arizona University

NH3 ammonia

NMGS northern Mexican gartersnake

pH potential of hydrogen – scale used to specify the acidity or basicity of an aqueous solution

Reclamation Bureau of Reclamation

San Bernardino NWR San Bernardino National Wildlife Refuge

SFD snake fungal disease (Ophidiomyces ophiodiicola)

USFWS U.S. Fish and Wildlife Service

Symbols

≈ approximately

> greater than

≥ greater than or equal to

< less than

% percent (aka parts per hundred)

± plus or minus

°C temperature in degrees Celsius

°F temperature in degrees Fahrenheit

Definitions

For the purposes of this document, vegetation layers are defined as follows:

Canopy – The canopy is the uppermost strata within a plant community. The canopy is exposed to the sun and captures the majority of its radiant energy.

Understory – The understory comprises plant life growing beneath the canopy without penetrating it to any extent. The understory exists in the shade of the canopy and usually has lower light and higher humidity levels. The understory includes subcanopy trees and the shrub and herbaceous layers.

Shrub layer – The shrub layer is comprised of woody plants between 0.5 and 2.0 meters in height.

Herbaceous layer – The herbaceous layer is most commonly defined as the forest stratum composed of all vascular species that are 0.5 meter or less in height.

CONTENTS

Page

Foreword ...... v

Executive Summary ...... ES-1 Conceptual Ecological Models ...... ES-1 Conceptual Ecological Model Structure ...... ES-3 Results ...... ES-5

Chapter 1 – Introduction ...... 1-1 Northern Mexican Gartersnake Reproductive Ecology ...... 1-2 Conceptual Ecological Model Purposes ...... 1-3 Conceptual Ecological Model Structure ...... 1-4

Chapter 2 – NMGS Life-Stage Model ...... 2-1 Introduction to the NMGS Life Cycle ...... 2-2 NMGS Life Stage 1 – Neonates/Juveniles ...... 2-2 NMGS Life Stage 2 – Adults ...... 2-2

Chapter 3 – Critical Biological Activities and Processes ...... 3-1 Basking ...... 3-2 Brumation ...... 3-3 Chemical Stress ...... 3-4 Competition...... 3-5 Disease ...... 3-6 Dispersal ...... 3-6 Ecdysis ...... 3-7 Foraging ...... 3-8 Predation ...... 3-9

Chapter 4 – Habitat Elements ...... 4-1 Anthropogenic Disturbance ...... 4-3 Competitors ...... 4-4 Environmental Contaminants...... 4-4 Food Availability ...... 4-6 Genetic Diversity ...... 4-7 Hydrologic Regime ...... 4-8 Infectious Agents ...... 4-8 Predators ...... 4-9 Substrate ...... 4-10 Surface Water Connectivity ...... 4-11 Temperature ...... 4-12 Turbidity ...... 4-13 Vegetation Structure ...... 4-13

i Page

Chapter 5 – Controlling Factors ...... 5-1 Fire Management ...... 5-3 Fisheries Introduction & Fisheries Management ...... 5-4 Grazing ...... 5-5 Nuisance Species Introduction & Management ...... 5-6 Off-Site Land Management & Use ...... 5-8 On-Site Vegetation Management ...... 5-8 On-Site Water Management ...... 5-9 Recreational Activities ...... 5-9 Site Maintenance ...... 5-10 Wastewater & Other Contaminant Inflows ...... 5-10 Water Storage-Delivery System Design & Operations ...... 5-11

Chapter 6 – Conceptual Ecological Model by Life Stage ...... 6-1 NMGS Life Stage 1—Neonates/Juveniles ...... 6-4 NMGS Life Stage 2—Adults ...... 6-9

Chapter 7 – Causal Relationships Across Life Stages ...... 7-1 Effects of Critical Biological Activities and Processes on Life-Stage Outcomes ...... 7-1 Effects of Critical Biological Activities and Processes on Each Other ...... 7-3 Effects of Habitat Elements on Critical Biological Activities and Processes ...... 7-4 Effects of Habitat Elements on Each Other ...... 7-6 Effects of Controlling Factors on Habitat Elements ...... 7-7 Causal Relationships with High Understanding ...... 7-10 Potentially Influential Causal Relationships with Low Understanding and/or Unknown Magnitude ...... 7-12

Chapter 8 – Discussion and Conclusions ...... 8-1

Literature Cited ...... L-1

Acknowledgments ...... A-1

ii

Tables

Table Page

1 NMGS life stages and life-stage outcomes ...... 2-1 2 NMGS critical biological activities and processes and their distribution between life stages ...... 3-1 3N MGS habitat elements and the critical biological activities and processes that are proposed to directly affect them between the two life stages ...... 4-2 4 NMGS controlling factors and the habitat elements they are proposed to directly affect among the two NMGS life stages ...... 5-2 5 Direct effects of critical biological activities and processes on NMGS life-stage outcomes ...... 7-2 6 Direct effects of critical biological activities and processes on each other across the two NMGS life stages ...... 7-3 7 NMGS habitat elements and the critical biological activities and processes they are proposed to directly affect across the two NMGS life stages ...... 7-5 8 Direct effects of habitat elements on each other across the two NMGS life stages ...... 7-6 9 Direct effects of controlling factors on habitat elements across the two NMGS life stages ...... 7-8 10 Direct effects of controlling factors on each other across the two NMGS life stages ...... 7-10

Figures

Figure Page

1 Proposed NMGS life history model...... 2-1 2 Diagram conventions for LCR MSCP species conceptual ecological models...... 6-3 3 CEM master diagram for NMGS life stage 1 – Neonates/juveniles life stage controlling factors, habitat elements, critical biological activities and processes, and life-stage outcome...... 6-7 4 CEM master diagram for NMGS life stage 2 – adult life stage controlling factors, habitat elements, critical biological activities and processes, and life-stage outcomes...... 6-13

iii Attachments

Attachment

1 Species Conceptual Ecological Model Methodology for the Lower Colorado River Multi-Species Conservation Program

2 Northern Mexican Gartersnake (Thamnophis eques megalops) (NMGS) Habitat Data

iv

Foreword

The Lower Colorado River Multi-Species Conservation Program (LCR MSCP) Habitat Conservation Plan requires the creation, and long-term stewardship, of habitat for 20 covered species. This is both an exciting and daunting challenge— exciting, in that success would mean a major conservation achievement in the lower Colorado River landscape, and daunting, in that we need to simultaneously manage our lands for the benefit of 20 species in a mosaic of land cover types. To do so, we need to develop a common understanding of the habitat requirements of each species and the stewardship required to meet those needs.

To provide a framework to capture and share the information that forms the foundation of this understanding, conceptual ecological models (CEMs) for each covered species have been created under the LCR MSCP’s Adaptive Management Program. The LCR MSCP’s conceptual ecological models are descriptions of the functional relationships among essential components of a species’ life history, including its habitat, threats, and drivers. They tell the story of “what’s important to the animal” and how our stewardship and restoration actions can change those processes or attributes for the betterment of their habitat. As such, CEMs can provide:

• A synthesis of the current understanding of how a species’ habitat works. This synthesis can be based on the published literature, technical reports, or professional experience.

• Help in understanding and diagnosing underlying issues and identifying land management opportunities.

• A basis for isolating cause and effect and simplifying complex systems. These models also document the interaction among system drivers.

• A common (shared) framework or “mental picture” from which to develop management alternatives.

• A tool for making qualitative predictions of ecosystem responses to stewardship actions.

• A way to flag potential thresholds from which system responses may accelerate or follow potentially unexpected or divergent paths.

• A means by which to outline further restoration, research, and development and to assess different restoration scenarios.

v

• A means of identifying appropriate monitoring indicators and metrics.

• A basis for implementing adaptive management strategies.

Most natural resource managers rely heavily upon CEMs to guide their work, but few explicitly formulate and express the models so they can be shared, assessed, and improved. When this is done, these models provide broad utility for ecosystem restoration and adaptive management.

Model building consists of determining system parts, identifying the relationships that link these parts, specifying the mechanisms by which the parts interact, identifying missing information, and exploring the model’s behavior (Heemskerk et al. 20031). The model building process can be as informative as the model itself, as it reveals what is known and what is unknown about the connections and causalities in the systems under management.

It is important to note that CEMs are not meant to be used as prescriptive management tools but rather to give managers the information needed to help inform decisions. These models are conceptual and qualitative. They are not intended to provide precise, quantitative predictions; rather, they allow us to virtually “tweak the system” free of the constraints of time and cost to develop a prediction of how a system might respond over time to a variety of management options; for a single species, documented models are valuable tools, but for 20 species, they are imperative. The successful management of multiple species in a world of competing interests (species versus species); potentially conflicting needs, goals, and objectives; long response times; and limited resources depends on a CEM to help land managers experiment from the safety of the desktop. Because quantitative data can be informative, habitat parameters that have been quantified in the literature are presented (attachment 2) in this document for reference purposes.

These models are intended to be “living” documents that should be updated and improved over time. The model presented here should not be viewed as a definitive monograph of a species’ life history but rather as a framework for capturing the knowledge and experience of the LCR MSCP’s scientists and land stewards. While ideally the most helpful land management tool would be a definitive list of do’s and don’ts, with exact specifications regarding habitat requirements that would allow us to engineer exactly what the species we care about need to survive and thrive, this is clearly not possible. The fact is, that despite years of active management, observation, and academic research on many of the LCR MSCP species of concern, there may not be enough data to support developing such detailed, prescriptive land management.

1 Heemskerk, M., K. Wilson, and M. Pavao-Zuckerman. 2003. Conceptual models as tools for communication across disciplines. Conservation Ecology 7(3):8: https://www.ecologyandsociety.org/vol7/iss3/art8/

vi

The CEMs for species covered under the LCR MSCP are based on, and expand upon, methods developed by the Sacramento- San Joaquin Delta Ecosystem Restoration Program (ERP): https://www.dfg.ca.gov/ERP/conceptual_models.asp. The ERP is jointly implemented by the California Department of Fish and Wildlife, U.S. Fish and Wildlife Service, and the National Marine Fisheries Service. The Bureau of Reclamation (Reclamation) participates in this program. (See attachment 1 for an introduction to the CEM process.)

Many of the LCR MSCP covered species are migratory. These models only address the species’ life history as it relates to the lower Colorado River and specifically those areas that are potentially influenced by LCR MSCP land management. The models DO NOT take into account ecological factors that influence the species at their other migratory locations.

Finally, in determining the spatial extent of the literature used in these models, the goals and objectives of the LCR MSCP were taken into consideration. For species whose range is limited to the Southwest, the models are based on literature from throughout the species’ range. In contrast, for those species whose breeding range is continental (e.g., yellow-billed cuckoo [Coccyzus americanus occidentalis]) or west-wide, the models primarily utilize studies from the Southwest.

How to Use the Models

There are three important elements to each CEM:

(1) The narrative description of the species’ various life stages, critical biological activities and processes, and associated habitat elements.

(2) The figures that provide a visual snapshot of all the critical factors and causal links for a given life stage.

(3) The associated workbooks. Each CEM has a workbook that includes a worksheet for each life stage.

This narrative document is a basic guide, meant to summarize information on the species’ most basic habitat needs; the figures are a graphic representation of how these needs are connected; and the accompanying workbook is a tool for land managers to see how on-the-ground changes might potentially change outcomes for the species in question. Reading, evaluating, and using these CEMs requires that the reader understand all three elements; no single element provides all the pertinent information in the model. While it seems convenient to simply read the narrative, we strongly recommend the reader have the figures and workbook open and refer to them while reviewing this document.

vii

It is also tempting to see these products, once delivered, as “final.” However, it is more accurate to view them as “living” documents, serving as the foundation for future work. Reclamation will update these products as new information is available, helping to inform land managers as they address the on-the-ground challenges inherent in natural resource management.

The knowledge gaps identified by these models are meant to serve only as an example of the work that could be done to further complete our understanding of the life history of the LCR MSCP covered species. However, this list can in no way be considered an exhaustive list of research needs. Additionally, while identifying knowledge gaps was an objective of this effort, evaluating the feasibility of addressing those gaps was not. Finally, while these models were developed for the LCR MSCP, the identified research needs and knowledge gaps reflect a current lack of understanding within the wider scientific community. As such, they may not reflect the current or future goals of the LCR MSCP. They are for the purpose of informing LCR MSCP decision making but are in no way meant as a call for Reclamation to undertake research to fill the identified knowledge gaps.

John Swett, Program Manager, LCR MSCP Bureau of Reclamation September 2015

viii

Executive Summary

This document presents a conceptual ecological model (CEM) for the northern Mexican gartersnake (Thamnophis eques megalops) (NMGS). The purpose of this model is to help the Lower Colorado River Multi-Species Conservation Program (LCR MSCP), Bureau of Reclamation, identify areas of scientific uncertainty concerning NMGS ecology, the effects of specific stressors, the effects of specific management actions aimed at species habitat restoration, and the methods used to measure NMGS habitat and population conditions. The CEM methodology follows that developed for the Sacramento-San Joaquin River Delta Regional Ecosystem Restoration Implementation Plan (DiGennaro et al. 2012), with modifications. (Note: Attachment 1 provides an introduction to the CEM process. We recommend that those unfamiliar with this process read the attachment before continuing with this document.)

The CEM addresses the present NMGS population and its distribution within the greater lower Colorado River (LCR) ecosystem along the main stem river and within the Bill Williams River watershed. Currently, NMGS populations are known to exist at the Beal Lake Conservation Area, Topock Marsh, and locations in the Bill Williams River watershed on the Bill Williams River, Big Sandy River, and Santa Maria River (O’Donnell et al. 2019; Sabin 2018). Since the species was rediscovered on the river in 2015, surveys have only been conducted at six locations at the Havasu National Wildlife Refuge around Topock Marsh, including the Beal Lake Conservation Area willow marsh. Unsurveyed populations and potential habitat may exist elsewhere in the LCR. Historically, NMGS were recorded from Reaches 6–7 in Yuma, Arizona (circa 1889), and in Reach 3 in Ft. Mohave (circa 1904) (Holycross et al. 2006; C. Ronning 2020, personal communication; Rosen and Schwalbe 1988). The model addresses the landscape as a whole rather than any single managed area.

The research questions and gaps in scientific knowledge identified through the modeling effort serve as examples of topics the larger scientific community could explore to improve the overall understanding of the ecology and conservation of NMGS. These research questions and knowledge gaps may or may not be relevant to the goals of the LCR MSCP. As such, they are not to be considered guidance for the Bureau of Reclamation or the LCR MSCP, nor are these knowledge gaps expected to be addressed under the program.

CONCEPTUAL ECOLOGICAL MODELS

CEMs integrate and organize existing knowledge concerning: (1) what is known about an ecological resource, with what certainty, and the sources of this information, (2) critical areas of uncertain or conflicting science that demand

ES-1 Northern Mexican Gartersnake (Thamnophis eques megalops) (NMGS) Basic Conceptual Ecological Model for the Lower Colorado River

resolution to better guide management planning and action, (3) crucial attributes to use while monitoring system conditions and predicting the effects of experiments, management actions, and other potential agents of change, and (4) how we expect the characteristics of the resource to change as a result of altering its shaping/controlling factors, including those resulting from management actions.

The CEM applied to NMGS expands on the methodology developed for the Sacramento-San Joaquin River Delta Regional Ecosystem Restoration Implementation Plan (DiGennaro et al. 2012). The model distinguishes the major life stages or events through which the individuals of a species must pass to complete a full life cycle. It then identifies the factors that shape the likelihood that individuals in each life stage will survive to the next stage in the study area and thereby shapes the abundance, distribution, and persistence of the species in that area.

Specifically, the NMGS conceptual ecological model has five core components:

• Life stages – These consist of the major growth stages and critical events through which the individuals of a species must pass in order to complete a full life cycle.

• Life-stage outcomes – These consist of the biologically crucial outcomes of each life stage, including the number of individuals surviving to the next life stage (e.g., from juvenile to adult), and the number of offspring produced (fertility rate). The rates of the outcomes for an individual life stage depend on the rates of the critical biological activities and processes for that life stage.

• Critical biological activities and processes – These consist of the activities in which the species engages and the biological processes that take place during each life stage that significantly affect its life-stage outcome rates. Examples of activities and processes for a reptile species may include basking, brumation, and foraging. Critical biological activities and processes typically are “rate” variables.

• Habitat elements – These consist of the specific habitat conditions, the quality, abundance, and spatial and temporal distributions of which significantly affect the rates of the critical biological activities and processes for each life stage. These effects on critical biological activities and processes may be either beneficial or detrimental. Taken together, the suite of natural habitat elements for a life stage is called the “habitat template” for that life stage. Defining the natural habitat template may involve estimating specific thresholds or ranges of suitable values for particular habitat elements outside of which one or more critical biological activities or processes no longer fully support desired life-stage outcome rates—if the state of the science supports such estimates.

ES-2 Executive Summary

• Controlling factors – These consist of environmental conditions and dynamics—including human actions—that determine the quality, abundance, and spatial and temporal distributions of important habitat elements. Controlling factors are also called “drivers.” There may be a hierarchy of such factors affecting the system at different scales of time and space (Burke et al. 2009). For example, the availability of suitable habitat for a riparian reptile may depend on factors such as local hydrology, substrate, and vegetation structure, which in turn may depend on factors such as the water storage-delivery system design and operation (dam design, reservoir morphology, and dam operations), which in turn is shaped by climate, land use, vegetation, water demand, and watershed geology.

This CEM identifies the causal relationships among these components for each life stage. A causal relationship exists when a change in one condition or property of a system results in a change in some other condition or property. A change in the first condition is said to cause a change in the second condition. The CEM method applied here assesses four variables for each causal relationship: (1) the character and direction of the effect, (2) the magnitude of the effect, (3) the predictability (consistency) of the effect, and (4) the certainty of a present scientific understanding of the effect. CEM diagrams and a linked spreadsheet tool document all information on the model components and their causal relationships. Software tools developed specifically for the LCR MSCP’s conceptual ecological models allow users to query the CEM spreadsheet for each life stage and generate diagrams that selectively display query results concerning the CEM for each life stage.

CONCEPTUAL ECOLOGICAL MODEL STRUCTURE

The information used to construct this CEM includes the most recent LCR MSCP species account (Sabin 2018). This publication summarizes and cites large bodies of earlier studies. Where appropriate and accessible, those earlier studies are directly cited. The CEM also integrates information from the U.S. Fish and Wildlife Service (USFWS) listing and critical habitat designation reports (USFWS 2014, 2020); research by the Arizona Division of Game and Fish (Bourne 2020; O’Donnell et al. 2019); websites and recent journal articles; studies of NMGS in other areas outside the LCR (Boyarski 2008, 2011, 2012, 2013 each in USFWS 2014; Boyarski et al. 2015 in Emmons et al. 2016a; Emmons 2017; Emmons and Nowak 2013, 2016 in Northern Arizona University 2020; Emmons et al. 2016a, 2016b; Nowak and Boyarski 2012; Nowak et al. 2011, 2014, 2019; Rosen and Schwalbe 1988; Sprague and Bateman 2018; and Young and Boyarski 2012, 2013); general works on the types of aquatic and adjacent upland southwestern settings in which NMGS can occur; and the expert knowledge of LCR MSCP biologists.

ES-3 Northern Mexican Gartersnake (Thamnophis eques megalops) (NMGS) Basic Conceptual Ecological Model for the Lower Colorado River

However, the purpose of this report is not to provide an updated literature review. Rather, its purpose is to integrate the available information and knowledge into a CEM so it can be used for adaptive management.

The NMGS conceptual ecological model identifies two life stages based on the aforementioned sources of information. Further, the CEM identifies two or more life-stage outcomes for each life stage as follows:

• Neonates/Juveniles: neonate/juvenile survival • Adults: adult survival and adult fertility

Chapter 2 defines and discusses these life stages and life-stage outcomes in detail. The CEM includes growth as a life-stage outcome for both life stages based on the extensive literature on snake biology, indicating that the rates of growth and attained sizes in both life stages affect survival rates, adult mating success, and fertility.

The NMGS conceptual ecological model identifies nine critical biological activities and processes that affect one or more of these life-stage outcomes. Chapter 3 defines and discusses these critical biological activities and processes in detail. The nine critical biological activities and processes are as follows, in alphabetical order: basking, brumation, chemical stress, competition, disease, dispersal, ecdysis, foraging, and predation.

The NMGS conceptual ecological model distinguishes 13 habitat elements that affect the rates, timing, magnitude, distribution, or other aspects of 1 or more critical biological activities or processes for 1 or more life stages. Chapter 4 defines and discusses these habitat elements in detail. The 13 habitat elements are as follows, in alphabetical order: anthropogenic disturbance, competitors, environmental contamination, food availability, genetic diversity, hydrologic regime, infectious agents, predators, substrate, surface water connectivity, temperature, turbidity, and vegetation structure.

Finally, the NMGS conceptual ecological model distinguishes 11 controlling factors that affect the distribution, quality, composition, abundance, and other features of 1 or more of these habitat elements. Because the LCR ecosystem is highly regulated, the controlling factors almost exclusively concern human activities. Chapter 5 defines and discusses these controlling factors in detail. The 11 controlling factors are as follows, in alphabetical order: fire management, fisheries introduction & fisheries management, grazing, nuisance species introduction & management, off-site land management & use, on-site vegetation management, on-site water management, recreational activities, site maintenance, wastewater & other contaminant inflows, and water storage-delivery system design & operation.

ES-4 Executive Summary

RESULTS

The CEM identifies the following direct, strong (high-magnitude) causal relationships among these controlling factors, habitat elements, critical biological activities and processes, and life-stage outcomes:

• Three controlling factors—nuisance species introduction & management, on-site water management, and water storage-delivery system design and operation—have direct, high-magnitude effects on one or more habitat elements relevant to one or both NMGS life stages. These relationships are of high or medium understanding.

The CEM identifies the following controlling factors, habitat elements, and critical biological activities and processes that affect at least five other components:

• One habitat element, hydrologic regime, has direct, medium-magnitude effects on at least five other components (four habitat elements and/or one critical biological activity or process). Four out of five of the effects of hydrologic regime on other model components are very well understood, and the effect of the hydrologic regime on predators is moderately well understood.

• One habitat element—anthropogenic disturbance— has direct effects on six other components (three habitat elements and three critical biological activities or processes). While considered of low magnitude, most of these relationships are of low understanding, and further study may modify the magnitude of some of these relationships.

• Two habitat elements—substate and vegetation structure—each have direct effects on five other components (critical biological activities or processes and/or habitat elements). While most relationships are considered of low magnitude (except the vegetation structure’s effects on predation, which is of medium magnitude), they are all moderately well understood. In particular, the fact that these habitat elements affect so many critical biological activities or processes (three each) highlights the importance of substrate and vegetation structure to NMGS.

• In addition to the high-magnitude effects that nuisance species introduction & management has on two habitat elements (predators and competitors), mentioned above, this controlling factors also has direct effects on four additional habitat elements (total of six). These include food availability, infectious agents, substrate, and vegetation structure. While these relationships are of low or medium magnitude, all of them are of medium or high understanding.

ES-5 Northern Mexican Gartersnake (Thamnophis eques megalops) (NMGS) Basic Conceptual Ecological Model for the Lower Colorado River

The CEM includes links with unknown magnitude based on established ecological principles and knowledge of particular features of snake biology and ecology in general for which there currently is no documentation for NMGS or any closely related species in particular. Specifically:

• The CEM identifies the effects of disease on adult fertility, neonate/ juvenile survival, and adult survival to be of unknown magnitude.

• The CEM identifies the effect of infectious agents on disease to be of unknown magnitude.

• The CEM identifies the effect of genetic diversity on disease to be of unknown magnitude.

• The CEM identifies the effects of chemical stress on adult survival, neonate/juvenile survival, and adult fertility to be of unknown magnitude.

• The CEM identifies the effect of ecdysis on chemical stress to be of unknown magnitude.

• The CEM identifies the effect of disease and ecdysis on each other (bi-directional relationship) to be of unknown magnitude.

Finally, the CEM also identifies several potentially important causal relationships with high magnitude and high or medium understanding. These links represent the best-understood aspects of NMGS ecology. Their medium and high ratings for link understanding reflect cumulative knowledge from several detailed studies of NMGS and their habitat. These better understood, high-magnitude relationships include:

• The CEM identifies the effects of nuisance species introduction & management on predators and competitors to both be of high magnitude and high understanding.

• The CEM identifies the effects of water storage-delivery system design & operation on one controlling factor—on-site water management—to be of high magnitude and medium understanding.

The CEM identifies the effects of water storage-delivery system design & operation on two habitat elements—hydrologic regime and surface water connectivity—to be of high magnitude and medium understanding.

ES-6

Chapter 1 – Introduction

This document presents a conceptual ecological model (CEM) for the northern Mexican gartersnake (Thamnophis eques megalops) (NMGS). The purpose of this model is to help the Lower Colorado River Multi-Species Conservation Program (LCR MSCP), Bureau of Reclamation (Reclamation), identify areas of scientific uncertainty concerning NMGS ecology, the effects of specific stressors, the effects of specific management actions aimed at species habitat restoration, and the methods used to measure NMGS habitat and population conditions. The CEM methodology follows that developed for the Sacramento-San Joaquin River Delta Regional Ecosystem Restoration Implementation Plan (DiGennaro et al. 2012), with modifications. (Note: Attachment 1 provides an introduction to the CEM process. We recommend that those unfamiliar with this process read the attachment before continuing with this document.)

The CEM addresses the present NMGS population and its distribution within the greater lower Colorado River (LCR) ecosystem along the main stem river and within the Bill Williams River watershed. Currently, NMGS populations are known to exist at the Beal Lake Conservation Area, Topock Marsh, and locations in the Bill Williams River watershed on the Bill Williams River, Big Sandy River, and Santa Maria River (O’Donnell et al. 2019; Sabin 2018). Since the species was rediscovered on the river in 2015, surveys have only been conducted at six locations at the Havasu National Wildlife Refuge (Havasu NWR) around Topock Marsh, including the Beal Lake Conservation Area willow marsh. Unsurveyed populations and potential habitat may exist elsewhere in the LCR. Historically, NMGS were recorded from Reaches 6–7 in Yuma, Arizona (circa 1889), and in Reach 3 in Ft. Mohave (circa 1904) (Holycross et al. 2006; C. Ronning 2020, personal communication; Rosen and Schwalbe 1988). The model addresses the landscape as a whole rather than any single managed area.

The basic source of information for this CEM consists of the most recent LCR MSCP species account (Sabin 2018). This publication summarizes and cites large bodies of earlier studies. Where appropriate and accessible, those earlier studies are directly cited. The CEM also integrates information from the U.S. Fish and Wildlife Service (USFWS) listing and critical habitat designation reports (USFWS 2014, 2020); research by the Arizona Division of Game and Fish (Bourne 2020; O’Donnell et al. 2019); websites and recent journal articles; studies of NMGS in other areas outside the LCR (Boyarski 2008, 2011, 2012, 2013 each in USFWS 2014); Boyarski et al. 2015 in Emmons et al. 2016a; Emmons 2017; Emmons and Nowak 2013, 2016 in Northern Arizona University (NAU) 2020; Emmons et al. 2016a, 2016b; Nowak and Boyarski 2012; Nowak et al. 2011, 2014, 2019; Rosen and Schwalbe 1988; Sprague and Bateman 2018; and Young and Boyarski 2012, 2013); general works on the types of aquatic and adjacent upland southwestern settings in which NMGS can occur; and the expert

1-1 Northern Mexican Gartersnake (Thamnophis eques megalops) (NMGS) Basic Conceptual Ecological Model for the Lower Colorado River

knowledge of LCR MSCP biologists. However, the purpose of this report is not to provide an updated literature review; rather, its purpose is to integrate the available information and knowledge into a CEM so it can be used for adaptive management.

This document is organized as follows: The remainder of chapter 1 provides a general description of the reproductive ecology of the NMGS, describes the purpose of the CEM, and introduces the underlying concepts and structure of the CEM. Succeeding chapters present and explain the CEM for NMGS within the LCR and identify possible implications of this information for management, monitoring, and research needs.

NORTHERN MEXICAN GARTERSNAKE REPRODUCTIVE ECOLOGY

NMGS are year-round residents of the Southwestern United States, currently distributed in clusters of occurrences in Arizona in the northern, eastern, and extreme southeastern tributary watersheds to the Gila River in central and southeastern Arizona and western New Mexico. In Arizona, these tributaries include the Santa Cruz, San Pedro, and Agua Fria Rivers, and the Verde River and its smaller tributaries (e.g., Oak Creek, Tonto Creek). NMGS are also found in reaches of the Bill Williams River, including its tributaries, the Big Sandy and Santa Maria Rivers and at Havasu National Wildlife Refuge (Bourne 2020; Cotten et al. 2013; Emmons 2017; O’Donnell et al. 2019; Sabin 2018; USFWS 2014, 2020).

Rosen and Schwalbe (1988) describe NMGS preferred habitats to be source-area wetlands, large river riparian woodlands and forests, and streamside gallery forests. Proximity to permanent water is the most important habitat feature. Specifically, NMGS are found in streams (even those with spatially or ephemerally intermittent flow [see USFWS 2020]), protected backwaters, cienegas, and other wetlands, flooded areas, ponds (Emmons and Nowak [2013] and Servoss et al. [2007] in Sabin 2018), and even stock tanks, impoundments, or canals (Conant 2003; Degenhardt et al. 1996 in Sabin 2018; NatureServe 2020 and references therein; USFWS 2014, 2020; Woodin, III 1950).

NMGS are usually found within 15 meters [m] (50 feet [ft]) of water (Rosen and Schwalbe 1988). Radio telemetry studies at the Bubbling Ponds Fish Hatchery also found that NMGS remained close to water (within ≈6 m [≈20 ft]) during their active season from March to October. In general, they rarely move farther than 50–100 m from the water’s edge during this time (NAU 2020).

1-2 Chapter 1 – Introduction

Terrestrial habitat is also important to NMGS; it is required “to support life- history functions such as immigration, emigration, and brumation” (USFWS 2013). Although the snakes typically remain in close proximity to permanent water, they will move into and through upland habitat. During the overwintering season at the Bubbling Ponds Fish Hatchery, NMGS moved into upland habitat to hibernate (Sprague and Bateman 2018). The distance from water to brumation sites ranges from an average of 39.9 ± 8.5 m (≈131 ft) along the Verde River (Emmons 2017) to a range of from 0.7 to 383 m (2 to 1,257 ft) along the Tonto River (Nowak et al. 2019; USFWS 2020 and references therein).

Prospective mates typically find and select each other in spring during April and May. Mating is also reported to occur in fall (Boyarski 2012 in USFWS 2014; Myrand et al. 2021); Rosen and Schwalbe (1988) observed high levels of male activity at Findley Tank coupled with early onset of follicular enlargement. Female gartersnakes (Thamnophis spp.) can retain male sperm for at least 7 months before using the sperm to fertilize the eggs (Halpert et al. 1982).

There is no information about where mating occurs or if there are environmental cues that lead to summer birthing. However, during late May and early June, when snakes would be gestating, researchers at the Bubbling Ponds Fish Hatchery observed that female NMGS remained underground or basked in open meadows, clearings, and other open sunnier habitats (Boyarski et al. 2015 in Emmons et al. 2016a; Emmons and Nowak 2016 in NAU 2020). At this same location, females appeared to give birth near shallow water ponds and backwaters with rushes (Juncaceae), cattails (Typhus sp.), and tules (Cyperaceae). Neonates tended to remain in or near these habitats (NAU 2020). Females give birth to 7–38 young in June through early July (Emmons and Nowak 2016 in NAU 2020); Nowak and Boyarski 2012; Rosen and Schwalbe 1988). Some records note that only half of any population gives birth in a year (Rosen and Schwalbe 1988).

CONCEPTUAL ECOLOGICAL MODEL PURPOSES

Adaptive management of natural resources requires a framework to help managers understand the state of knowledge about how a resource “works,” what elements of the resource they can affect through management, and how the resource will likely respond to management actions. The “resource” may be a population, species, habitat, or ecological complex. The best such frameworks incorporate the combined knowledge of many professionals accumulated over years of investigations and management actions. CEMs capture and synthesize this knowledge (Fischenich 2008; DiGennaro et al. 2012). The CEM methodology followed here is a crucial foundation for carrying out effects analyses as described by Murphy and Weiland (2011, 2014) and illustrated by Jacobson et al. (2016).

1-3 Northern Mexican Gartersnake (Thamnophis eques megalops) (NMGS) Basic Conceptual Ecological Model for the Lower Colorado River

CEMs explicitly identify: (1) the variables or attributes that best characterize resource conditions, (2) the factors that most strongly shape or control these variables under both natural and altered (including managed) conditions, (3) the character, strength, and predictability of the ways in which these factors do this shaping/controlling, and (4) how the characteristics of the resource vary as a result of the interplay of its shaping/controlling factors.

By integrating and explicitly organizing existing knowledge in this way, a CEM summarizes and documents: (1) what is known, with what certainty, and the sources of this information, (2) critical areas of uncertain or conflicting science that demand resolution to better guide management planning and action, (3) crucial attributes to use while monitoring system conditions and predicting the effects of experiments, management actions, and other potential agents of change, and (4) how the characteristics of the resource would likely change as a result of altering its shaping/controlling factors, including those resulting from management actions.

A CEM thus translates existing knowledge into a set of explicit hypotheses. The scientific community may consider some of these hypotheses well tested, but others less so. Through the model, scientists and managers can identify which hypotheses, and the assumptions they express, most strongly influence management actions. A CEM thus helps guide management actions based on the results of monitoring and experimentation. These results indicate whether expectations about the results of management actions—as clearly stated in the CEM—have been met or not. Both expected and unexpected results allow managers to update the model, improving certainty about some aspects of the model, while requiring changes to other aspects, to guide the next cycle of management actions and research. A CEM, through its successive iterations, becomes the record of improving knowledge and the ability to manage the system.

CONCEPTUAL ECOLOGICAL MODEL STRUCTURE

The CEM methodology used here expands on that developed for the Sacramento- San Joaquin River Delta Regional Ecosystem Restoration Implementation Plan (DiGennaro et al. 2012). The expansion incorporates recommendations of Burke et al. (2009), Kondolf et al. (2008), and Wildhaber et al. (2007, 2011) to provide greater detail on causal linkages and outcomes and explicit demographic notation in the characterization of life-stage outcomes (McDonald and Caswell 1993). Attachment 1 provides a detailed description of the methodology. The resulting model is a “life history” model, as is common for CEMs focused on individual species and their population dynamics (Wildhaber et al. 2007, 2011).

1-4 Chapter 1 – Introduction

That is, a CEM distinguishes the major life stages or events through which the individuals of a species must pass to complete a full life cycle, including reproducing, and the biologically crucial outcomes of each life stage. These biologically crucial outcomes minimally include the number of individuals recruited to the next life stage (e.g., juvenile to adult) or to the next age class within a single life stage, termed the recruitment rate, and the number of viable offspring produced, termed the fertility rate. The CEM then identifies the factors that shape the rates of these outcomes in the study area and thereby shapes the abundance, distribution, and persistence of the species in that area.

The NMGS conceptual ecological model has five core components as explained further in attachment 1:

• Life stages – These consist of the major growth stages and critical events through which the individuals of a species must pass in order to complete a full life cycle.

• Life-stage outcomes – These consist of the biologically crucial outcomes of each life stage, including the number of individuals surviving to the next life stage (e.g., from juvenile to adult), and the number of offspring produced (fertility rate). The rates of the outcomes for an individual life stage depend on the rates of the critical biological activities and processes for that life stage.

• Critical biological activities and processes – These consist of the activities in which the species engages and the biological processes that take place during each life stage that significantly affect its life-stage outcomes rates. Examples of activities and processes for a reptile species may include basking, brumation, and foraging. Critical biological activities and processes typically are “rate” variables.

• Habitat elements – These consist of the specific habitat conditions, the quality, abundance, and spatial and temporal distributions of which significantly affect the rates of the critical biological activities and processes for each life stage. These effects on critical biological activities and processes may be either beneficial or detrimental. Taken together, the suite of natural habitat elements for a life stage is called the “habitat template” for that life stage. Defining the natural habitat template may involve estimating specific thresholds or ranges of suitable values for particular habitat elements, outside of which one or more critical biological activities or processes no longer fully support desired life-stage outcome rates—if the state of the science supports such estimates.

1-5 Northern Mexican Gartersnake (Thamnophis eques megalops) (NMGS) Basic Conceptual Ecological Model for the Lower Colorado River

• Controlling factors – These consist of environmental conditions and dynamics—including human actions—that determine the quality, abundance, and spatial and temporal distributions of important habitat elements. Controlling factors are also called “drivers.” There may be a hierarchy of such factors affecting the system at different scales of time and space (Burke et al. 2009). For example, the availability of suitable habitat for a reptile may depend on factors such as local hydrology, substrate, and vegetation structure, which in turn may depend on factors such as the water storage-delivery system design and operation (dam design, reservoir morphology, and dam operations), which in turn is shaped by climate, land use, vegetation, water demand, and watershed geology.

The process of identifying the life stages, life-stage outcomes, critical biological activities and processes, habitat elements, and controlling factors for a CEM begins with a review of the LCR MSCP and other major accounts for the species of interest, accounts for better known but closely related or ecologically similar species, and LCR MSCP management concerns as expressed in the LCR MSCP Habitat Conservation Plan (LCR MSCP 2004), and annual work plans (LCR MSCP 2018a). The process also follows conventions for life history CEMs focused on individual species and their population dynamics in the relevant branch of zoology for the species of interest. Further, the process is guided by an overarching need to ensure that the CEM helps the LCR MSCP identify areas of scientific uncertainty concerning the ecology and specific habitat requirements of the species it has been charged with conserving, the effects of specific stressors on these species, the effects of specific management actions aimed at habitat and species conservation, and the appropriate methods with which to monitor species and habitat conditions. Each CEM is developed in consultation with experts in the LCR MSCP, and submitted in draft form for review by the LCR MSCP, to ensure that the CEM meets management needs. Terminology for life stages, life- stage outcomes, critical biological activities and processes, habitat elements, and controlling factors is standardized across CEMs where feasible and appropriate.

The process of identifying the life stages for a CEM recognizes that the life cycle of any species can be divided into multiple life stages. There is no rule for how many life stages a CEM must include, and different scientists may lump together or divide up the life cycle into a different set of life stages. The process of identifying the life stages for the LCR MSCP conceptual ecological models takes into account the following two criteria for lumping versus splitting life stages. First, knowledge of the species in the LCR valley prior to river regulation and the general ecological literature for similar species indicates that there could be differences in habitat requirements, threats, behaviors, or management requirements for individuals in different portions of the life cycle. Second, a single life stage may encompass several age classes. However, unless there are strong ecological reasons to distinguish individual age classes or groups of age

1-6 Chapter 1 – Introduction

classes as separate life stages, the LCR MSCP conceptual ecological models combine different age classes into the fewest life stages that make good ecological sense.

The process of identifying the life-stage outcomes for a CEM follows the conventions for life history CEMs focused on individual species and their population dynamics in the relevant branch of zoology for the species of interest, as noted above. These conventions recognize three possibilities: (1) The outcomes for an individual life stage may consist exclusively of survival. For example, the outcome of a juvenile life stage may consist only of survival to become an adult. (2) The outcomes for an individual life stage may consist of both survival and participation in reproduction, when participation in reproduction constitutes a distinct life stage for the species. (3) Alternatively, the outcomes for an individual life stage may consist of both survival and fertility, the latter of which concerns the production of viable fertilized eggs in the absence of parental care or the production of viable newborn in the presence of parental care. This third possibility pertains either to a life stage in which all individuals participate in reproduction, or to a life stage that focuses only on some subset of adults that engages in reproduction in a single year, such as “Breeding Adult.” Several of the species of concern to the LCR MSCP are subject to management goals concerning their genetic integrity; however, the present CEMs focus only on demographic outcomes unless the LCR MSCP Adaptive Management Program specifically requests that the CEM also include outcomes related to genetic integrity.

The process of identifying the critical biological activities and processes for a CEM focuses on identifying three possibilities in the literature: (1) activities necessary to achieve one or more life-stage outcomes, such as feeding, mating, migrating, avoiding or escaping hazards, or resting in (relatively) safe settings, (2) biological processes that individuals must undergo to achieve one or more life- stage outcomes, such as maturing sexually, developing adult morphology and strength, or mating, and (3) biological processes that individuals will experience during the life stage that affect their fitness or survival, such as encounters with predators and/or competitors, or experiences with physical or physiological stress that reduces fitness. Critical biological activities and processes thus may be either beneficial or detrimental to fitness, survival, or reproduction. Critical biological activities and processes may affect life-stage outcomes directly or only indirectly through their effects on other critical biological activities or processes. For example, disease may not always result in death (i.e., may not always directly affect survivorship), but it may make an individual weaker or disoriented and, therefore, less able to forage or be more vulnerable to depredation.

Ordinarily, only the life-stage outcomes of an individual life stage—survival and fertility—affect demographic dynamics in the next life stage; however, in some circumstances, critical activities or processes for one life stage also may affect dynamics in the next life stage. Most commonly, such transgenerational

1-7 Northern Mexican Gartersnake (Thamnophis eques megalops) (NMGS) Basic Conceptual Ecological Model for the Lower Colorado River

dynamics involve patterns of parental investment in raising offspring. For example, preparing a nest for eggs, protecting the eggs during incubation, and caring for the nestlings after the eggs hatch are all critical activities for breeding adult birds that have energetic and other costs for these adults. At the same time, these activities constitute crucial features of the environment—i.e., habitat elements—for the eggs and nestlings that affect their access to food and vulnerability to predators.

The process of identifying the critical biological activities and processes for a CEM recognizes that the critical biological activities and processes for any species can be combined or split into different categories in different ways. A single critical biological activity or process may encompass several more specific variables, behaviors, or changes. There is no rule for how many critical biological activities and processes a CEM must include or for determining which specific variables, behaviors, or changes to lump together under the heading of a single critical biological activity or process and which to split under separate headings. As with the process of identifying the life stages for the LCR MSCP conceptual ecological models, the process of identifying the critical biological activities and processes for a CEM looks for information on the species within its historic range and information in the general ecological literature for similar species indicating that there could be differences in habitat requirements, threats, or management requirements for different possible critical biological activities or processes.

The process of identifying the habitat elements for each life stage in a CEM focuses on identifying physical or biological environmental conditions that: (1) are necessary or beneficial for the successful participation of individuals of a life stage in particular beneficial critical biological activities or processes, (2) may limit or prevent the successful participation of individuals of a life stage in particular beneficial critical biological activities or processes, or (3) may result in the participation of individuals of a life stage in particular detrimental critical biological activities or processes. Habitat elements thus shape the rates of beneficial or detrimental critical biological activities or processes. Further, habitat elements may affect critical biological activities or processes directly, indirectly through their effects on other habitat elements, or both. For example, the herbaceous vegetation in a marsh may benefit an aquatic species directly by providing protective cover and plant litter on which the aquatic species may feed or indirectly by helping maintain cooler water temperatures, stabilizing the marsh substrate, and providing habitat for insects on which the aquatic species also may feed. However, the same marsh vegetation may also provide habitat for invertebrate or vertebrate species that may prey on the aquatic species of interest.

The process of identifying the habitat elements for each life stage in a CEM also recognizes that the key physical or biological environmental conditions affecting the individuals of a life stage can be combined or split into different categories in different ways. A single habitat element may encompass several more specific

1-8 Chapter 1 – Introduction

variables or properties of the physical or biological environment. There is no rule for how many habitat elements a CEM must include or for determining which specific properties of the physical or biological environment to lump together under the heading of a habitat element and which to split under separate headings. The process of identifying the habitat elements for each life stage in a CEM lumps together properties of the physical or biological environment that closely covary with each other over space and time along the LCR because these properties are shaped by the same controlling factors and laws of physics or chemistry and/or because these properties strongly interact with each other and, therefore, are not independent. A CEM also may lump together properties of the physical or biological environment when there is not sufficient knowledge to split these properties into separate habitat elements in ways that would help the LCR MSCP manage the species of concern. Finally, CEMs lump together properties of the physical or biological environment that have similar effects or management implications across multiple life stages even if these effects or implications differ in their details between life stages. Lumping together such closely related properties under the heading for a single habitat element across all life stages makes comparison and integration of the CEMs for the individual life stages across the entire life cycle less difficult. On the other hand, a CEM may split properties of the physical or biological environment into separate habitat elements if they do not meet any of these criteria.

Finally, the process of identifying the controlling factors for each life stage in a CEM focuses on environmental conditions and dynamics—including human actions—that (1) determine the quality, abundance, and spatial and temporal distributions of important habitat elements and (2) are within the scope of potential human manipulation, most particularly manipulation by the LCR MSCP and its conservation partners along the LCR valley. The specific or “immediate” controlling factors identified in a CEM necessarily exist and vary in a larger context of human institutions and policies and both short- and long-term dynamics of climate and geology; however, a CEM does not address this larger context. The process of identifying the controlling factors for each life stage in a CEM also recognizes that a controlling factor may affect a habitat element directly or indirectly through its effects on either another controlling factor or another habitat element.

The process of identifying the controlling factors for each life stage in a CEM also recognizes that the key drivers affecting the habitat elements for that life stage can be combined or split into different categories in different ways. A single controlling factor may encompass several more specific variables or human activities. There is no rule for how many controlling factors a CEM must include. The process of identifying the controlling factors for each life stage in a CEM lumps together types of human activities in particular that closely covary with each other over space and time along the LCR because of the institutions and policies driving them and/or because these activities strongly interact with

1-9 Northern Mexican Gartersnake (Thamnophis eques megalops) (NMGS) Basic Conceptual Ecological Model for the Lower Colorado River

each other and, therefore, are not independent. A CEM also may lump together human activities when there is not sufficient knowledge to split these into separate categories in ways that would help the LCR MSCP manage the species of concern. Finally, CEMs lump together human activities as controlling factors when these activities have similar effects or management implications across multiple life stages and across multiple species of concern to the LCR MSCP, even if these effects or implications differ in their details between life stages and species. Lumping together such closely related activities under the heading for a single controlling factor across multiple species and multiple life stages of these species makes comparison and integration of CEMs across the LCR MSCP less difficult.

Each CEM not only identifies these five components for each species, it also identifies the causal relationships among them that affect life-stage outcome rates. Further, a CEM assesses each causal linkage based on four variables to the extent possible with the available information: (1) the character and direction of the effect, (2) the magnitude of the effect, (3) the predictability (consistency) of the effect, and (4) the status (certainty) of a present scientific understanding of the effect. Attachment 1 provides detailed definitions and criteria for assessing these four variables for each causal link. Each CEM attempts to include all possible “significant” causal linkages among controlling factors, habitat elements, critical biological activities and processes, and life-stage outcomes for each life stage. “Significant” here means that, based on the available literature and knowledge of experts in the LCR MSCP, the linkage has been proposed to exist or appears reasonably likely to exist and to have the potential to affect management of the species.

A CEM for each life stage thus identifies the causal relationships that most strongly support or limit the rates of its life-stage outcomes, support or limit the rate of each critical biological activity or process, and support or limit the quality, abundance, and distribution of each habitat element (as these affect other habitat elements or affect critical biological activities or processes). In addition, the model for each life stage highlights areas of scientific uncertainty concerning these causal relationships, the effects of specific management actions aimed at these relationships, and the suitability of the methods used to measure habitat and population conditions. Attachment 1 provides further details on the assessment of causal relationships, including the use of diagrams and a spreadsheet tool to record the details of a CEM and to summarize the findings. Software tools developed in association with these CEMs allow users to query the CEM spreadsheet for each life stage and generate diagrams that selectively display query results concerning the CEM for each life stage. For example, a query may selectively identify all links with high magnitude but low understanding or it may identify the critical biological activities or processes for a life stage with the greatest number of poorly understood drivers or effects.

1-10

Chapter 2 – NMGS Life-Stage Model

A life stage consists of a biologically distinct portion of the life history of a species during which individuals undergo distinct developments in body form and function, engage in distinct behaviors, use distinct sets of habitats, and/or interact with their larger ecosystems in ways that differ from those associated with other life stages. This chapter proposes a life-stage model for NMGS within the LCR on which to build the CEM. Except where noted, the sources for the following information are Sabin (2018), USFWS (2014, 2020) and references therein. Table 1 and figure 1 summarize the proposed NMGS life-stage model.

Table 1.—NMGS life stages and life-stage outcomes

Life stage Life-stage outcomes 1. Neonates/juveniles • Neonate/juvenile survival 2. Adults • Adult survival • Adult fertility

Figure 1.—Proposed NMGS life history model. Squares indicate life stages, diamonds indicate life-stage outcomes, and arrows indicate life-stage transitions. S1-2 = survival of neonates/juveniles, F2-1 = adult fertility, and S2-2 = survival of adults.

2-1 Northern Mexican Gartersnake (Thamnophis eques megalops) (NMGS) Basic Conceptual Ecological Model for the Lower Colorado River

INTRODUCTION TO THE NMGS LIFE CYCLE

The NMGS conceptual ecological model for each of these life stages recognizes a minimum of one life-stage outcome: survival. Survival” for neonates/juveniles in the NMGS conceptual ecological model refers to the rate at which members of a local population survive through their entire life stage to enter—recruit into—the next life stage. “Survival” for adults in the NMGS conceptual ecological model refers to the rate at which individuals in a local population survive from year to year. Although no data are available, the USFWS (2014, 2020) reports that an overall lifespan of 10 years is likely for NMGS, with a maximum estimated longevity of 15 years.

The NMGS conceptual ecological model recognizes fertility as an additional life- stage outcome for the adults life stage. “Fertility” in the NMGS conceptual ecological model refers to the rate of live births per adult female.

NMGS LIFE STAGE 1 – NEONATES/JUVENILES

The NMGS neonates/juveniles life stage begins when the young snakes are born (NMGS are live-bearing rather than egg-laying reptiles) and lasts until sexual maturity is reached. Upon birth, young snakes disperse following the pheromone trails of other gartersnakes. They typically begin foraging almost immediately. This life stage is characterized by rapid growth, with young snakes shedding their skins an average of four times per year. Upon reaching maturity, growth slows, and the rate of shedding also slows down to 1–2 times per year. Although no specific data are available, it is likely that the predation rate of young snakes is higher than that of adults due, in part, to the smaller size of neonates/juveniles.

Some herpetologists consider the neonate life stage to be separate from juveniles and subadults. While there is good information about nesting behavior in reptiles that lay eggs, and maternal basking and parturition for those that are ovoviviperous or viviperous, once the young are born (or hatch), there is very little known about what happens when newly hatched or emerged young disperse.

NMGS LIFE STAGE 2 – ADULTS

The NMGS adults life stage begins when snakes reach sexual maturity. This occurs at 18 inches (in) (≈46 centimeters [cm]) of length (ca. 2 years) in males and about 24 in (≈61 cm) in length in females = [ca. 2–3 years of age]) (Rosen and Schwalbe 1988). Upon maturity, mating occurs in early spring (March until

2-2 Chapter 2 – NMGS Life-Stage Model

May), although occasionally pairs will mate in fall (Boyarski 2012 in USFWS 2014; Myrand et al. 2021; Rosen and Schwalbe 1988). Regardless of when mating occurs, females are ovoviviparous and typically give birth during the summer months, mid-May through early July or into August, to 7–26 young (mean ≈14 young) (Rosen and Schwalbe 1988); Nowak and Boyarski (2012) reported a gravid female giving birth to 38 young at the Bubbling Ponds Fish Hatchery located on the Verde River in Arizona. At this location, during the spring gestation period, female NMGS either remained underground or basked in sunny habitats—meadows or clearings (Boyarski et al. 2015 in Emmons et al. 2016a; Emmons and Nowak 2016 in NAU 2020). Emmons and Nowak (2016 in NAU 2020) observed females apparently moving to dense vegetation in or near areas of shallow water to give birth. Not all females give birth every year. Rosen and Schwalbe (1988) note that only half of the females they studied bore young in any given year. Similarly, Sprague and Bateman (2018) also note that not all females gestated in both years of their study.

Adult gartersnakes do not provide parental care; after birth, neonates are on their own. Adult NMGS remain active for most of the year, spring through fall, but they will seek shelter for the winter months (typically late November through February) in underground refuges. Snakes may live about 10 years, with a maximum estimated longevity of 15 years (USFWS 2014, 2020).

2-3

Chapter 3 – Critical Biological Activities and Processes

Critical biological activities and processes consist of activities in which the species engages and biological processes that take place during each life stage that significantly shape the rate(s) of the outcome(s) for that life stage. Critical biological activities and processes are “rate” variables (i.e., the rate [intensity] of these activities and processes, taken together, determine the rate of recruitment of individuals from one life stage to the next).

The NMGS conceptual ecological model identifies nine critical biological activities and processes that affect one or more NMGS life stages. Some of these activities or processes differ in their details among life stages; however, grouping activities or processes across all life stages into broad types makes it easier to compare the individual life stages to each other across the entire life cycle. Table 2 lists the nine critical biological activities and processes and their distribution across life stages.

Table 2.—NMGS critical biological activities and processes and their distribution between life stages (Xs indicate life stages to which each critical biological activity or process applies.)

Life stage 

Critical biological activity or process  Neonates/juveniles Adults

Basking X X Brumation X X Chemical stress X X Competition X X Disease X X Dispersal X X Ecdysis X X Foraging X X Predation X X

3-1 Northern Mexican Gartersnake (Thamnophis eques megalops) (NMGS) Basic Conceptual Ecological Model for the Lower Colorado River

Except where noted, the sources for the following information are Sabin (2018) and USFWS (2014, 2020). These publications summarize all earlier studies. Where appropriate and accessible, those earlier studies are directly cited. The CEM also integrates information from both older and more recent works as well as the expert knowledge of LCR MSCP biologists. In addition, where appropriate, the discussions of individual critical biological activities and processes draw upon literature concerning gartersnakes of the Southwestern United States or elsewhere. The following paragraphs discuss the nine critical biological activities and processes in alphabetical order.

BASKING

Exposure to temperatures outside their range of tolerance presumably renders NMGS in each life stage vulnerable to reduced metabolic rates, reduced growth, impaired performance, disease, stress, and mortality. As ectothermic animals, they rely on ambient temperatures to regulate body heat. However, gartersnakes generally are adaptable species and have evolved ways to cope with a variety of habitat conditions across their range.

NMGS are most active during the summer months (July through August and June and September) between temperatures of between 22–33 degrees Celsius (°C) (71–91 degrees Fahrenheit [°F] Rosen 1991). If temperatures are higher, snakes will retreat to underground burrows or aquatic habitats to remain cooler (Sprague and Bateman 2018). In the winter months, NMGS brumate to avoid the coldest temperatures (See below, “Brumation”).

Basking is a daily thermoregulatory behavior in which a snake exposes all or part of its body to thermal radiation while being immobile. This serves to raise the snake’s body temperature to a thermal optimum necessary for locomotion, digestion (see Huey et al. 1989; Tattersall et al. 2004 and references therein), and other physiological activities as well as for embryonic development in gestating females (O’Donnell and Arnold 2005).

Radio telemetry studies of NMGS at the Bubbling Ponds Fish Hatchery found that during gestation, some females basked in mottled shade, close to water. In other cases, females remained underground, basking on occasion in sunny habitats such as meadows and other openings (Boyarski et al. 2015 in Emmons et al. 2016a; Emmons and Nowak 2016 in NAU 2020). During the active season of the year (March – October), male and female snakes frequented sunny edges of thick herbaceous vegetation to bask, often with just their tail or other part of their body exposed to direct sunlight.

3-2 Chapter 3 – Critical Biological Activities and Processes

Activity observations of NMGS at the same fish hatchery found that NMGS were inactive 64% of the time and active 16% of the time (unknown 20%) (Boyarski 2013 in USFWS 2014). In contrast, past observations by Rosen (1991) at another location found snakes foraging and moving 60% of the time and resting/hiding for the remainder (13% basking on vegetation, 18% basking on the ground, 9% resting under cover). It is not known why there was such a difference between studies.

BRUMATION

Brumation is a period of reduced activity engaged in by snakes and many other reptiles to survive periods of low temperature. During this time, a snake’s body temperature cools down significantly, and oxygen consumption declines. Studies of T. sirtalis during brumation in high latitude sites partially submerged in water found that oxygen consumption was reduced by 54.8%, and cardiac activity was lower by 77% (Costanzo 1989a). Dehydration, weight loss, and energy depletion are major factors affecting overwintering survival (Davenport 1992).

In the northern part of their range, NMGS retreat to terrestrial brumation sites to overwinter near riparian areas late each autumn (early November to mid- December), usually underground in burrows created by tree roots or small animals or in debris piles, and cavities under boulders, or in rock piles (Nowak et al. 2011; Sprague and Bateman 2018; USFWS 2020). On milder winter days, NMGS individuals occasionally emerge from their brumation sites to bask or move to alternate brumation sites (Emmons 2017; NAU 2020), retreating as temperatures fluctuate. Mating occurs upon emergence in early spring, typically beginning in March (C. Ronning 2020, personal communication).

Recent radio telemetry studies at the Bubbling Ponds Fish Hatchery found that NMGS at that location hibernated on a rocky slope or wooded sites with a southeasterly aspect and high percentage of forb coverage. Other brumation habitats used by NMGS at this site included mesquite bosque, meadows, and rodent burrows along aquatic edges. Sites were found 0–170 m (558 ft) away from the water’s edge (Boyarski et al. 2015 in Emmons et al. 2016a; Emmons and Nowak 2016 in NAU 2020; Myrand et al. 2017 in Emmons 2017). Along the Tonto River in Arizona, 14 NMGS were found at brumation sites at distances from water that ranged from 0.7 to 383 m (2–1,257 ft) (Nowak et al. 2019). Emmons (2017) found that winter retreat sites along the Verde River used by female NMGS were located 39.9 ± m (≈131 ft) from water and included burrows, flood debris piles, rocky slopes, and junk piles.

At the Bubbling Ponds Fish Hatchery, females avoided areas with minimal to no ground cover and selected sites with a higher canopy cover of vegetation greater

3-3 Northern Mexican Gartersnake (Thamnophis eques megalops) (NMGS) Basic Conceptual Ecological Model for the Lower Colorado River

than 1 m (3.3 ft) in height. Males chose locations with high amounts of vegetation, particularly shrubs, that were farther from water than females (males were 25.32 m [83 ft] from water; females were 23.7 m [77.8 ft] from water). Before entering brumation sites, researchers noted that the snakes regularly moved back and forth between water and upland during a short transition period (Sprague and Bateman 2018).

There is little information available about what cues migration to the winter hibernation site in gartersnakes, other than temperature or possibly photoperiod, and specifically whether NMGS overwinter singly or with other snakes and whether individual NMGS change sites from year to year or exhibit site fidelity to a particular hibernation area. Emmons (2017) observed that on two occasions at a Verde River site, two pairs of female snakes overwintered in the same underground brumation site, confirming communal denning, and that one radio- tracked female returned to a wintering site within 9 m (29.5 ft) of the one she had used the previous winter, exhibiting “general site fidelity.” Many T. sirtalis exhibit fidelity to specific hibernacula (Gregory 1977; Gregory and Skebo 1998; Gregory and Stewart 1975). However, dispersing individuals will move to new brumation areas periodically, potentially enhancing the genetic diversity of the population when mating occurs the following spring.

There is no information about whether NMGS use brumation sites to aestivate, avoiding times of drought or extreme temperatures during summer. They may simply take advantage of their aquatic habitat for temperature moderation (see chapter 4, “Temperature”).

CHEMICAL STRESS

NMGS in every life stage are vulnerable to stress and mortality due to exposure to harmful chemicals such as pesticides or to natural chemicals at extreme concentrations. Pesticide use on agricultural fields adjacent to wetlands and streams could result in contamination of NMGS habitat either through wind transport of sprayed chemicals, through chemical leaching into shallow groundwater, or directly via agricultural runoff. The pesticides used may be toxic to prey of NMGS, reducing food availability and/or causing sublethal or lethal poisoning of NMGS via ingestion of treated insects or fishes that bioaccumulate toxins. Similarly, environmental contaminants such as mercury and other heavy metals can bioaccumulate with adverse effects on NMGS (USFWS 2014). Drewett et al. (2013 in USFWS 2014) observed that the more aquatic the species of gartersnake, the greater the likelihood of mercury exposure and bioaccumulation occurring from ingestion of fish and other aquatic prey. While selenium in some backwaters is monitored for fishes in the LCR, in addition to

3-4 Chapter 3 – Critical Biological Activities and Processes

temperature, dissolved oxygen, specific conductivity, and pH (C. Ronning 2020, personal communication), no research has been conducted directly on NMGS and chemical stress in the LCR.

COMPETITION

All species face competition from other species and other members of their own species for the resources they need to survive, grow, and reproduce. Non-native species are significant competitors with NMGS for resources, especially food. Northern crayfish (Oronectes virilis), red swamp crayfish (Procambarus clarkia), American bullfrogs (Lithobates catesbeiana), and numerous fish species (e.g., bass, various catfish species [Siluriformes spp.], bullhead (Ameiurus sp.), sunfish (Lepomis sp.), crappie [Pomoxis sp.], and brown trout [Salmo trutta]) directly compete for fishes and other food resources, removing smaller-sized species or age classes on which NMGS depend (Carpenter 2005; USFWS 2014 and references therein). (One new potential non-native competitor may be the southern watersnake [Nerodia fasciata], first observed in Mittry Lake and Imperial Dam during surveys in 2016 [Californiaherps.com 2021; Myatt 2016; B. Sabin 2021, personal communication]). With reduced prey availability, NMGS must increase foraging, expending more energy on feeding activity that could otherwise be allocated to growth and reproduction (Rosen et al. 2001).

Feral hogs (Sus scrofa) may also compete with NMGS. They are present at the Beal Lake Conservation Area and Topock Marsh and prey on some of the same species as do NMGS (e.g., fishes, frogs, and lizards). During NMGS surveys, gartersnakes were not found in the end of the Beal Lake Conservation Area willow marsh being used as hog wallows. It is not known whether the snakes were not detected because they were not present, or because they were responding to the presence and activities of feral hogs. They may have been avoiding the area due to the hogs’ modification of the physical habitat or because there was less food available due to competition with the hogs (which are voracious feeders), or because NMGS were eaten by or avoiding the threat of the hogs (C. Ronning 2020, personal communication).

Possible competition between NMGS and related gartersnake species may also occur. The checkered gartersnake (T. marcianus marcianus) has been reported in some locations where they co-occur (Rosen and Schwalbe 1988; Rosen et al. 2001). (High densities of checkered gartersnakes have been detected in the LCR at Mittry Lake and surrounding canals [Munes et al. 2016]). Checkered gartersnakes may be less susceptible to bullfrog predation; hence, they have greater survivorship than NMGS when both are in the presence of that non-native species. Arizona Game and Fish Department (AZGFD) surveys on the Bill Williams River, Big Sandy River, Santa Maria River , Burro Creek, and

3-5 Northern Mexican Gartersnake (Thamnophis eques megalops) (NMGS) Basic Conceptual Ecological Model for the Lower Colorado River

Francis Creek found black-necked gartersnakes (Thamnophis cyrtopsis) at some sites and NMGS at others—the two species were never captured together at the same locations (O’Donnell et al. 2019). Both snake species consume the same prey, and it is possible that this distribution is the result of resource competition.

DISEASE

This process refers to diseases caused either by lack of genetic diversity or by infectious agents. Although there is little information available about NMGS in relation to disease susceptibility, NMGS in all life stages are conceivably susceptible to diseases. One of the newest potential infectious agents is snake fungal disease (SFD) (Ophidiomyces ophiodiicola), which is prevalent in the eastern and mid-western United States and has been found in a number of snake species, including watersnakes that use similar aquatic habitat. A new non-native species, the southern watersnake, has been found in the LCR and may harbor and/or spread this or other pathogens into NMGS populations should the watersnakes expand their range (Californiaherps.com 2021; Myatt 2016; B. Sabin 2021, personal communication). Unfortunately, in 2019, researchers from NAU confirmed the presence of SFD from two northern Arizona snakes (see NAU Gartersnake Project website: https://news.nau.edu/snake-fungal-disease/#.X9TW16pKgci).

Boyarski (2008 in USFWS 2014) observed what is thought to be a tapeworm (Spirometra sp.) in NMGS, and Nowak et al. (2014) reported an observation at the Bubbling Ponds Fish Hatchery of the transmission of a nematode parasite, Macdonaldius sp., from mother to neonate NMGS, with two different mechanisms hypothesized. Thirteen species of helminth (= 11) and trematode (= 2) parasites have been found in NMGS (Sabin 2018 and references therein). It is not known whether or how those parasites affect fertility or survival.

Injuries from predation attempts, especially to a snake’s tail, can increase the likelihood of infection and disease (Fitch 2003; USFWS 2014 and references therein).

DISPERSAL

Although there is little information about dispersal in NMGS, it is a highly aquatic species and likely disperses along stream corridors and other water bodies. NMGS are most often found in or near permanent water bodies and, as described

3-6 Chapter 3 – Critical Biological Activities and Processes

by USFWS (2020), these include “perennially or spatially intermittent streams that provide both aquatic and terrestrial habitat that allows for immigration, emigration, and maintenance of population connectivity…”

Neonates disperse after birth, although it is not known how far they will travel; adults may also disperse into new habitat if suitable. Often, gartersnakes use scent trails to follow conspecifics (Costanzo 1989b and references therein; Heller and Halpern 1981). NMGS use terrestrial habitat for brumation, gestation, and for dispersal, and these terrestrial habitats also reduce the impacts of high flow events (USFWS 2020). Specifically, the USFWS (2020) has identified riparian habitat adjacent to streams that provides “dense vegetation or other natural structural components that provide cover” as critical terrestrial habitat for NMGS. Preliminary data reported by Emmons (2014 in USFWS 2014) recorded movements of 528 ft (161m) away from nearest water, up to 0.4 mi (0.61 kilometers in a single day), and terrestrial movement data are available from other studies. The USFWS (2020) reported an individual NMGS tracked for over 1 year that traveled 4,852 ft (1,478.89 m) and eight other snakes that traveled between 587.9 ft (179.2 m) to 2,580 ft (481.58 m). Juveniles may disperse greater distances than adult snakes, as they establish their own home ranges (USFWS 2020). In the related Oregon gartersnake (Thamnophis atratus hydrophilus), a juvenile snake was found to have dispersed 2.2 mi (3.6 kilometers) from its original capture location (Welsh et al. 2010). Topography is another important habitat factor in determining how far snakes will move away from water (USFWS 2020).

The range of NMGS has been significantly reduced in recent years due to the construction of dams and reservoirs, and other water control diversions and dewatering along the LCR and its tributaries as well as habitat conversion and/or fragmentation. In addition, the introduction and spread of non-native species, including bullfrogs, crayfish, and predatory fishes have decimated NMGS numbers through competition for prey and by direct predation (USFWS 2014); the presence of these introduced species, along with the recently observed southern watersnake, in a habitat may prevent NMGS dispersal and/or recolonization.

ECDYSIS

All snakes undergo the process of ecdysis, or shedding – the sloughing of dead surface skin cells. The old skin layer is usually shed in one piece during a specific shedding period (Porter 1972). In juvenile snakes, ecdysis can occur four or more times a year, as they are growing rapidly. Once maturity is reached, the growth rate slows, and snakes typically shed only once or twice a year. Shedding of the outer skin layer helps remove external parasites as the old skin

3-7 Northern Mexican Gartersnake (Thamnophis eques megalops) (NMGS) Basic Conceptual Ecological Model for the Lower Colorado River

layer is replaced and may also serve as a “method of toxin excretion” (toxins such as mercury, lead, and chromium concentrate in the skin [Burger 1992]). The frequency of shedding varies by “gender, reproductive state, body size, feeding frequency, diet, and color morph” (King and Turmo 1997 and references therein). Although not well studied in the wild, if there is a problem with incomplete shedding, especially if eye caps remain and accumulate, or if constriction bands of previously shed skin form on the tail, blocking normal blood flow, causing avascular necrosis, snakes may suffer from infection or blindness, and even death, from tail tip loss.

The period of time leading up to ecdysis is a time of some vulnerability for snakes as their eye caps have clouded and vision is reduced. They will seek refuge until shedding is complete. In most snakes, suitable humidity and other environmental conditions are important to ensure proper shedding.

Lab studies with T. sirtalis demonstrated that when snakes were in the blue-eyed state pre-shed, they fed less, were less active overall, but more likely to strike at moving stimuli. King and Turmo (1997) surmised that reduced activity allowed the snakes to devote more energy to the shedding process; the increased tendency to strike may be due to an inability to distinguish threat from non-threat or the fact that the shedding process itself may make snakes more sensitive to threats. Overall, these behavioral changes may offset the risk of predation during a time of increased vulnerability.

FORAGING

NMGS neonates/juveniles and adults must find and acquire food to survive. Foraging success can also affect reproductive output (fecundity), as was observed in related checkered gartersnakes (Ford and Seigel 1994). NMGS are “aquatic- terrestrial generalist” feeders (Drummond and Macías-García 1983 in USFWS 2014), preying mainly on amphibians (frogs and tadpoles, and salamanders when available); small fishes; invertebrates such as earthworms, slugs, and leeches; and occasionally small mice and lizards. They preferentially prey on native species such as the lowland leopard frog (Rana = Lithobates yavapiensis) (LLFR), Chiricahua leopard frog (Lithobates chiricahuensis), and Gila topminnow (Poeciliopsis occidentalis).

Native anuran prey, especially LLFR, as well as Chiricahua leopard frogs and Colorado River toads may be preferred, but NMGS can subsist on non-native prey if necessary (USFWS 2020 and references therein). Where native prey are lacking, NMGS have been reported to feed on non-native species such as young of bullfrogs and mosquitofish (Gambusia affinis), red shiners (Cyprinella lutrensis) and introduced trout species, as well as subadult green sunfish (Lepomis

3-8 Chapter 3 – Critical Biological Activities and Processes

cyanellus), bluegills (Lepomis macrochirus), or largemouth bass (Micropteris salmoides) (Emmons and Nowak 2013; Holycross et al. 2006). Similarly, at the Beal Lake Conservation Area and Topock Marsh, NMGS prey on non-native mosquitofish, crayfish, and bullfrog tadpoles and juveniles, in addition to freshwater shrimp (Bourne 2020; C. Ronning 2020, personal communication). However, in some instances, non-native predatory fishes, especially spiny-rayed species, may be a food of last resort and may not even be compatible prey, as NMGS have been known to regurgitate such species soon after eating or suffer injuries or mortality (Boyarski et al. 2015 in Emmons et al. 2016a; USFWS 2014), although Emmons et al. (2016a) observed an instance in which a gartersnake survived puncture after consumption of spiny-rayed fish. NMGS at the Page Springs State Fish Hatchery and Bubbling Ponds Fish Hatchery have been observed to prey on non-native Chinese mystery snails (Cipangopaludina chinensis), with fatalities reported when the snake’s lower jaw became permanently wedged in the snail shell (Young and Boyarski 2012). Observations by Emmons et al. (2016a) in upper-middle Verde River populations suggest that NMGS may readily consume non-native, even “harmful” prey in watersheds that lack sufficient native prey and do well—survive and reproduce.

Depending on air temperature, NMGS forage when prey are available, day or night (Holm and Lowe 1995 in USFWS 2014). The snakes forage along the banks of water bodies using a variety of foraging behaviors, and will ambush their prey on land or in water; actively hunt the edges of thick riparian vegetation in open water and riffles, or on vegetation mats; and take advantage of fishes and other prey concentrated in drying pools (Rosen and Schwalbe 1988). Researchers at the Bubbling Ponds Fish Hatchery also observed NMGS foraging in deeper water (several meters in depth) without thick vegetative cover (NAU 2020).

Foraging activity varies in response to prey availability. When native prey are available but prey size is small, foraging activity will need to increase for snakes to acquire sufficient resources, which may expose them to greater predation risk. With few or no native prey of any size available, NMGS may simply reduce foraging activity to avoid predation (Rosen and Schwalbe 1988), although they may switch to non-native prey. Snakes may also reduce foraging activity immediately after feeding, during which time individuals will remain quiet to digest their meal.

PREDATION

Predation is a threat to NMGS at all life stages and obviously affects survival. Every animal species has evolved strategies that permit its persistence despite

3-9 Northern Mexican Gartersnake (Thamnophis eques megalops) (NMGS) Basic Conceptual Ecological Model for the Lower Colorado River

predation, including specific behaviors, body features, or reproductive strategies that allow it to avoid, escape, defend against, or counterbalance losses from predation.

Although they are well camouflaged, gartersnakes will reduce foraging or other activities to minimize exposure to predators. When encountered, they may try to escape under cover, but if cornered, gartersnakes will hiss, strike repeatedly without biting (Degenhardt et al. 1996 in Sabin 2018; Woodin, III 1950), or bite. They also secrete a foul-smelling liquid from their anal glands to deter predators (Degenhardt et al. 1996 in Sabin 2018).

Neonates and young juveniles are more vulnerable to predation. As Morafka et al. (2000) describe, they are usually the “smallest, softest, easiest and most naïve age class and they occur at high densities when they first emerge as a brood.” Possible naïveté aside, newborn snakes in general have an innate predator avoidance repertoire that includes using aggressive behavior and/or tail vibrations to deter predators.

When adjacent to water, NMGS will escape into it (Sprague and Bateman 2018) or dive under water (Brennan 2008). On land, Brennan also observed that “If persistently harassed it [NMGS] has been observed to tuck head down against the ground and under the coils hold the head and forebody motionless and “crawl” the wiggling tail away from the body, “perhaps to divert a predator’s attention away from head and to the tail.”

In addition to predation resulting in death of the individual, unsuccessful predation attempts that cause injury to the snake can lead to disease or other infirmity that reduces individual fitness or results in later death. For example, many gartersnakes lose part of their tails while trying to escape from predators, and tails do not grow back in snakes, unlike with some lizard or amphibian species. Willis et al. (1982) has shown that snakes smaller than 12 in (30 cm) usually do not survive the loss of part of the tail.

Tail damage can also increase infection rates, reducing overall health. Tail damage can affect locomotion by reducing swimming and crawling speeds. Slower movement speeds, in turn, can increase susceptibility to predation and reduce foraging success. Tail damage can also negatively affect reproduction. Females are more susceptible to predation and tail damage, as they bask more often when carrying young. Larger, older snakes that lose tail sections experience reduced reproductive success—tail tips are important for proper alignment with the cloaca for mating (Shine et al. 1999).

Potential predators of NMGS reported in the literature include native birds (wading birds, common mergansers [Mergus merganser], and belted kingfishers [Megaceryle alcyon]), as well as red-tailed hawks [Buteo jamaicensis] and other

3-10 Chapter 3 – Critical Biological Activities and Processes

raptors); mammals (raccoons [Procyon lotor], skunks [Mephitis sp.], coyotes [Canis latrans], river otters [Lutra canadensis], foxes [likely Urocyon cinereoargenteus], and bobcats [Lynx rufus], etc.); spiny softshell turtles (Apalone spinifera); and other snake species (whipsnakes [Coluber sp.], kingsnakes [Lampropeltis sp.], and regal ring-necked snakes [Diadophis punctatus regalis]) (Nowak et al. 2011; Sabin 2018 and reference therein; USFWS 2014 and references therein). Non-native predators include feral hogs and aquatic animals such as brown trout and introduced species of spiny-rayed fish, particularly largemouth bass, bullfrogs, and crayfish (see chapter 4, “Predators”).

3-11

Chapter 4 – Habitat Elements

Habitat elements consist of specific conditions in the physical or biotic environment; the quality, abundance, spatial and temporal distributions, or other properties of which significantly affect the rates of critical activities and processes for one or more life stages.

This chapter identifies 13 habitat elements that may affect 1 or more critical biological activities or processes among the 2 NMGS life stages. Some of these habitat elements differ in their details among life stages. For example, different species may prey on different life stages of NMGS. However, using the same labels for the same kinds of habitat elements across all life stages makes it possible to compare the CEMs for individual life stages across the entire life cycle.

The NMGS conceptual ecological model includes habitat elements identified in species accounts and scientific studies demonstrating or positing a direct effect on one or more critical biological activities or processes for one or more NMGS life stages or for similar or related species in similar habitats. Table 3 lists the 13 habitat elements and the critical biological activities or processes that they may directly affect across all NMGS life stages. Habitat elements may also directly affect each other.

Except where noted, the sources for the information in this chapter are Sabin (2018) and USFWS (2014, 2020). These publications summarize all earlier studies. Where appropriate and accessible, those earlier studies are directly cited. The identification also integrates information from the expert knowledge of LCR MSCP biologists and from more recent studies, including investigations by Bourne (2020); Boyarski (2008, 2011, 2012, 2013 each in USFWS 2014); Boyarski et al. (2015 in Emmons et al. 2016a); Emmons 2017; Emmons and Nowak 2013, 2016 in NAU 2020); Emmons et al. (2016a, 2016b); Nowak and Boyarski (2012); Nowak et al. (2011, 2014, 2019); O’Donnell et al. (2019); Rosen and Schwalbe (1988); Sprague and Bateman (2018), and Young and Boyarski (2012, 2013). The results are hypotheses that make the best use of the available information.

The following paragraphs discuss the 13 habitat elements. As with all such tabulations of habitat associations, inferences that particular habitat characteristics may be critical to a species or life stage require evidence and CEMs for why each association matters to species viability (Rosenfeld 2003; Rosenfeld and Hatfield 2006).

4-1 Northern Mexican Gartersnake (Thamnophis eques megalops) (NMGS) Basic Conceptual Ecological Model for the Lower Colorado River

Table 3.—NMGS habitat elements and the critical biological activities and processes that are proposed to directly affect them between the two life stages (Xs indicate which habitat elements may affect each critical biological activity or process.)

Affected critical biological activity or process 

 Causal habitat element Basking Brumation Chemical stress Competition Disease Dispersal Ecdysis Foraging Predation

Anthropogenic disturbance X X X Competitors X Environmental contaminants X Food availability X X Genetic diversity X X Hydrologic regime X Infectious agents X Predators X Substrate X X X Surface water connectivity X Temperature X X X Turbidity X Vegetation structure X X X Note: Substrate indirectly affects foraging via food availability and turbidity. Anthropogenic disturbance indirectly affects competition via competitors. No habitat elements directly affect ecdysis.

The diagrams and other references to habitat elements elsewhere in this document identify the habitat elements by a one-to-three-word short name. However, each short name in fact refers to a longer, complete name. For example, “food availability” is the short name for “The diversity, size, abundance, and spatial and temporal distributions of the species on which NMGS feed.” The following paragraphs provide both the short and full names for each habitat element and a detailed definition, addressing the elements in alphabetical order.

4-2 Chapter 4 – Habitat Elements

ANTHROPOGENIC DISTURBANCE

Full name: Human activity within or surrounding a given habitat patch, including noise, pollution, and other disturbances associated with human activity. Whether due to recreation, land management, or scientific research activities, the presence of humans can disturb NMGS, causing changes in behavior that might ultimately affect survival, or cause direct mortality through encounters with vehicular traffic (Boyarski 2011 as summarized from multiple sources in USFWS 2014; Rosen and Schwalbe 1988) or with fearful or antagonistic humans. Increased human activity at a site may also degrade snake habitat through the trampling of dense streamside vegetation (USFWS 2014).

Species of conservation concern such as NMGS may be sensitive to disturbance by field investigations not only by the intrusion of field crews but by specific activities (such as those resulting in capture) and by handling following capture. Movements of field investigators and their equipment between sites also can transfer infectious agents and potentially harmful chemicals (including hand cleaners and lotions, insect sprays, or sunscreen) unless the investigators take adequate precautions (Todd and Glaudas 2013).

NMGS in the LCR have been surveyed and monitored in recent years, and these studies may be ongoing in an effort to restore and manage existing populations. Techniques include visual surveys usually in combination with some moderate trapping. Trapping and netting of fishes also occurs in backwaters containing razorback suckers (Xyrauchen texanus) and bonytail [Gila elegans] (C. Ronning 2020, personal communication). The use of minnow traps for fisheries or NMGS survey or management, while recommended by the AZGFD and USFWS, has resulted in accidental deaths by drowning of NMGS when submerged for too long under water (USFWS 2014). New, approved trapping guidelines include ways of floating traps to avoid accidental drowning (C. Ronning 2020 personal communication). Radio telemetry research may cause stress, illness, or mortality as a result of surgical implantation of a transmitter (Sprague and Bateman 2018). Protocols for all studies include methods to minimize stress or harm to individuals or populations of NMGS (LCR MSCP 2004; https://www.lcrmscp.gov/species/no_mexican_garternake.html).

Anthropogenic disturbance is considered to be a habitat element, as it is an environmental characteristic with which snakes of either life stage must contend.

4-3 Northern Mexican Gartersnake (Thamnophis eques megalops) (NMGS) Basic Conceptual Ecological Model for the Lower Colorado River

COMPETITORS

Full name: The full range of potential competitors of NMGS for food or other resources. Non-native species such as bullfrogs and crayfish are major competitors. Gartersnakes feed mainly on amphibians and fishes, with native frog species being their preferred prey. Bullfrogs, which are widespread and abundant, can significantly reduce or eliminate local amphibian populations, removing frog species preferred as prey by NMGS. Rosen et al. (2001) attributed the collapse of the prey base at the San Bernardino National Wildlife Refuge (San Bernardino NWR) in Arizona to bullfrog consumption of local frogs (specifically the LLFR and Chiricahua leopard frog) that resulted in a catastrophic decline in the resident NMGS population.

Crayfish and fishes also compete for food resources, feeding on amphibians and smaller aquatic invertebrates that NMGS may consume. Northern crayfish tend to capture smaller prey items, which reduces the same prey size class required by smaller snakes (snakes are somewhat gape limited regarding prey size) (Carpenter 2005). Competition may also occur between Marcy’s checkered gartersnakes (T. marcianus marcianus) or black-necked gartersnakes (T. cyrtopsis) and NMGS, as has been hypothesized by Rosen and Schwalbe (1988) and O’Donnell et al. (2019), respectively. The southern watersnake is also starting to invade the LCR in the southern region (Myatt 2016) and may become a potential competitor, as it feeds on native fishes and amphibians. More research is needed to determine the ecological relationships of each of these snake species with NMGS.

Feral hogs are another potential competitor. They are abundant at some of the sites where NMGS have been found (e.g., Beal Lake Conservation Area and Topock Marsh). Although primarily plant feeders, browsing on leaves and twigs and rooting around for underground tubers, they prey on some of the same species as do NMGS (e.g., fishes, frogs, and lizards), potentially competing with NMGS (U.S. Department of Agriculture 2019).

ENVIRONMENTAL CONTAMINANTS

Full name: The concentrations of chemical contaminants in locations with NMGS or with potentially suitable habitat for NMGS that could harm NMGS or species with which NMGS interact. This habitat element addresses chemicals that may drift in the air to reach NMGS from land-use activities upwind as well as naturally occurring dissolved substances or chemical pollutants in water such as ammonia (NH3), nitrate/nitrite, perchlorate, selenium, mercury,

4-4 Chapter 4 – Habitat Elements

other metals; artificial organic compounds and their derivatives; and organic waste products of combustion. Contaminants in the LCR arrive from both point and non-point sources (see “Chapter 5 – Controlling Factors”).

Atmospheric deposition of mercury may also pose a risk to NMGS along the LCR and Bill Williams River valleys. Mercury, presumed to be from atmospheric deposition, has been found in fish tissue along the LCR and in Alamo Lake on the Bill Williams River. Lithobatid frogs (Rana) may bioaccumulate mercury even in the larval stage (Bank et al. 2007; Unrine et al. 2004). Gartersnakes also bioaccumulate toxins in fatty tissue, and those species with a more aquatic prey base like the NMGS, which feed mainly on fishes and frogs, have a greater risk of exposure to toxic levels of mercury (Drewett et al. 2013 in USFWS 2014).

High NH3 concentrations resulting from livestock grazing have been detected in and downstream from Alamo Lake. Detected concentrations were recognized as harmful to fishes and anurans, both prey species of NMGS (Arizona Department of Environmental Quality 2016a, 2016b, 2018). NH3 can cause mortality in all life stages for frogs and can also cause deformities in developing embryos (Mann et al. 2009).

Selenium is regularly monitored by the LCR MSCP in marshes and backwater habitats and is considered of management concern if it is in high enough concentrations or is found to be biologically accumulating in the food chain at high levels. Currently, this is not occurring where NMGS are present, but if that changes and selenium is detected as bioaccumulating, there may be an effect on NMGS (C. Ronning 2020, personal communication).

Salinity at current levels is not considered a major concern to NMGS at this time, as marshes and backwaters are regularly monitored for salinity to manage for fishes. It may be a potential concern in disconnected, unmanaged marsh patches if present in NMGS habitat, especially if access to other suitable aquatic habitat is unavailable (C. Ronning 2020, personal communication).

The environmental contaminant habitat element also includes herbicides, pesticides, fertilizers, and industrial wastes. The literature reviewed and experts consulted to prepare this CEM do not identify any other specific airborne chemical contaminants of potential concern for NMGS habitat sites along the LCR or Bill Williams River valleys. However, airborne pesticide drift from agricultural spraying is a frequently noted suspect in amphibian declines in Arizona, the Western United States, and worldwide, which could affect the native frog species preferred by NMGS (See Braun et al. 2019).

In general, contaminants are a concern for the NMGS prey base because amphibians—including members of the R. pipiens complex—suffer deformities and altered patterns of growth, neurologic and reproductive impairment, and

4-5 Northern Mexican Gartersnake (Thamnophis eques megalops) (NMGS) Basic Conceptual Ecological Model for the Lower Colorado River

mortality from agricultural, industrial, and urban pollutants that cause harm either directly or following bioaccumulation (Bank et al. 2007; Blaustein et al. 2003; Brühl et al. 2013; Davidson et al. 2002; Mann et al. 2009; Smalling et al. 2015; Unrine et al. 2004; U.S. Environmental Protection Agency 2016). More information is needed about the specific effects of various chemicals on growth and development, fertility, and survival in NMGS.

FOOD AVAILABILITY

Full name: The diversity, size, abundance, and spatial and temporal distributions of the species on which NMGS feed. This habitat element refers to the availability of food resources (e.g., fishes, amphibians, or invertebrates) that individual NMGS will encounter during each life stage as well as the density and spatial and temporal distributions of the food supply.

According to Rosen and Schwalbe (1988), prey abundance is one of the most important factors in habitat selection for NMGS as “they must feed regularly and heavily to maintain health and gain weight.” Rosen et al. (2001) found that foraging more often for smaller prey reduced growth and reproduction. Food availability directly affects reproduction: feeding before pregnancy in a related species, the checkered gartersnake, was shown to positively affect litter size (Ford and Seigel 1989a), while food intake during pregnancy is allocated to increasing postpartum mass in a female, helping to ensure survival through hibernation to the subsequent breeding season (Ford and Seigel 1989b; Gregory and Skebo 1998).

Although NMGS regularly use upland habitat for basking, brumation, and shedding, the snakes rely on aquatic habitat for most of their foraging, feeding in water or along streambanks and wetland edges (Rosen and Schwalbe 1988). Gartersnakes locate their prey by chemical cues and by sight; they require clear water to locate their fish or amphibian prey, which they capture by grasping in their mouth.

Gartersnakes feed on a variety of organisms, depending on the size of both snake and prey. Snakes swallow their food whole; thus, food availability is somewhat constrained by the size of the snake in relation to the size of the prey item. Young snakes feed on smaller items such as minnows, tadpoles, or recently metamorphosed toads (they do not seem to be negatively affected by bufotoxins), worms and other invertebrates, both aquatic and terrestrial. Larger adult snakes feed on larger amphibians (anurans) and fishes, and even small mammals. Despite their ability to consume a variety of prey, NMGS feed preferentially on native amphibian and fish species, favoring LLFR and Chiricahua leopard frogs. It had been thought that in the absence of native prey species, NMGS may fail to thrive; Rosen et al. (2001) attributed competition with bullfrogs that removed

4-6 Chapter 4 – Habitat Elements

native amphibians with the crash of the San Bernardino NWR population of NMGS. However, recent work by Emmons et al. (2016a) has identified NMGS populations in the Verde River that survive and reproduce on predominately non- native fishes and amphibian prey. Similarly, NMGS have been found at the Havasu NWR, where native LLFR are thought to have been extirpated; NMGS are likely feeding on non-native species such as bullfrogs and mosquitofish (B. Sabin 2021, personal communication). Native prey may be preferred, but NMGS can subsist on non-native prey if necessary (USFWS 2020 and references therein).

In northern Mexico, NMGS have been known to congregate at amphibian breeding ponds, possibly to take advantage of the large pulse of available prey as frogs or salamanders congregate to mate (d’Orgeix et al. 2013; Macías-García and Drummond 1988; USFWS 2014). Other authors hypothesize that birth in NMGS may have evolved to coincide with the emergence of toad metamorphosis from aquatic habitats (USFWS 2014), ensuring an abundant prey resource for neonates.

GENETIC DIVERSITY

Full name: The genetic diversity of NMGS (sub)populations. This habitat element refers to the genetic homogeneity versus heterogeneity of a (sub)population during each life stage. The greater the heterogeneity, the greater the possibility that individuals of a given life stage will have genetically encoded abilities to survive their encounters with the diverse stresses presented by their environment and/or take advantage of the opportunities presented (Allendorf and Leary 1986; Wood et al. 2018). Increased habitat fragmentation or other factors that lead to population isolation, such as reduced capacity for dispersal, will reduce heterogeneity.

The lineage of NMGS in the Lower Colorado River Basin differs genetically from sister populations in Mexico, exhibiting lower genetic diversity (Wood et al. 2018). For this reason, there is concern over genetic isolation among NMGS local/LCR populations as a result of habitat loss and fragmentation of surface water connectivity in addition to other factors such as non-native species that block dispersal along water courses (Wood et al. 2018). This is particularly concerning because these isolated populations are very small and are likely to be below the minimum effective population size for NMGS. To address these concerns, genetic management and translocation strategies have been proposed, such as assisted dispersal or translocation into appropriate habitat, new or restored (Wood et al. 2018).

4-7 Northern Mexican Gartersnake (Thamnophis eques megalops) (NMGS) Basic Conceptual Ecological Model for the Lower Colorado River

HYDROLOGIC REGIME

Full name: The magnitude, timing, and duration of surface water availability and movement at locations with, or potentially suitable for, NMGS habitat within the LCR ecosystem, and the patterns of variation in these, over time. This habitat element addresses features of the surface water hydrologic regime that may affect NMGS occupancy, behaviors, and life-stage outcomes at individual sites, including water permanence, depth, wetted area, movement (flow or turnover), and their temporal variability, including extreme events (droughts and floods).

Rosen and Schwalbe (1988) report that proximity to permanent water is one of the most important habitat features for NMGS in Arizona; NMGS are usually found within 15 m (50 ft) of water. Radio telemetry studies at the Bubbling Ponds Fish Hatchery also found that NMGS remained close to water (within ≈6 m [≈20 ft]) during their active season from March to October. In a study at Tonto Creek, Nowak et al. (2019) observed the mean distance to water for NMGS to range from 1.18–95.25 m (3.87 to 312.5 ft). Emmons (2017) found that the mean distance to water of NMGS along the Verde River was 14.7±1.4 m (≈48 ft). Specifically, NMGS are found in streams (including those with spatially or ephemerally intermittent flow [see USFWS 2020]), protected backwaters, cienegas, and other wetlands, flooded areas, and ponds (Emmons and Nowak [2013]; Servoss et al. [2007] in Sabin 2018), and even stock tanks, impoundments, or canals (Conant 2003; Degenhardt et al. 1996 in Sabin 2018; NatureServe 2020 and references therein; USFWS 2014, 2020; Woodin, III 1950).

INFECTIOUS AGENTS

Full name: The types, abundance, and distribution of infectious agents and their vectors. This habitat element refers to the spectrum of viruses, bacteria, fungi, and parasites that individual NMGS are likely to encounter during each life stage. The effects of disease and other infectious agents on NMGS are poorly understood.

Non-lethal infections may make the affected individuals vulnerable to mortality from other causes, and other sources of stress correspondingly may increase susceptibility to disease (Blaustein et al. 2003; Davidson et al. 2002). SFD has been spreading through the eastern and mid-western United States. There had been no cases reported in Arizona, although Barber et al. (2016) reported an instance of SFD in a Texas blotched water snake (Nerodia erythrogaster transversa), which uses similar aquatic habitat as NMGS. However, in 2019,

4-8 Chapter 4 – Habitat Elements

researchers from NAU confirmed the presence of SFD from two northern Arizona snakes (see NAU Gartersnake Project website: https://news.nau.edu/snake- fungal-disease/#.X9TW16pKgci). Also, a new, non-native snake, the southern watersnake, has been found in the LCR (at Mittry Lake and Imperial Dam) and may spread unknown pathogens into NMGS habitat should the species disperse more widely (Myatt 2016; B. Sabin 2021, personal communication).

PREDATORS

Full name: The abundance and distribution of species that depredate NMGS during both life stages. This habitat element refers to a set of closely related variables that affect the likelihood that different kinds of predators will encounter and successfully prey on NMGS during each life stage. The variables of this element include the species and sizes of the fauna that prey on NMGS during different life stages and the density and spatial distribution of these fauna in the streams and wetland corridor habitats used by NMGS. Susceptibility to predation is related, in part, to aquatic and terrestrial herbaceous vegetation cover.

Avian predators include wading birds, common mergansers, belted kingfishers, and raptors (USFWS 2014). Emmons et al. (2016b) reported an incidence of red- tailed hawk predation of an adult NMGS that had been previously transmitted and found in a regurgitated pellet along Tonto Creek, Arizona. NMGS are preyed upon by reptiles such as spiny softshell turtles, other snakes such as kingsnakes, whipsnakes, and regal ring-necked snakes (Sabin 2018 and references therein; USFWS and references therein); mammalian predators include raccoons, skunks, coyotes, foxes, and bobcats (Brennan et al. 2009 in USFWS 2014; Rosen and Schwalbe 1988). Researchers at NAU have also hypothesized that population expansion of the river otter (Nowak et al. 2011) and common black hawk (Buteogallus anthracinus) (Etzel et al. 2014) may affect NMGS populations in future years (see NAU Gartersnake Project website: https://news.nau.edu/snake- fungal-disease/#.X9TW16pKgci).

A number of non-native species are important predators of NMGS. Non-native crayfish are currently found in a majority of the subbasins where NMGS are found. Larger crayfish can be significant predators on smaller juvenile snakes (USFWS 2014 and references therein). Bullfrogs also share habitat with NMGS, both preferring calmer backwaters, pools, and side channels in stream habitat. Adult and subadult bullfrogs, depending on size, regularly prey on NMGS. Bullfrogs are widespread throughout Arizona and the range of the gartersnake and have been implicated in the collapse of the NMGS population and prey base at the San Bernardino NWR (Rosen and Schwalbe 1988). One result of high levels of bullfrog predation is a skewing of the age classes of NMGS—since it is easier for bullfrogs to eat smaller snakes, with high bullfrog numbers, the snake

4-9 Northern Mexican Gartersnake (Thamnophis eques megalops) (NMGS) Basic Conceptual Ecological Model for the Lower Colorado River

population skews toward larger, older snakes, with fewer newborn and juvenile age classes (Holm and Lowe 1995 in USFWS 2014; Rosen and Schwalbe 1988). Feral hogs are another potential predator. They are abundant at some of the sites where NMGS have been found (e.g., Beal Lake Conservation Area and Topock Marsh). Although primarily plant feeders, browsing on leaves and twigs and rooting around for underground tubers, they readily, opportunistically prey on invertebrates and other animals, including snakes (U.S. Department of Agriculture 2019).

Non-native fishes also prey on NMGS, particularly brown trout and other larger predatory spiny-rayed fish that have been introduced to many of Arizona’s aquatic habitats, such as largemouth bass (Holycross et al. 2006; New Mexico Department of Game and Fish 2013; Rosen and Schwalbe 1988; Young and Boyarski 2013).

SUBSTRATE

Full name: The types (texture) and stability of the substrate at locations with or potentially suitable for NMGS habitat within the LCR ecosystem and spatial and temporal variation in these features. This habitat element refers to the particle size distribution of the benthic and riparian sediment, including upland soil friability at locations with or potentially suitable for NMGS habitat; the frequency and magnitude of substrate movement, erosion (scour), and deposition (burial); and the presence of surface features such as rock piles, logs and other large woody debris, or smaller flood-deposited debris piles. These features may affect the suitability of individual locations for NMGS during different life stages, providing brumation sites, basking locations, or protection from predators. Such features may also be important for species that may prey on or compete with NMGS, or serve as prey themselves, and may affect or be affected by other habitat elements such as herbaceous vegetation and aspects of the hydrologic regime.

Although NMGS likely select habitat based in part on the availability of surface features mentioned above, it is not known whether NMGS select among sites or habitats (aquatic or terrestrial) based on substrate type. Research on LLFR (Cotten and Leavitt 2014; Leavitt et al. 2017), an important prey species of NMGS, has shown that LLFR do not select habitat based on substrate type, and this may be the same for gartersnakes (See Braun et al. [2019] for discussion of substrate and LLFR).

It is known that NMGS often use animal burrows for brumation, making terrestrial soil friability an important habitat feature. It is also known that in aquatic habitats, sedimentation does affect visibility for searching for prey (see

4-10 Chapter 4 – Habitat Elements

below, “Turbidity”). Erosion of streamside soils into water and subsequent deposition can smother amphibian eggs, and suspended particles may clog amphibian and fish gills, depending on the intensity of erosion. Conversely, flow events that scour and thereby reopen pools can improve habitat for anurans (e.g., LLFR [Sredl et al. 1997]) or fishes. Both species groups are important prey for NMGS. Sedimentation may also change the water depth in marshes, creating new shallow water habitat or degrading habitat conditions, depending on the pattern of deposition. NMGS at the Havasu NWR are associated with shallow marshes, while researchers at the Bubbling Ponds Fish Hatchery observed NMGS foraging in deeper water (several meters in depth) (NAU 2020). Observations of the western aquatic gartersnake (Thamnophis couchii) in California found that water depth used for foraging was correlated with the age class of the snake, with younger snakes using shallower calmer waters and adults often foraging in deeper, faster-moving waters (Lind 1990); however, it is not known what water depth is optimal for NMGS for either life stage.

Sedimentation can result from runoff after severe storms, erosion after fire, and from the construction and maintenance of dams and other water control/diversion structures (USFWS 2014). Increased erosion and siltation can also result from the burrowing activity of non-native crayfish in streambanks (Fernandez and Rosen 1996 in USFWS 2014) and benthic disturbance from some non-native fishes.

SURFACE WATER CONNECTIVITY

Full name: The capacity of surface water connections to support or inhibit both upstream and downstream movement of aquatic or semiaquatic organisms, as determined by the spatial distribution of natural and artificial barriers, including falls and dry reaches, that may prevent or inhibit such movement. Surface water connections may include natural water courses, artificial water courses including diversion and drainage ditches and canals, and artificial water bodies such as impoundments. Surface water connectivity affects the taxonomic and genetic composition of the fish and frog assemblages in different reaches of a stream network by affecting the ability of these fauna to move within the network, including escaping reaches of the network that have become isolated by interruptions in surface water connectivity, returning to reaches previously isolated by such interruptions, or colonizing new habitat (e.g., cattle tanks) created by human activities (Benda et al. 2004; Fullerton et al. 2010; Levick et al. 2008; Meyer et al. 2007; Pringle 2003). For a given water- dependent species or group of species, the severity of fragmentation (loss of connectivity) in a network depends on the number and spatial extent of barriers, their relative placement within the network, and their permeability (ability to allow some passage of the subject fauna) (Fullerton et al. 2010).

4-11 Northern Mexican Gartersnake (Thamnophis eques megalops) (NMGS) Basic Conceptual Ecological Model for the Lower Colorado River

NMGS use terrestrial habitats for brumation, basking, ecdysis, and mating. However, they are semiaquatic (“mesic environment obligates”) and forage predominantly in or adjacent to permanent water bodies—in the water or in dense vegetation along streambanks or wetlands edges. There is little information about dispersal in NMGS, but it is likely that surface water connectivity would be important to the snakes—as the young disperse after birth, they would be expected to follow water courses given their need to be near permanent water. Some researchers (e.g., Wood et al. 2018) have expressed concerns over tendencies toward genetic isolation among NMGS populations in the LCR as a result of habitat loss and fragmentation of surface water connections.

Once NMGS become sexually mature, they may remain within an established home range. Although there are no data to support this conjecture, it is likely that historically, surface water connectivity was important to dispersal in the snakes, especially if NMGS were distributed along the entire LCR and its tributaries pre- dam construction.

TEMPERATURE

Full name: The maximum or minimum temperature in a habitat patch. This habitat element refers to the temperature extremes in NMGS habitats. NMGS are most active when the air temperature is between 22–33 °C (71–91 °F) (Rosen 1991). Concerning optimal water temperatures, Holm and Lowe (1995 in USFWS 2014) found that NMGS were active when water temperatures were between 22.1–22.5 °C (71.8–72.5 °F) at Scotia Canyon in the Huachuca Mountains, Arizona. Ruibal (1962 in Sredl 2018) found that eggs of LLFR (a preferred prey species of NMGS) in a population in southern California developed fully in waters between 11 and 29 °C (≈52–84 °F) but did not find LLFR egg masses in waters warmer than 25 °C (77 °F). During the winter months, when colder temperatures prevail, they brumate individually or, on occasion, communally in underground burrows (see chapter 3, “Brumation” for more details). On warmer winter days, they may emerge to bask in the sun; during cooler summer nights, they may seek warm rocks on which to bask overnight (see chapter 3, “Basking” for more details). Semiaquatic snakes may also use the water to thermoregulate as air and water temperatures fluctuate during the day (Nelson and Gregory 2000; Osgood 1970; Sprague and Bateman 2018). It is not known if NMGS aestivate during times of excessive heat.

Extreme ambient summer temperatures that excessively warm aquatic habitats used by NMGS may affect water chemistry, species interactions, and the aquatic prey base, affecting NMGS survival. (See Braun et al. 2019 for additional information on synergetic effects of higher temperatures on water temperature and survival of anurans.)

4-12 Chapter 4 – Habitat Elements

TURBIDITY

Full name: The turbidity of water in a wetland, pond, or stream, including its magnitude and spatial and temporal distributions. Gartersnakes are visual predators, searching for prey as they swim along streambanks or among dense emergent vegetation. As such, they need relatively clear water in order to locate their prey (Sabin 2018; USFWS 2014), although they are aided by sensory cues (chemoreception). Water turbidity can have many causes (e.g., dredging and other construction or maintenance activities, sedimentation from runoff from weather events, recreation [motorboat disturbances of shallows – see Asplund 2000], substrate characteristics, dense algal blooms that block sunlight from penetrating the water column, and invasive common carp (Cyprinus carpio) that churn the benthic substrate during feeding and spawning [Cucherousset and Olden 2011; Lougheed et al. 1998]). Note that the density and arrangement of emergent marsh vegetation can also affect water turbidity caused by weather-related factors such as wind.

VEGETATION STRUCTURE

Full name: The species composition, density, spatial and temporal distributions, and vertical structure of aquatic or terrestrial herbaceous vegetation and shrubs. NMGS aquatic habitats occur embedded in a diversity of upland habitats: pine or oak woodlands, grasslands with sacaton (Sporobolus spp.), or in riparian habitat with willows and cottonwood (Brennan and Holycross 2006; New Mexico Department of Fish and Game 2013). As long as suitable aquatic and riparian habitat is available adjacent to uplands, the upland vegetation community type (species composition) may be of less importance to NMGS.

Of key importance is the presence of thick herbaceous vegetation adjacent to permanent water bodies that provides foraging habitat while offering protection from predation and cover during basking (Szaro et al. 1985 in Center for Biological Diversity 2003). Equally important may be the vegetation structure of overwintering sites—radio telemetry research by Sprague and Bateman (2018) found that NMGS had precise vegetative habitat requirements with males favoring locations farther from water with high amounts of vegetation, including shrubs, and females selecting burrows in areas with a high percentage of canopy cover 1 m in height. See above, “Brumation” and “Substrate” for further discussion of brumation sites and the importance of burrows and other components≥ of habitat structure.

4-13 Northern Mexican Gartersnake (Thamnophis eques megalops) (NMGS) Basic Conceptual Ecological Model for the Lower Colorado River

Sprague and Bateman (2018) observed that during the active season (March through October) NMGS frequented sloping banks in low vegetation along streams or wetland edges. Females were often found near shrubs. In this study, NMGS were also found in marshy habitats (Sprague and Bateman 2018). Trapping by the USFWS at the Havasu NWR found them in shallow, dense marshes and not in areas of open water (e.g., found at the willow marsh site at the Beal Lake Conservation Area and Glory Hole but not at the nearby Beal Lake Slough, South Dike Interior, South Dike East, or Five Mile Landing (Bourne 2020). Other studies found adult NMGS in areas with less vegetative cover, often foraging in deeper water (Emmons and Nowak 2016 in NAU 2020).

Typical aquatic vegetation along streams and in and adjacent to cienegas in Arizona that support NMGS includes knot grass (Paspalem distichum), spike rush (Eleocharis spp.), bulrush (Scirpus spp.), and cattails. Deer grass (Muhlenbergia spp.), sacaton, velvet mesquite (Prosopis velutina) with Fremont cottonwood (Populus fremontii), willow (Salix spp.), and seep willow (Baccharis salicifolis) are also commonly present (Rosen and Schwalbe 1988). NMGS will also use bank vegetation with introduced species such as Bermudagrass (Cynadon dactylon) (Rosen and Schwalbe 1988).

(Note: Cotten and others [Cotten 2011; Cotten and Grandmaison 2013; Cotten and Leavitt 2014] suggest that high densities of some herbaceous species – common reed, cattails, and bulrush – reduce habitat quality for LLFR [see Braun et al. 2019 for discussion]. As this leopard frog is a preferred prey of NMGS, these plants may not provide ideal habitat for NMGS that rely on native frog prey.) Instream vegetation, in particular filamentous algae, is also an important component of NMGS habitat. It provides food for anuran tadpoles (prey of NMGS) and cover for snakes and anurans. However, foraging crayfish can remove this structural component (Creed 1994; Fernandez and Rosen 1996 in USWFS 2014; also see chapter 5 “Nuisance Species Introduction & Management.”)

4-14

Chapter 5 – Controlling Factors

Controlling factors consist of environmental conditions and dynamics, both natural and anthropogenic, that affect the abundance, spatial and temporal distributions, and quality of habitat elements. Controlling factors may also significantly directly affect some critical biological activities or processes. A hierarchy of controlling factors exists, with long-term dynamics of climate and geology at the top. However, this CEM focuses on 11 immediate controlling factors that are within the scope of potential human manipulation, particularly manipulation by the LCR MSCP and its conservation partners.

The 11 controlling factors identified in this CEM do not constitute individual variables; rather, each identifies a category of variables (including human activities) that share specific features that make it useful to treat them together. In particular, each controlling factor covers activities with similar effects or management implications across multiple life stages and across multiple species of concern to the LCR MSCP. Categorizing such activities together across multiple species and multiple life stages of these species makes it easier to compare and integrate the CEMs across the LCR MSCP.

Table 4 lists the 11 controlling factors included in the NMGS conceptual ecological model and the habitat elements they directly affect. Controlling factors affect habitat elements indirectly, as well, through their effects on other controlling factors or through the cascading effects of habitat elements on each other.

5-1 Northern Mexican Gartersnake (Thamnophis eques megalops) (NMGS) Basic Conceptual Ecological Model for the Lower Colorado River

Table 4.—NMGS controlling factors and the habitat elements they are proposed to directly affect among the two NMGS life stages (Xs indicate which controlling factors may affect each habitat element.)

Habitat element 

 Controlling factor Anthropogenic disturbance Competitors Environmental contaminants Food availability Genetic diversity regime Hydrologic Infectious agents Predators Substrate Surface water connectivity Temperature Turbidity Vegetation structure

Fire management X X X Fisheries introduction & fisheries management X X X X Grazing X Nuisance species introduction & management X X X X X X Off-site land management & use X X On-site vegetation management X X X On-site water management X X Recreational activities X Site maintenance X X Wastewater & other contaminant inflows X X Water storage-delivery system design & operations X X Note: Controlling factors affect turbidity indirectly via effects on substrate. Ambient temperature and genetic diversity are not directly affected by any controlling factor. Temperature is determined by regional climate and local weather conditions as well as indirect effects of controlling factors.

5-2 Chapter 5 – Controlling Factors

FIRE MANAGEMENT

Full name: Any fire management (whether prescribed fire or fire suppression) that could affect NMGS or their habitat. Effects may include creation of habitat that supports or excludes NMGS; a reduction in the food supply/prey base of fishes, frogs or invertebrates; or support of species that pose threats to NMGS such as predators, competitors, or carriers of infectious agents.

The LCR MSCP and other land management agencies along the LCR and Bill Williams River valleys may use prescribed fire as a management tool (e.g., burning marshes to remove cattails) and actively manage wildfires through fire suppression and the construction of fuel breaks. Wildfire is a natural type of disturbance in the riparian plant communities of the LCR valley, and wildfires today also occur through human accidents (Conway et al. 2010; Mac Nally et al. 2004; Meyer 2005; LCR MSCP 2018a). In fact, wildfires have occurred recently at LCR MSCP conservation areas (Hunters Hole, Cibola Valley Conservation Area, and Yuma East Wetlands) and in riparian habitat at the Havasu NWR, Cibola National Wildlife Refuge-Island Unit, and adjacent to the Laguna Division Conservation Area (J. Hill and C. Ronning 2018, joint personal communication with D. Braun; LCR MSCP 2018a, 2018b).

Fire may affect water chemistry and turbidity (see chapter 4, “Environmental Contaminants” and “Substrate”) through the release of soluble and insoluble materials into the water after a burn; however, the effect will vary depending on the severity and intensity of the fire (and whether wildfire or prescribed burn), season of the year, type of aquatic system, surrounding upland habitat, and subsequent precipitation events that increase runoff from burned areas if offsite (Bixby et al. 2015; Meixner and Wohlgemuth 2004; New Mexico Environment Department 2014; Tecle and Neary 2015). Whether fire affects NMGS depends on many factors, including which habitats are involved, whether they are streams, pools, larger wetlands, or adjacent uplands; the availability of deep hibernacula or burrows in which to shelter in upland habitats; and/or whether prey populations are affected. Without deep hibernacula or burrows in which to shelter, fire may injure or kill NMGS. Shallow burrows may not provide sufficient protection from heat and smoke. Proximity to safe sites is important—snakes are expected to actively move and avoid danger, but they may not be able to get out of harm’s way if a marsh burns and they are unable to access refuge in unburned areas in the marsh or uplands (C. Ronning 2021, personal communication).

In New Mexico populations, Chiricahua leopard frogs were eliminated from one site after a catastrophic wildfire followed by a rain event resulted in erosion and sedimentation that altered the canyon breeding sites for this species—an important prey item for NMGS.

5-3 Northern Mexican Gartersnake (Thamnophis eques megalops) (NMGS) Basic Conceptual Ecological Model for the Lower Colorado River

Climate change is projected to affect fire frequency in the Southwestern United States (Westerling et al. 2006) and specifically along the LCR (USFWS 2013).

FISHERIES INTRODUCTION & FISHERIES MANAGEMENT

Full Name: The types, frequencies, and duration of activities carried out by the LCR MSCP, USFWS, States, and Tribes to manage native fish species and introduced recreational fisheries along the LCR channel and off-channel habitats, including in conservation areas and wildlife refuges, and the legacies of past and unofficial introductions. Examples of such activities include fish monitoring, seining, electroshocking (may disturb resting or foraging gartersnakes if present), dredging to deepen backwaters to improve razorback sucker and bonytail habitat in this section, or construction of riprap shoreline cover (depending on the location) (see LCR MSCP research and monitoring fisheries activities reports at http://www.lcrmscp.gov/fish/fish_res_mon.html). Fisheries management may alter the species composition of the aquatic community and introduce competitors into the system (e.g., stocked bonytail may compete with NMGS for the same prey). Depending on the management activities that are implemented, improvement of stream and backwater habitats and water quality for fishes could benefit NMGS that also use these habitats. However, use of rotenone to remove non-native fish species would clearly be detrimental to the NMGS prey base. (See below, “Nuisance Species Introduction & Management”)

Rotenone is a piscicide that has been widely and regularly used in many streams and water bodies in Arizona (and elsewhere) over multiple years. It is considered an important tool for management and restoration of native fish populations. However, the length of the interim period between when fish populations are removed/killed and when restocking of new native fishes occurs—essentially, the time where the water body is fishless—does affect the prey base for NMGS. Because there are strict environmental standards in place related to fish management and effects on non-target reptiles, the use of rotenone is not considered a “substantial” risk in Arizona (USFWS 2014).

Mechanical methods for managing fish populations include the use of electroshocking and minnow traps. Although electroshocking will temporarily paralyze any NMGS in the water within range of the electroshocker, this is not considered a “substantial” threat (USFWS 2014). The use of minnow traps, either by managers or by recreational fishers – (see below, “Recreational Activities”), if flooded or improperly deployed, may result in the death of NMGS by drowning

5-4 Chapter 5 – Controlling Factors

(if traps are completely submerged below water and access to air unavailable) or by entrapment in wire mesh and subsequent drowning or predation. This is more common with juvenile snakes due to their small size and tendency to try and wriggle through the traps to get at bait fish inside. The USFWS (2014) reports that there is not enough information to determine the magnitude of this threat; however, the loss of any reproductive females in small or declining populations could be problematic. For this reason, best management practices are being implemented to guide the use of minnow traps to minimize any risk of harm to snakes (C. Ronning 2020, personal communication).

Fish hatcheries raise and release both native and non-native sport fishes into Arizona’s waters. Two of the State fish hatcheries (Page Springs and Bubbling Ponds) support populations of NMGS, and their land management activities could affect these populations (Sabin 2018; USFWS 2014). To address this possibility, hatchery staff receive snake awareness training and know the places they are most likely to encounter the species. In addition, there are strict protocols in place that discuss guidelines for management and maintenance activities (e.g., timing of weeding and mowing, the use of heavy equipment near snakes, and safe speed limits around ponds) (B. Sabin 2020, personal communication).

The LCR MSCP has lead responsibility for the management of covered native fish species in the LCR ecosystem under the Habitat Conservation Plan (LCR MSCP 2004). In turn, the States bordering the LCR recognize and oversee the sport fisheries along the LCR, its reservoirs and connected backwaters, and its tributaries. The fishes recognized by these States as sport fishes include intentionally introduced and/or stocked species and accidental introductions. The States and recreational fishers have also introduced bait and forage species to support the sport fisheries. These bait and forage species may be caught as sport fishes and may also be considered (by the States) to be nuisance species. Arizona lists the official sport fishes for the State (https://www.azgfd.com/fishing/species/) and State records for any caught along the LCR (https://www.azgfd.com/Fishing/records/). Recreational fishers also may transplant desired sport fishes to water bodies where management agencies have tried to keep them out (LCR MSCP 2017; Wolff et al. 2012).

GRAZING

Full name: The grazing activity on riparian habitats along the LCR and in surrounding areas that could affect NMGS or their habitat. Overgrazing by cattle (Bovidae), burros (Equus asinus), or mule deer (Odocoileus hemionus) across the arid Southwestern United States has substantially degraded riparian habitat in many areas (see Appendix G in USFWS 2002). Reclamation staff and researchers have observed mule deer and burros browsing on some LCR sites,

5-5 Northern Mexican Gartersnake (Thamnophis eques megalops) (NMGS) Basic Conceptual Ecological Model for the Lower Colorado River

which may affect vegetation communities if populations change such that overgrazing occurs. Wild burros, in particular, are present along the Bill Williams River and prevalent on Planet Ranch. They have created trails in vegetation, dug holes to groundwater in sandy areas, and heavily grazed available vegetation at Planet Ranch and along sections of the river where snakes may occur. To reduce the impact of these animals, nuisance burro gathers and removal operations have been conducted at Planet Ranch and may be conducted in the future if needed (Bureau of Land Management 2020; C. Ronning 2020, personal communication).

High-density cattle grazing may remove the dense streamside vegetation and other cover on which gartersnakes depend for thermal cover and to avoid predation (Rosen and Schwalbe 1988; USFWS 2014 and references therein). The removal of vegetative cover in small, isolated populations could cause extirpation in one season. Small, isolated populations are extremely vulnerable, especially where populations are also isolated by dewatered stretches and overland migration is minimal (NatureServe 2020; Rosen and Schwalbe 1988).

Krueper (1993) and Krueper et al. (2003) report that fencing cattle out of sensitive riparian habitats in the San Pedro Riparian National Conservation Area led to improved habitat quality for riparian birds, which would likely benefit NMGS inhabiting dense vegetation in similar riparian environs. Other reports document greater abundance of western gartersnakes (Thamnophis elegans vagrans) in locations where grazing activity was controlled (Szaro et al. 1985 in USFWS 2014). Apart from vegetation removal, overgrazing may also result in direct mortality to snakes from trampling (Chapman 2005 in USFWS 2014; Szaro et al. 1985 in USFWS 2014).

Isolated populations of NMGS would be most vulnerable to habitat degradation from overgrazing (Rosen and Schwalbe 1988), and unmanaged livestock grazing is a concern in some Mexican populations (USFWS 2014). However, in the LCR and other subbasins where NMGS are present, overgrazing by any of these species currently is not considered a major threat. Land managers today are better at keeping cattle out of riparian areas (USFWS 2014).

NUISANCE SPECIES INTRODUCTION & MANAGEMENT

Full name: The intentional or unintentional introduction of nuisance species (animals and plants as well as microbes) and/or their control that affect NMGS survival and reproduction. The nuisance species may poison, infect, prey on, compete with, or present alternative food resources for NMGS during

5-6 Chapter 5 – Controlling Factors

one or more life stages; cause other alterations to the riparian food web that affect NMGS; or affect physical habitat features such as vegetation cover or water chemistry. For example, encroachment of non-native species such as common reed or saltcedar (Tamarix sp.) can alter riparian vegetation structure, as will the use of herbicides to control these invasive plants. Non-native species are often intentionally introduced, but unintentional, seemingly minor introductions as when an individual dumps unwanted aquarium fish into ponds can also be problematic (B. Sabin 2020, personal communication).

The introduction of exotic species, including non-native American bullfrogs, northern crayfish, and non-native spiny-rayed fish is considered the “most significant and pervasive of all threats” to NMGS populations throughout their range (USFWS 2014). In 2016, seven non-native southern watersnakes were trapped in Mittry Lake near Yuma, Arizona, confirming the presence of a thriving population (Myatt 2016). Not only do these and other non-natives compete for and reduce the abundance of native fishes and amphibians on which NMGS depend, they may also prey directly on young gartersnakes, causing populations to decline (Holycross et al. 2006; New Mexico Department of Game and Fish 2013; Paroz et al. 2009; Rosen et al. 2001; as summarized from multiple sources in USFWS 2014; Rosen and Schwalbe 1988).

Introduced species often alter critical habitats. The burrowing activities of non- native crayfish and some non-native fish can affect the benthic substrate, reducing habitat quality for NMGS. Crayfish have also been shown to remove vegetation cover in aquatic habitats through their foraging activity, altering the vegetation structure needed by NMGS and their prey (Creed 1994; Fernandez and Rosen 1996 in USWFS 2014).

Feral hogs may also be a concern for NMGS. Hogs occur on the Havasu NWR and have been sighted elsewhere in the LCR. Through their rooting and wallowing activities, they disturb soil and increase erosion, destroy vegetation, and negatively affect water quality. Hogs may also compete with NMGS for prey and may opportunistically prey on the snakes. The U.S. Department of Agriculture – Animal and Plant Health Inspection Service (APHIS) is working to eradicate them from the Havasu NWR (Neskey 2018; U.S. Department of the Interior 2016); it is unknown what effect they may have on NMGS at this and other locations where they may occur.

The spread of saltcedar has been an ongoing management issue in the LCR, particularly as it affects habitat for riparian birds. Little is known about any potential effects on NMGS and habitat use. However, replacement of native stands with saltcedar following a fire may be a potential future concern because dense saltcedar litter may blanket the ground and exclude native ground cover.

5-7 Northern Mexican Gartersnake (Thamnophis eques megalops) (NMGS) Basic Conceptual Ecological Model for the Lower Colorado River

Saltcedar presence may also limit habitat suitability in currently unoccupied reaches of the Bill Williams River National Wildlife Refuge, should NMGS disperse to those areas on the refuge (C. Ronning 2020, personal communication).

Lastly, the use of chemicals to control introduced species can be problematic (see below, “Off-Site Land Management & Use” and “On-Site Vegetation Management.” For example, the use of algicides to control non-native golden algal (Prymnesium parvum) blooms may affect other organisms in aquatic systems (Sallenave 2010). (Note: Under certain environmental conditions, blooms of golden algae may produce a toxin harmful to many gilled aquatic organisms [Brooks et al. 2011; Roelke et al. 2011; Sallenave 2010]). If this were to occur, many prey species of NMGS (e.g., amphibian larvae, fishes) could be affected.

OFF-SITE LAND MANAGEMENT & USE

Full name: Activities occurring on lands adjacent to streams, rivers, pools and backwater habitats that may affect NMGS or their prey. Agricultural and other land use activities on adjacent lands may result in erosion and sediment deposition into water bodies, streams, and backwater habitats utilized by NMGS. Increased nutrient loads may lead to eutrophication, altering water chemistry and reducing water quality and increasing turbidity.

The use of pesticides/herbicides was listed as a potential threat to NMGS by the USFWS (2014). Wetlands often accumulate toxins from surrounding watershed lands. In particular, drainage waters from agricultural lands that include pesticides may cause sublethal poisoning of NMGS via ingestion of treated insects or fishes that bioaccumulate toxins and/or a reduction in the food supply if pollutants are toxic to prey of NMGS.

ON-SITE VEGETATION MANAGEMENT

Full name: The types, frequencies, and durations of actions taken to manage the taxonomic composition, abundance, condition, and spatial distribution of vegetation at locations with, or potentially suitable for, NMGS habitat within the LCR ecosystem. This factor addresses vegetation management at the scale of individual sites managed to achieve specific habitat or hydrologic goals. The LCR MSCP and other land managers along the LCR and Bill Williams River valleys use a range of methods to manage vegetation on lands under their authorities, including prescribed fire, surface irrigation and subirrigation, planting, fertilizing, thinning or hand removal of dead or down vegetation or

5-8 Chapter 5 – Controlling Factors

the application of herbicides, disking and plowing, mowing, or the use of any associated vehicles (LCR MSCP 2014, 2018a; C. Ronning 2020, personal communication). Agencies and irrigation and drainage districts may also remove vegetation to maintain roads and canals under their authorities. Each of these activities can result in injuries and mortalities if NMGS are not moved out of harm’s way during management. Threats and minimization measures have been developed for the Beal Lake Conservation Area.

ON-SITE WATER MANAGEMENT

Full name: Water management at the scale of individual sites intentionally supplied by surface water diversions and/or groundwater withdrawals or passively supplied by seepage from adjacent surface water diversions. NMGS can occupy manmade systems, including water canals, earthen stock tanks, and impoundments (Conant 2003; Degenhardt et al. 1996 in Sabin 2018; Woodin, III 1950). This factor addresses the types, frequencies, and durations of official activities that affect the delivery and distribution of regulated water within sites managed to support NMGS habitat. In particular, this addresses water management for ponds and wetlands maintained by the LCR, including the Beal Lake Conservation Area at Havasu Wildlife Management Area. NMGS were first observed at Beal Lake Conservation Area in 2015 (Sabin 2018), and surveys at the site in 2019 (Bourne 2020) resulted in the capture of NMGS at Willow Marsh and Glory Hole. Dewatering of sites used by NMGS will negatively affect NMGS populations in part by eliminating the prey base (USFWS 2014). Once established at a location with a hydrologic regime determined by on-site water management, NMGS will likely depend on that management to maintain suitable hydrologic conditions for their continuing presence at the site.

In some cases, on-site water management may be used to mitigate negative impacts associated with other forms of habitat management (e.g., draining wetlands prior to application of prescribed fire to allow snakes or other species to move and find refuge elsewhere) (C. Ronning 2021, personal communication). On-site water management may also include actions to reduce or terminate water applications at a site (e.g., to reallocate water to other sites within the limits of Reclamation’s water rights).

RECREATIONAL ACTIVITIES

Full name: The disturbance to NMGS from recreational activities. Even non- consumptive human activity can have negative effects on wildlife (reviewed by

5-9 Northern Mexican Gartersnake (Thamnophis eques megalops) (NMGS) Basic Conceptual Ecological Model for the Lower Colorado River

Boyle and Samson 1985). This is a broad category that encompasses the types of recreational activities (e.g., boating, fishing, horseback riding, wildlife viewing, camping, and off-road vehicle use) as well as the frequency and intensity of those activities. The impacts may consist of disturbance and habitat alteration (e.g., the use of motorized watercraft can affect turbidity and water chemistry) (Asplund 2000). In the LCR, the backwaters of the Havasu NWR and Topock Marsh, as well as the river waters at the Bill Williams River National Wildlife Refuge, are closely regulated to protect nesting waterbirds from boat wake (https://www.fws.gov/refuge/Havasu/activities/boating.html), which may also benefit other species using aquatic habitat such as NMGS. Increased recreational use of an area may attract predators if resulting garbage is not managed properly. The risk of wildfires may increase (see above, “Fire Management”), streamside vegetation important to NMGS may be trampled, and there may be more incidences of human/snake interaction (Rosen and Schwalbe 1988; as summarized from multiple sources in USFWS 2014).

SITE MAINTENANCE

Full name: Any site maintenance activities related to infrastructure, such as road maintenance, that affect NMGS habitat. Road construction or grading, mowing, and/or vegetation removal from the sides of roads and adjacent berms or waterways, and the removal and maintenance of riprap, may interfere with or kill basking or resting NMGS. Road activity in particular has been identified as a major problem for snakes (Sabin 2018; USFWS 2014 and references therein).

WASTEWATER & OTHER CONTAMINANT INFLOWS

Full name: Contamination from main stem diversions that may include contaminants such as high concentrations of heavy metals. Areas along the LCR have been found to harbor high concentrations of contaminants that may negatively affect species of concern directly or indirectly via effects on habitat or food availability. For example, selenium monitoring is a component of LCR marsh management and backwater restoration, especially in more southern sections of the LCR – Imperial Reservoir, Cibola Lake and Topock Marsh. High mercury levels have been detected along the Bill Williams River, including Alamo Lake, and in past years, Alamo Lake also has been found to have high NH3 levels due to deposition of organic materials (Ajami et al. 2005).

5-10 Chapter 5 – Controlling Factors

Pesticide use is also very common along the river and may contaminate surface water (B. Sabin 2021, personal communication). It is not known what effect these contaminants have on NMGS.

WATER STORAGE-DELIVERY SYSTEM DESIGN & OPERATIONS

Full name: The types, frequencies, and durations of coordinated basin-scale activities that regulate the elevation of surface water along the LCR main stem. The Colorado River through the LCR valley consists of a chain of reservoirs separated by flowing reaches. The water moving through these systems is highly managed to allow for storage and delivery (diversion) to numerous international, Federal, State, Tribal, and municipal users and for hydropower generation. This system of water management and its infrastructure, together with regulated discharges from the Upper Colorado River Basin and local weather conditions, determine surface water elevations and groundwater elevations along the main stem LCR and its floodplain (LCR MSCP 2004). The dams along and above the LCR also trap essentially all of the sediment and both coarse and fine organic matter that would have flowed past their locations prior to their construction. River regulation and entrenchment of the river along flowing reaches have eliminated almost all opportunities for the river to deliver pulses of water onto its former floodplain and have altered water table elevations throughout the valley. Reclamation, the USFWS, and other agencies have rights to use some of the water in the LCR for the use of that water on lands managed as wildlife habitat (LCR MSCP 2014, 2018a).

This factor includes river and off-channel water management, including pumping of groundwater and diversion of river water to manage water levels in refuge ponds, as well as dewatering and flushing of marsh or other aquatic habitats. The amount of water, flooding frequency, flow, water depth and stability, etc., all affect development and maintenance of streams, backwater habitats, and fringing marshes with sufficient water depth, water quality, vegetation density, and species composition for NMGS. In the LCR, currently NMGS populations are known to exist at the Beal Lake Conservation Area, Topock Marsh, and locations in the Bill Williams River watershed on the Bill Williams River, Big Sandy River, and Santa Maria River (O’Donnell et al. 2019; Sabin 2018) (see above, “On-Site Water Management”).

5-11

Chapter 6 – Conceptual Ecological Model by Life Stage

This chapter contains two sections, each presenting the CEM for a single NMGS life stage. For each life stage, the text and diagrams identify its life-stage outcomes; its critical biological activities and processes; the habitat elements that support or limit the success of its critical biological activities and processes; the controlling factors that determine the abundance, distribution, and other important qualities of these habitat elements; and the causal links among them.

The model for each life stage assesses the character and direction, magnitude, predictability, and scientific understanding of each causal link based on the following definitions (see attachment 1 for further details):

• Character and direction categorizes a causal relationship as positive, negative, or complex. “Positive” means that an increase in the causal node results in an increase in the affected node, while a decrease in the causal node results in a decrease in the affected node. “Negative” means that an increase in the causal node results in a decrease in the affected element, while a decrease in the causal node results in an increase in the affected node. Thus “positive” or “negative” here do not mean that a relationship is beneficial or detrimental. The terms instead provide information analogous to the sign of a correlation coefficient. “Complex” means that there is more going on than a simple positive or negative relationship. Positive and negative relationships are further categorized based on whether they involve any response threshold in which the causal agent must cross some value before producing an effect. In addition, the “character and direction” attribute categorizes a causal relationship as uni- or bi-directional. Bi-directional relationships involve a reciprocal relationship in which each node affects the other.

• Magnitude refers to “… the degree to which a linkage controls the outcome relative to other drivers” (DiGennaro et al. 2012). Magnitude takes into account the spatial and temporal scale of the causal relationship as well as the strength (intensity) of the relationship at any single place and time. The present methodology separately rates the intensity, spatial scale, and temporal scale of each link on a three-part scale from “Low” to “High” and assesses overall link magnitude by averaging the ratings for these three. If it is not possible to estimate the intensity, spatial scale, or temporal scale of a link, the subattribute is rated as “Unknown” and ignored in the averaging. If all three subattributes are “Unknown,” however, the overall link magnitude is rated as “Unknown.” Just as the terms for link character provide information analogous to the sign of a correlation coefficient, the terms for link magnitude provide information analogous to the size of a correlation coefficient.

6-1 Northern Mexican Gartersnake (Thamnophis eques megalops) (NMGS) Basic Conceptual Ecological Model for the Lower Colorado River

• Predictability refers to “… the degree to which current understanding of the system can be used to predict the role of the driver in influencing the outcome. Predictability … captures variability… [and recognizes that] effects may vary so much that properly measuring and statistically characterizing inputs to the model are difficult” (DiGennaro et al. 2012). A causal relationship may be unpredictable because of natural variability in the system or because its effects depend on the interaction of other factors with independent sources for their own variability. Just as the terms for link character provide information analogous to the sign of a correlation coefficient, the terms for link predictability provide information analogous to the size of the range of error for a correlation coefficient. The present methodology rates the predictability of each link on a three-part scale from “Low” to “High.” If it is not possible to rate predictability due to a lack of information, then the link is given a rating of “Unknown” for predictability.

• Scientific understanding refers to the degree of agreement represented in the scientific literature and among experts in understanding how each causal relationship works—its character, magnitude, and predictability. Link predictability and understanding are independent attributes. A link may be highly predictable but poorly understood or poorly predictable but well understood. The present methodology rates the state of scientific understanding of each link on a three-part scale from “Low” to “High.”

Constructing a CEM for each life stage involves identifying, assembling, and rating each causal link one at a time. Analyses of the resulting information for each life stage can then help identify the causal relationships that most strongly support or limit life-stage outcomes, support or limit the rate of each critical biological activity or process, and support or limit the quality of each habitat element, as that element affects other habitat elements or affects critical biological activities or processes. Analyses also can help identify which, among these potentially high-impact relationships, are not well understood.

All potential causal links—among controlling factors, habitat elements, critical biological activities and processes, and life-stage outcomes—affecting each life stage are recorded on a spreadsheet. This spreadsheet is then used to record information on the character and direction, magnitude, predictability, and scientific understanding for each causal link, along with the underlying rationale and citations, for each life stage. Software tools developed in association with these CEMs then allow users to generate a “master” diagram for each life stage from the data in the spreadsheet—or, more usefully, to query the CEM spreadsheet for each life stage and generate diagrams that selectively display query results concerning that life stage.

6-2 Chapter 6 – Conceptual Ecological Model by Life Stage

This report includes the master diagram for each life stage. The master diagrams display all causal links, of all character types and directions, magnitudes, predictabilities, and levels of understanding. The results can be visually complex but are included with this report to give the reader an overall sense of the CEM for each life stage.

The master CEM diagram for each life stage shows the controlling factors, habitat elements, critical biological activities and processes, and life-stage outcomes for that life stage. The diagram displays information on the character and direction, magnitude, predictability, and scientific understanding of every link. The diagrams use a common set of conventions for identifying the controlling factors, habitat elements, critical biological activities and processes, and life-stage outcomes as well as for displaying information about the causal links. Figure 2 illustrates these conventions.

Link Magnitude (line thickness) Controlling Fac tor High – thick line Medium – medium line

Link# Low – thin line Unknown – very thin line

Habitat Element Link Understanding (line color) High – black line

Link# Medium – blue line Low – red line

Critical Biological Activity or Process Link Predictability (link label color) High – black text Link# Medium – blue text Low – red text

Life-Stage Outcome Unknown – grey text

Figure 2.—Diagram conventions for LCR MSCP species conceptual ecological models.

6-3 Northern Mexican Gartersnake (Thamnophis eques megalops) (NMGS) Basic Conceptual Ecological Model for the Lower Colorado River

The conventions for displaying information about the causal links are as follows: Links are represented by arrows, the point of which indicates the direction of causation. Bi-directional causal links are represented by arrows with points at both ends. The thickness of the arrow represents link magnitude, and the color of the arrow represents link understanding. Each arrow has a label that uniquely identifies the link. The number to the left of the decimal place indicates the life stage (1…N), while the number to the right of the decimal place provides a unique index value for each link. The color of the label represents link predictability.

The discussions of each life stage in this chapter and of both life stages considered together in chapter 7 include analyses of the information contained in the spreadsheet. The analyses highlight causal chains that strongly affect the outcomes for each life stage and identify important causal relationships with high scientific uncertainty. The latter constitutes topics of potential importance for adaptive management investigation.

NMGS LIFE STAGE 1 – NEONATES/JUVENILES

As described in chapter 2, this life stage begins when NMGS adult females give birth to their live young. Upon birth, the young independently move away into the habitat and forage on their own. The life stage ends when the young attain sufficient size to become sexually mature. This life stage has one life-stage outcome (see figure 1): survival. Figure 3 (at the end of this section) presents the complete CEM for this life stage, showing all controlling factors, habitat elements, critical biological activities and processes, life-stage outcomes, and their linkages.

The CEM identifies five critical biological activities or processes affecting neonate/juvenile survival outcomes for this life stage as shown on figure 3. However, the CEM identifies no critical biological activities or processes to have a high-magnitude effect; rather, two—predation and foraging—do have medium- magnitude direct effects.

Predation is the most commonly proposed cause of mortality among NMGS in both life stages, thought to contribute significantly (along with habitat loss and competition) to NMGS extirpation across large areas of its historic range. As a result, link understanding is rated as high.

The three other critical biological activities or processes proposed to affect neonate/juvenile survival are brumation, chemical stress, and disease. The CEM proposes that disease and chemical stress have an effect of unknown magnitude, and brumation has a low-magnitude effect on survival. The effect of brumation is

6-4 Chapter 6 – Conceptual Ecological Model by Life Stage

of medium understanding, while chemical stress and disease are of low understanding due to a lack of studies of these possible effects in NMGS or any closely related species (see figure 3).

The CEM proposes that several of the critical biological activities and processes affect each other, possibly compounding their effects on survival. Specifically, disease and ecdysis are proposed to affect each other, and predation and foraging affect each other. The bi-directional effect of disease and ecdysis on each other is proposed to be unknown, with low understanding, for the relationship in either direction. The predation and foraging interactions are proposed to be of low magnitude, with medium understanding for the foraging/predation interaction. In addition, competition has a medium-magnitude effect on foraging, with medium understanding; basking has a low-magnitude effect on predation, with medium understanding; and edysis has an unknown-magnitude effect on chemical stress, with low understanding.

The CEM does not identify any habitat elements with high-magnitude effects on any critical biological activity or process. There are two habitat elements with medium-magnitude direct effects on critical activities or processes: competitors effect on competition and food availability effects on foraging. The effect of competitors on competition and of food availability on foraging are both very well understood in a general sense, although more specific research on interactions at LCR sites would be beneficial.

Two other habitat elements have medium-magnitude direct effects on predation: hydrologic regime and predators. The effect of hydrologic regime and predators on predation are of high understanding. The presence of suitable water bodies, especially if connected, can support abundant and diverse aquatic predators of NMGS (USFWS 2014). Lastly, environmental contamination affects chemical stress with a medium-magnitude effect, with medium understanding.

One habitat element, genetic diversity, has a bi-directional relationship with the critical biological activity or process of dispersal. This relationship is proposed to be a low-magnitude effect, with low understanding.

The CEM identified two controlling factors that have high-magnitude effects on other controlling factors: On-site water management directly affects fisheries introduction & fisheries management, while water storage-delivery system design & operation directly affects on-site water management. Each of these interactions are well understood, with medium understanding. The recreational activities controlling factor has a medium-magnitude effect on the fire management controlling factor, with medium understanding. Lastly, the fisheries introduction & fisheries management activities controlling factor has a medium-magnitude effect on recreational activities, with low understanding.

6-5 Northern Mexican Gartersnake (Thamnophis eques megalops) (NMGS) Basic Conceptual Ecological Model for the Lower Colorado River

The CEM identified three controlling factors that have high-magnitude effects on habitat elements: Nuisance species introduction & management directly affects both competitors and predators, and on-site water management and water storage- delivery system design & operation both directly affect the hydrologic regime and surface water connectivity. The relationships between nuisance species introduction & management with competitors and with predators are very well understood. The other hydrologic relationships are of medium understanding. Non-native competitors and predators are particularly problematic for NMGS, and the availability of water is a critical determinant of hydrologic regime and connectivity. These relationships are very well understood.

Six controlling factors have medium-magnitude effects on various habitat elements. Fire management has a medium-magnitude effect on vegetation structure, with medium understanding. Fisheries introduction & fisheries management has a medium-magnitude effect on competitors and predators, with low understanding and a medium-magnitude effect on food availability, with medium understanding. Nuisance species introduction & management has medium-magnitude effects on food availability and infectious agents, both with medium understanding. Off-site land management & use has medium-magnitude effects on environmental contamination and substrate, both with medium understanding. On-site vegetation management has medium-magnitude effects on environmental contaminants, substrate, and vegetation structure, each with medium understanding. Lastly, wastewater & other contaminant inflows has medium-magnitude effects on environmental contaminants & infectious agents, both of low understanding.

Although the interaction is of low magnitude, one controlling factor has a direct effect on life stage outcomes: Fire management can directly affect neonate/juvenile survival. This relationship is of medium understanding.

6-6 Chapter 6 – Conceptual Ecological Model by Life Stage

Controlling Factor

Link#

Habitat Element

Link#

Critical Biological Activity or Process

Link#

Life-Stage Outcome

Link Magnitude (line thickness) High – thick line

Medium – medium line

Low – thin line Unknown – very thin line

Link Understanding (line color)

High – black line Medium – blue line Low – red line

Link Predictability (link label color)

High – black text Medium – blue text Low – red text Unknown – grey text

Figure 3.—CEM master diagram for NMGS life stage 1 – Neonates/juveniles life stage controlling factors, habitat elements, critical biological activities and processes, and life-stage outcome

6-7 Chapter 6 – Conceptual Ecological Model by Life Stage

NMGS LIFE STAGE 2—ADULTS

As described in chapter 2, this life stage begins when the young garter snakes reach the size at which they become sexually mature. This typically occurs at 2 years of age for males and 3 years of age for female NMGS and lasts through the remainder of the snake’s life. NMGS adults may live up to 10 years in the wild and may grow throughout their remaining lives (USFWS 2014). This life stage has two life-stage outcomes (see figure 1): adult survival and adult fertility. Figure 4 (at the end of this section) presents the complete CEM for this life stage, showing all controlling factors, habitat elements, critical biological activities and processes, life-stage outcomes, and their linkages.

The CEM identifies five critical biological activities or processes affecting adult survival outcomes for this life stage as shown on figure 4. However, the CEM identifies no critical biological activities or processes that have high-magnitude effects; rather, two—predation and foraging—do have medium-magnitude direct effects.

Predation is the most commonly proposed cause of mortality among NMGS in both life stages, thought to contribute significantly (along with habitat loss and competition) to NMGS extirpation across large areas of its historic range. As a result, link understanding is rated as high.

The three other critical biological activities or processes proposed to affect adult survival are brumation, chemical stress, and disease. The CEM proposes that disease and chemical stress have an effect of unknown magnitude, and brumation has a low-magnitude effect on survival. The effect of brumation is of medium understanding, while chemical stress and disease are of low understanding due to a lack of studies of these possible effects in NMGS or any closely related species (see figure 4).

The CEM proposes that several of the critical biological activities and processes affect each other, possibly compounding their effects on survival. Specifically, disease and ecdysis are proposed to affect each other, and predation and foraging affect each other. The bi-directional effect of disease and ecdysis on each other is proposed to be unknown, with low understanding for the relationship in either direction. The predation and foraging interactions are proposed to be of low magnitude, with medium understanding for the foraging/predation interaction. In addition, competition has a medium-magnitude effect on foraging, with medium understanding; basking has a low-magnitude effect on predation with medium understanding; and edysis has an unknown-magnitude effect on chemical stress, with low understanding.

6-9 Northern Mexican Gartersnake (Thamnophis eques megalops) (NMGS) Basic Conceptual Ecological Model for the Lower Colorado River

The CEM does not identify any habitat elements with high-magnitude effects on any critical biological activity or process. There are two habitat elements with medium-magnitude direct effects on critical activities or processes: competitors effect on competition and food availability effects on foraging. The effect of competitors on competition and of food availability on foraging are both very well understood in a general sense, although more specific research on interactions at LCR sites would be beneficial.

Two other habitat elements have medium-magnitude direct effects on predation: hydrologic regime and predators. The effect of hydrologic regime and predators on predation are of high understanding. The presence of suitable water bodies, especially if connected, can support abundant and diverse aquatic predators of NMGS (USFWS 2014). Lastly, environmental contamination affects chemical stress with a medium-magnitude effect, with medium understanding.

The CEM identified two controlling factors that have high-magnitude effects on other controlling factors: On-site water management directly affects fisheries introduction & fisheries management, while water storage-delivery system design & operation directly affects on-site water management. Each of these interactions are well understood, with medium understanding. The recreational activities controlling factor has a medium-magnitude effect on the fire management controlling factor, with medium understanding. Lastly, the fisheries introduction & fisheries management activities controlling factor has a medium-magnitude effect on recreational activities, with low understanding.

One habitat element, genetic diversity, has a bi-directional relationship with the critical biological activity or process of dispersal. This relationship is proposed to have a low-magnitude effect, with low understanding.

The CEM identified three controlling factors that have high-magnitude effects on habitat elements: Nuisance species introduction & management directly affects both competitors and predators, and on-site water management and water storage- delivery system design & operation both directly affect the hydrologic regime and surface water connectivity. The relationships between nuisance species introduction & management with competitors and with predators are very well understood. The other hydrologic relationships are of medium understanding. Non-native competitors and predators are particularly problematic for NMGS, and the availability of water is a critical determinant of hydrologic regime and connectivity.

Six controlling factors have medium-magnitude effects on various habitat elements. Fire management affects vegetation structure with a medium- magnitude effect, with medium understanding. Fisheries introduction & fisheries management has a medium-magnitude effect on competitors and predators, with low understanding, and a medium-magnitude effect on food availability, with

6-10 Chapter 6 – Conceptual Ecological Model by Life Stage

medium understanding. Nuisance species introduction & management has medium-magnitude effects on food availability and infectious agents, both of medium understanding. Off-site land management & use has medium-magnitude effects on environmental contamination and substrate, both with medium understanding. On-site vegetation management has medium-magnitude effects on environmental contaminants, substrate, and vegetation structure, each with medium understanding. Lastly, wastewater & other contaminant inflows has medium-magnitude effects on environmental contaminants & infectious agents, both with low understanding.

Although the interaction is of low magnitude, one controlling factor has a direct effect on life stage outcomes: Fire management can directly affect adult survival. This relationship is of medium understanding.

6-11 Chapter 6 – Conceptual Ecological Model by Life Stage

Controlling Factor

Link#

Habitat Element

Link#

Critical Biological Activity or Process

Link#

Life-Stage Outcome

Link Magnitude (line thickness) High – thick line

Medium – medium line

Low – thin line Unknown – very thin line

Link Understanding (line color)

High – black line Medium – blue line Low – red line

Link Predictability (link label color)

High – black text Medium – blue text Low – red text Unknown – grey text

.—FigureCEM 4 master diagram for NMGS life stage 2 – adult life stage controlling factors, habitat elements, critical biological activities and processes, and life-stage outcomes.

6-13

Chapter 7 – Causal Relationships Across Life Stages

This chapter examines the information assembled for the CEM across both life stages to assess the following:

• Which critical biological activities and processes most strongly affect the life-stage outcomes across all life stages?

• Which critical biological activities and processes strongly affect other critical biological activities and processes across all life stages?

• Which habitat elements, through their abundance, distribution, and/or quality, most strongly affect the most influential biological activities or processes across all life stages?

• Which habitat elements, through their abundance, distribution, and/or quality, most strongly affect the abundance, distribution, and/or quality of other habitat elements across all life stages?

• Which controlling factors most strongly affect the most influential habitat elements across all life stages?

• Which of the most influential causal relationships appear to be the least understood in ways that could affect their management?

EFFECTS OF CRITICAL BIOLOGICAL ACTIVITIES AND PROCESSES ON LIFE-STAGE OUTCOMES

Table 5 shows which critical biological activities and processes directly affect each life-stage outcome, the estimated magnitude of each effect, and the estimated level of understanding of the effect. Five critical biological activities or processes—brumation, chemical stress, disease, foraging, and predation— directly affect at least one life-stage outcome (survival) in both life stages, and one critical biological activity or process—basking—only affects one life stage, adults (adult fertility).

7-1 Northern Mexican Gartersnake (Thamnophis eques megalops) (NMGS) Basic Conceptual Ecological Model for the Lower Colorado River

Table 5.—Direct effects of critical biological activities and processes on NMGS life- stage outcomes (The letter in a cell indicates the proposed link magnitude: H = high; M = medium; L = low; U = unknown. The letter color indicates the level of understanding of the effect: High = X; Medium = X; Low = X.)

Life-stage outcome 

 Critical biological activity or process Neonate/juvenile survival Adult fertility Adult survival

Basking M Brumation L L Chemical stress U U U

Competition Disease U U U

Dispersal Ecdysis Foraging M L M Predation M L M

Table 5 indicates that no critical biological activities or processes are proposed to have high-magnitude effects on any life-stage outcome in either life stage. Table 5 also indicates that only three critical biological activities or processes are proposed to have medium-magnitude effects on any life-stage outcome in either life stage. Basking is proposed to have a medium potential to affect adult fertility, with medium understanding; foraging and predation are proposed to have a medium potential to affect adult and neonate/juvenile survival, with high understanding of each. All other links shown in table 5 are proposed to have either low- or unknown-magnitudes effects, most of which are of low understanding.

Looked at another way, table 5 indicates that the CEM identifies foraging and predation as the most important critical biological activities or processes shaping snake survival, and basking as the most important activity affecting adult fertility.

7-2 Chapter 7 – Causal Relationships Across Life Stages

EFFECTS OF CRITICAL BIOLOGICAL ACTIVITIES AND PROCESSES ON EACH OTHER

Table 6 shows which critical biological activities and processes directly affect other critical biological activities and processes, thereby influencing life-stage outcomes indirectly across the two NMGS life stages, the estimated magnitude of these effects, and the estimated level of understanding of the effects. Five critical biological activities or processes directly affect at least one other critical biological activity or process in both life stages.

Table 6.—Direct effects of critical biological activities and processes on each other across the two NMGS life stages (The two characters in each cell indicate the ratings for the neonates/juveniles and adults life stages, from top to bottom, respectively. The letter in a cell indicates the proposed link magnitude: H = high; M = medium; L = low; and U = unknown. The letter color indicates the level of understanding of the effect: High = X; Medium = X; Low = X.) Affected critical biological activity or process 

 Causal critical biological activity or process Basking Brumation Chemical stress Competition Disease Dispersal Ecdysis Foraging Predation

L Basking L

Brumation Chemical stress M Competition M U Disease U Dispersal U U Ecdysis U U L Foraging L L Predation L

7-3 Northern Mexican Gartersnake (Thamnophis eques megalops) (NMGS) Basic Conceptual Ecological Model for the Lower Colorado River

The CEM proposes that only one critical biological activity or process has a medium-magnitude effect—competition’s effect on foraging, which is of medium understanding. The remainder of the effects of critical biological activities and processes on each other have low or unknown magnitudes. Three of the relationships are of proposed to be of low understanding, while the effects of basking on predation, foraging on predation, and predation on foraging are considered to be moderately well understood (of medium understanding). In particular:

• Disease and ecdysis have a bi-directional relationship, proposed to be of unknown magnitude, with low understanding. Disease may affect the shedding process, while frequent, incomplete shedding may lead to increased susceptibility to infection.

• Foraging and predation have a bi-directional relationship, proposed to be of low magnitude, with medium understanding. Increased foraging activity may increase the potential for predation; conversely, the threat of predation may reduce foraging activity.

EFFECTS OF HABITAT ELEMENTS ON CRITICAL BIOLOGICAL ACTIVITIES AND PROCESSES

Table 3 (see chapter 4) identifies which habitat elements affect which critical biological activities and processes across both NMGS life stages. These relationships are in fact largely identical across both life stages in the CEM. Table 7 shows which habitat elements directly affect which critical biological activities and processes as in table 3 and also indicates the proposed magnitude and level of understanding of these effects between the two NMGS life stages.

Roughly two-thirds of the entries in table 7 indicate effects of habitat elements on critical biological activities and processes, with low or unknown magnitude. Conversely, while table 7 indicates that there are no proposed high-magnitude effects of habitat elements on critical biological activities and processes, there are several with medium magnitude as follows. Specifically, the CEM proposes that:

• The habitat elements of anthropogenic activities and temperature affect three critical biological activities and processes each, although all the relationships are proposed to be of low magnitude.

• The habitat element of competitors affects the rate of competition experienced by NMGS in both life stages, with medium magnitude. The relationship is proposed to have a high level of understanding.

7-4 Chapter 7 – Causal Relationships Across Life Stages

• The presence of environmental contaminants affects the potential for chemical stress experienced by NMGS in both life stages, with medium magnitude. The relationship is proposed to have a medium level of understanding.

• Food availability directly affects foraging activity with a proposed medium-magnitude effect, with high understanding.

• The hydrologic regime and predators both directly affect predation with proposed medium-level effects, with high understanding.

Table 7.—NMGS habitat elements and the critical biological activities and processes they are proposed to directly affect across the two NMGS life stages (The letter in a cell indicates the proposed link magnitude: H = high; M = medium; L = low; and U = unknown. The letter color indicates the level of understanding of the effect: High = X; Medium = X; Low = X.) 

Affected critical biological activity or process

 Causal habitat elements Basking Brumation Chemical stress Competition Disease Dispersal Ecdysis Foraging Predation L L L Anthropogenic disturbance L L L M Competitors M M Environmental contaminants M L M Food availability L M U L Genetic diversity U L M Hydrologic regime M U Infectious agents U M Predators M L L L Substrate L L L L Surface water connectivity L L L L Temperature L L L L Turbidity L L L L Vegetation structure L L L

7-5 Northern Mexican Gartersnake (Thamnophis eques megalops) (NMGS) Basic Conceptual Ecological Model for the Lower Colorado River

EFFECTS OF HABITAT ELEMENTS ON EACH OTHER

Table 8 shows which habitat elements directly affect which other habitat elements, as in table 4; it also indicates the proposed magnitude and level of understanding of these effects. These relationships are identical across both NMGS life stages.

Table 8.—Direct effects of habitat elements on each other across the two NMGS life stages (The letter in a cell indicates the proposed link magnitude: H = high; M = medium; L = low; and U = unknown. The letter color indicates the level of understanding of the effect: High = X; Medium = X; Low = X.)

Affected habitat element 

 Causal habitat elements Anthropogenic disturbance Competitors Environmental contaminants Food availability Genetic diversity regime Hydrologic Infectious Agents Predators Substrate Surface water connectivity Temperature Turbidity Vegetation structure

Anthropogenic disturbance U L L

Competitors M Environmental contaminants M

Food availability

Genetic diversity

Hydrologic regime M M M Infectious agents

Predators Substrate L L Surface water connectivity M M

Temperature Turbidity

Vegetation structure L M

Table 8 shows that six habitat elements—food availability, genetic diversity, infectious agents, predators, temperature, and turbidity—have no direct effects on any others. Further, five of the remaining seven habitat elements have an effect on at least more than one habitat element. Food availability and predators are

7-6 Chapter 7 – Causal Relationships Across Life Stages

directly affected by four other habitat elements. There are no relationships of high magnitude; however, there are several medium-magnitude effects of habitat elements on other habitat elements as follows. Specifically, the CEM proposes that:

• The presence of competitors directly affects food availability for NMGS with a proposed medium-magnitude effect, with medium understanding.

• The presence of environmental contaminants directly affects food availability with a proposed medium-magnitude effect, with medium understanding.

• The hydrologic regime affects three other habitat elements—competitors, predators, and vegetation structure—with proposed medium-magnitude effects. The relationships between competitors and vegetation structure with hydrologic regime are proposed to be of high understanding, while the relationship of local hydrology with predators is proposed to be of medium understanding.

• Surface water connectivity affects competitors with a proposed medium- magnitude effect, and along with vegetation structure, they both affect predators with proposed medium-magnitude effects, with medium understanding.

EFFECTS OF CONTROLLING FACTORS ON HABITAT ELEMENTS

Table 4 (see chapter 5) identifies which controlling factors affect which habitat elements in the NMGS conceptual ecological model. These relationships are identical across all life stages in the CEM. Table 9 also presents this information but adds information on the magnitude and level of understanding of these effects.

7-7 Northern Mexican Gartersnake (Thamnophis eques megalops) (NMGS) Basic Conceptual Ecological Model for the Lower Colorado River

Table 9.—Direct effects of controlling factors on habitat elements across the two NMGS life stages (The letter in a cell indicates the proposed link magnitude: H = high; M = medium; L = low; and U = unknown. The letter color indicates the level of understanding of the effect: High = X; Medium = X; Low = X.)

Habitat elements 

 Controlling factors Anthropogenic disturbance Competitors Environmental contaminants Food availability Genetic diversity regime Hydrologic Infectious agents Predators Substrate Surface water connectivity Temperature Turbidity Vegetation structure Fire management L L M

Fisheries introduction & fisheries management L M M M

Grazing L

Nuisance species introduction & management H M M H L L

Off-site land management & use M M

On-site vegetation management L M M

On-site water management H H Recreational activities L Site maintenance L L

Wastewater & other contaminant inflows M M

Water storage-delivery system design & operation H H

Table 9 shows that 3 of the 11 controlling factors—nuisance species introduction & management, on-site water management, and water storage-delivery system design & operation—each has a high-magnitude effect on at least one habitat element. These high-magnitude effects of habitat elements on other habitat elements are as follows. Specifically, the CEM proposes that:

• Nuisance species introduction & management has high-magnitude effects on both competitors and predators. These relationships are very well understood.

• On-site water management directly affects both the hydrologic regime and surface water connectivity. These relationships are proposed to have medium understanding in the greater LCR ecosystem.

7-8 Chapter 7 – Causal Relationships Across Life Stages

• Water storage-delivery system design & operation has high-magnitude effects on the hydrologic regime and surface water connectivity. These relationships are well understood.

Table 9 shows that 6 out of 11 controlling factors have medium-level effects on habitat elements. These include:

• Fire management has medium-magnitude, moderately well understood effects on vegetation structure.

• Fisheries introduction & fisheries management is proposed to have medium-magnitude effects on competitors, predators, and food availability. The relationship between fisheries activities and food availability is moderately well understood, while the interactions between fisheries activities and competitors and with predators are of low understanding.

• Nuisance species introduction & management has medium-magnitude effects on food availability and infectious agents. Both relationships are moderately well understood.

• Off-site land management & use has direct, medium-magnitude effects on environmental contaminants and substrate, both with proposed medium understanding.

• On-site vegetation management has medium-magnitude, moderately well understood effects on substrate and vegetation structure.

• Finally, waste water contaminants & other inflows have medium- magnitude effects on environmental contaminants & infectious agents in the LCR ecosystem, but both are rated as having low understanding.

The CEM also recognizes that several controlling factors directly affect other controlling factors. Table 10 shows which controlling factors affect which others, with what proposed magnitude, and with what proposed level of understanding. Table 10 shows that two controlling factors have high-magnitude effects on other controlling factors as follows. Specifically, the CEM proposes that:

• On-site water management has high-magnitude, moderately well understood effects on fisheries introduction & fisheries management.

• Water storage-delivery system design & operations has high-magnitude and moderately well-understood effects on on-site water management.

7-9 Northern Mexican Gartersnake (Thamnophis eques megalops) (NMGS) Basic Conceptual Ecological Model for the Lower Colorado River

Table 10.—Direct effects of controlling factors on each other across the two NMGS life stages (The letter in a cell indicates the proposed link magnitude: H = high; M = medium; L = low; and U = unknown. The letter color indicates the level of understanding of the effect: High = X; Medium = X; Low = X.)

Affected controlling factor 

management

delivery system design & operations -

site land management & use site vegetation site water management - - -  Causal controlling factor Fire management Fisheries introduction fisheries& management Grazing Nuisance species introduction management & Off On On Recreational activities maintenanceSite Wastewater other & contaminant inflows Water storage

Fire management L

Fisheries introduction & fisheries management M

Grazing

Nuisance species introduction & management Off-site land management & use

On-site vegetation management On-site water management H Recreational activities M Site maintenance Wastewater & other contaminant inflows

Water storage-delivery system design & operations H

CAUSAL RELATIONSHIPS WITH HIGH UNDERSTANDING

Many causal relationships proposed in the CEM (see chapter 6 and above, this chapter) are rated as having high understanding. The CEM proposes these relationships based on established ecological principles, knowledge of snake biology and ecology in general, published information on the regulated LCR

7-10 Chapter 7 – Causal Relationships Across Life Stages

ecosystem overall, and detailed studies of NMGS. The latter detailed studies include investigations by Rosen and Swalbe 1988, ongoing research conducted by LCR and AZGFD biologists and the gartersnake research project at NAU, and the USFWS listing and critical habitat designation reports (2014, 2020 respectively).

The CEM identifies five causal relationships between individual controlling factors. These relationships apply to both NMGS life stages. The CEM rates none of these relationships as having high understanding.

The CEM identifies 28 causal relationships between controlling factors and habitat elements, which also apply to both NMGS life stages. The CEM rates 2 of these 28 causal relationships as having high understanding, based on general ecological principles and publications concerning resource conditions and management along the LCR valley, as discussed also in chapters 4 and 5. These relationships follow:

• Nuisance species introduction & management and (a) competitors and (b) predators.

The CEM identifies much less certainty in knowledge concerning possible causal interactions among habitat elements, between habitat elements and critical biological activities and processes, among critical biological activities and processes, between critical biological activities and processes and life-stage outcomes, or among life-stage outcomes. Chapters 3–6 provide detailed explanations. Specifically:

• The CEM identifies 14 causal relationships among habitat elements, all of which apply across both NMGS life stages, but it identifies only 2 of these as having high understanding. These two are the effects of the hydrologic regime on competitors and vegetation structure.

• The CEM identifies 23 causal relationships between habitat elements and critical biological activities and processes for both life stages. Four of these are of high understanding—competitors to competition, food availability to foraging, hydrologic regime to predation, and predators to predation.

• The CEM for both life stages identifies 7 causal relationships among critical biological activities and processes and 15 causal relationships between critical biological activities and processes and life-stage outcomes. There are no causal relationships of high understanding among the critical biological activities and processes. There are two relationships of high understanding between the critical biological activities and processes: foraging and predation to survival for both life stages.

7-11 Northern Mexican Gartersnake (Thamnophis eques megalops) (NMGS) Basic Conceptual Ecological Model for the Lower Colorado River

POTENTIALLY INFLUENTIAL CAUSAL RELATIONSHIPS WITH LOW UNDERSTANDING AND/OR UNKNOWN MAGNITUDE

Many causal relationships proposed in the CEM (see chapter 6 and above, this chapter) are rated as having low understanding. The CEM proposes these relationships based on established ecological principles, knowledge of snake biology and ecology in general, and suggestions in the literature on NMGS or on closely related species. However, few or no studies provide actual evidence concerning these relationships specifically for NMGS or any closely related species or concerning relevant habitat element conditions and dynamics in the greater LCR ecosystem. In some instances, the gaps in knowledge are so large that the CEM assigns a rating of unknown for link magnitude as well. The CEM includes links with unknown magnitude based on established ecological principles and knowledge of snake biology and ecology in general for which there is no documentation specifically for NMGS or any closely related species (e.g., for any other gartersnake). For example, the presence of infectious agents that affect snakes or their prey would increase the possibility of disease transmission, with impacts to NMGS. However, the literature does not contain sufficient information about specific infectious agents affecting NMGS at this time. Links rated as having low understanding appear in the NMGS life-stage diagrams in chapter 6 (figures 3–4) as red arrows; links with unknown magnitude appear in these diagrams as extremely thin red arrows.

7-12

Chapter 8 – Discussion and Conclusions

The NMGS conceptual ecological model has several notable features. First, the assessment of the causal relationships among controlling factors, habitat elements, critical biological activities and processes, and life-stage outcomes indicates the following strong (high-magnitude) causal relationships:

• Three controlling factors—nuisance species introduction & management, on-site water management, and water storage-delivery system design & operation—have direct, high-magnitude effects on one or more habitat elements relevant to one or both NMGS life stages. These relationships are of high or medium understanding.

The assessment of causal relationships among controlling factors, habitat elements, critical biological activities and processes, and life-stage outcomes also identifies numerous relationships with proposed intermediate (medium) and low magnitude. As knowledge about the species expands, the ratings of link magnitude for these proposed relationships, as well as for those currently assigned a high-magnitude rating, may change.

Some components of the model affect a high number of other components. The CEM identifies the following controlling factors, habitat elements, and critical biological activities and processes that affect at least five other components:

• One habitat element, hydrologic regime, has direct, medium-magnitude effects on at least five other habitat elements (4) and/or critical biological activities or processes (1). Four out of five of the effects of hydrologic regime on other model components are very well understood, and the effect of hydrologic regime on predators is moderately well understood.

• One habitat element—anthropogenic disturbance—has direct effects on six other components (three habitat elements and three critical biological activities or processes). While considered of low magnitude, most of these relationships are of low understanding, and further study may modify the magnitude of some of these relationships.

• Two habitat elements—substate and vegetation structure—each have direct effects on five other components (critical biological activities or processes and/or habitat elements). While most relationships are considered of low magnitude (except vegetation structure’s effects on predation, which is of medium magnitude), they are all moderately well understood. In particular, the fact that these habitat elements affect so many critical biological activities or processes (three each) highlights the importance of substrate and vegetation structure to NMGS.

8-1 Northern Mexican Gartersnake (Thamnophis eques megalops) (NMGS) Basic Conceptual Ecological Model for the Lower Colorado River

• In addition to the high-magnitude effects that nuisance species introduction & management has on two habitat elements (predators and competitors) mentioned above, this controlling factors also has direct effects on four additional habitat elements (total of six). These include food availability, infectious agents, substrate, and vegetation structure. While these relationships are of low or medium magnitude, all of them are of medium or high understanding.

The CEM includes links with unknown magnitude based on established ecological principles and knowledge of particular features of snake biology and ecology in general for which there currently is no documentation for NMGS or any closely related species in particular. Specifically:

• The CEM identifies effects of disease on adult fertility, neonate/juvenile survival, and adult survival to be of unknown magnitude.

• The CEM identifies the effect of infectious agents on disease to be of unknown magnitude.

• The CEM identifies the effect of genetic diversity on disease to be of unknown magnitude.

• The CEM identifies the effects of chemical stress on adult survival, neonate/juvenile survival, and adult fertility to be of unknown magnitude.

• The CEM identifies the effect of ecdysis on chemical stress to be of unknown magnitude.

• The CEM identifies the effect of disease and ecdysis on each other (bi-directional relationship) to be of unknown magnitude.

Finally, the CEM also identifies several potentially important causal relationships with high magnitude and high or medium understanding. These links represent the best-understood aspects of NMGS ecology. Their medium and high ratings for link understanding reflect cumulative knowledge from several detailed studies of NMGS and their habitat. These better understood, high-magnitude relationships include:

• The CEM identifies the effects of nuisance species introduction & management on predators and competitors to both be of high magnitude and high understanding.

8-2 Chapter 8 – Discussion and Conclusions

• The CEM identifies the effects of water storage-delivery system design & operation on one controlling factor—on-site water management—to be of high magnitude and medium understanding.

• The CEM identifies the effects of water storage-delivery system design & operation on two habitat elements—hydrologic regime and surface water connectivity—to be of high magnitude and medium understanding.

8-3

LITERATURE CITED

Ajami, H., D.P. Guertin, L.R. Levick, and K. Uhlman. 2005. NEMO Watershed Based Plan Bill Williams Watershed. University of Arizona, Tucson, Arizona.

Allendorf, F.W. and R.F. Leary. 1986. Heterozygosity and fitness in animals. Pages 57–76 in M.E. Soule (editor). Conservation Biology. Sinauer, Sunderland, Massachusetts.

Arizona Department of Environmental Quality. 2016a. 2016 Water Quality in Arizona 305(b) Assessment Report. Arizona Department of Environmental Quality, Phoenix, Arizona.

_____. 2016b. Bill Williams Watershed Water Quality Assessment, Page, Arizona 2016 305(b) Assessment Report. Arizona Department of Environmental Quality, Phoenix, Arizona.

_____. 2018. Arizona’s 2018 303(d) List of Impaired Waters. Arizona Department of Environmental Quality, 2018 303(d) List of Impaired Waters, Phoenix, Arizona.

Asplund, T.R. 2000. The Effects of Motorized Watercraft on Aquatic Ecosystems. PUBL-SS-948-00. Wisconsin Department of Natural Resources/University of Wisconsin, Madison.

Bank, M.S., J. Crocker, B. Connery, and A. Amirbahman. 2007. Mercury bioaccumulation in green frog (Rana clamitans) and bullfrog (Rana catesbeiana) tadpoles from Acadia National Park, Maine, USA. Environmental Toxicology and Chemistry 26:118–125.

Barber, D.M., V.A. Poole, C.R. Sanchez, P. Roady, and M.C. Allender. 2016. Snake fungal infection associated with Fusarium found in Nerodia erythrogaster transversa (blotched water snake) in Texas, USA. Herpetological Review 47(1):39–42.

Benda, L., N.L. Poff, D. Miller, T. Dunne, G. Reeves, G. Pess, and M. Pollock. 2004. The network dynamics hypothesis: how channel networks structure riverine habitats. BioScience 54:413–427.

Bixby, R.J., S.D. Cooper, R.E. Gresswell, L.E. Brown, C.N. Dahm, and K.A. Dwire. 2015. Fire effects on aquatic ecosystems: an assessment of the current state of science. Freshwater Science 34(4):1340–1350.

L-1 Northern Mexican Gartersnake (Thamnophis eques megalops) (NMGS) Basic Conceptual Ecological Model for the Lower Colorado River

Blaustein, A.R., J.M. Romansic, J.M. Kiesecker, and A.C. Hatch. 2003. Ultraviolet radiation, toxic chemicals and amphibian population declines. Diversity and Distributions 9:123–140.

Bourne, J. 2020. Surveys for the northern Mexican gartersnake (Thamnophis eques megalops) at Havasu National Wildlife Refuge. Colorado River Terrestrial and Riparian meeting, McLaughlin, Nevada.

Boyle, S.A. and F.B. Samson. 1985. Effects of nonconsumptive recreation on wildlife: a review. Wildlife Society Bulletin 13(2):110–116.

Braun, D.P, E.A. Johnson, and R. Unnasch. 2019. Lowland Leopard Frog (Rana = Lithobates yavapaiensis) (LLFR) Basic Conceptual Ecological Model for the Lower Colorado River. Submitted to the Lower Colorado River Multi-Species Conservation Program, Boulder City, Nevada, by Sound Science, LLC, Boise, Idaho, under contract No. R16PC00028.

Brennan, T.C. 2008. Online Field Guide to the Reptiles and Amphibians of Arizona. http://www.reptilesofaz.org

Brennan, T.C. and A.T. Holycross. 2006. A Field Guide to Amphibians and Reptiles in Arizona. Arizona Game and Fish Department, Phoenix, Arizona.

Brooks, B.W., J.P. Grover, and D.L. Roelke. 2011. Prymnesium parvum: an emerging threat to inland waters. Environmental Toxicology and Chemistry 30:1955–1964.

Brühl, C.A., T. Schmidt, S. Pieper, and A. Alscher. 2013. Terrestrial pesticide exposure of amphibians: An underestimated cause of global decline? Nature: Scientific Reports 3:1–4.

Bureau of Land Management. 2020. Planet Ranch Nuisance Wild Burro Gather. https://www.blm.gov/programs/wild-horse-and-burro/herd- management/gathers-and-removals/arizona/2020-planet-ranch-nuisance- gather

Burger, J. 1992. Trace element levels in pine snake hatchlings: tissue and temporal differences. Archives of Environmental Contamination and Toxicology 22:209–213.

L-2 Literature Cited

Burke, M., K. Jorde, and J.M. Buffington. 2009. Application of a hierarchical framework for assessing environmental impacts of dam operation: changes in streamflow, bed mobility and recruitment of riparian trees in a western North American river. Journal of Environmental Management 90:S224– S236.

Californiaherps.com. 2021. Southern Watersnake (Nerodia fasciata). http://www.californiaherps.com/snakes/pages/n.fasciata.html

Carpenter, J. 2005. Competition for food between an introduced crayfish and two fishes endemic to the Colorado River Basin. Environmental Biology 72:335–342.

Center for Biological Diversity. 2003. Petition to List the Mexican Gartersnake, Thamnophis eques megalops, as an Endangered or Threatened Species Under the Endangered Species Act. Petition submitted to the U.S. Department of the Interior.

Conant, R. 2003. Observations on gartersnakes of the Thamnophis eques complex in the lakes of Mexico’s Transvolcanic Belt, with descriptions of new taxa. American Museum Novitates 3406:1–64.

Conway, C.J., C.P. Nadeau, and L. Piest. 2010. Fire helps restore natural disturbance regime to benefit rare and endangered marsh birds endemic to the Colorado River. Ecological Applications 20:2024–2035.

Costanzo, J.P. 1989a. A physiological basis for prolonged submergence in hibernating garter snakes Thamnophis sirtalis: evidence for an energy- sparing adaptation. Physiological Zoology 62(2):580–592.

_____. 1989b. Conspecific scent trailing by gartersnakes (Thamnophis sirtalis) during autumn: further evidence for use of pheromones and den location. Journal of Chemical Ecology 15(11):2531–2538.

Cotten, T. 2011. Lowland Leopard Frog (Rana yavapaiensis) and Colorado River Toad (Bufo alvarius) Distribution and Habitat Use in the Greater Lower Colorado River Ecosystem. Lower Colorado River Multi-Species Conservation Program, Boulder City, Nevada.

Cotten, T. and D. Grandmaison. 2013. Lowland Leopard Frog and Colorado River Toad Distribution and Habitat Use in the Greater Lower Colorado River Ecosystem, 2012 Annual Report. Submitted to the Lower Colorado River Multi-Species Conservation Program, Boulder City, Nevada.

L-3 Northern Mexican Gartersnake (Thamnophis eques megalops) (NMGS) Basic Conceptual Ecological Model for the Lower Colorado River

Cotten, T.B. and D.J. Leavitt. 2014. Lowland Leopard Frog and Colorado River Toad Distribution and Habitat Use in the Greater Lower Colorado River Ecosystem. Submitted to the Lower Colorado River Multi-Species Conservation Program, Boulder City, Nevada.

Cotten, T.B., J.D. Miller, and D.D. Grandmaison. 2013. Geographic distribution: Thamnophis eques (Mexican gartersnake). Herpetological Review 44(1):111.

Creed, R.P., Jr. 1994. Direct and indirect effects of crayfish grazing in a stream community. Ecology 75(7):2091–2103.

Cucherousset, J. and J.D. Olden. 2011. Ecological impacts of non-native freshwater fishes. Fisheries 36:215–230.

Davenport, J. 1992. Animal Life at Low Temperatures. Springer, Netherlands.

Davidson, C., H.B. Shaffer, and M.R. Jennings. 2002. Spatial tests of the pesticide drift, habitat destruction, UV-B, and climate-change hypotheses for California amphibian declines. Conservation Biology 16:1588–1601.

DiGennaro, B., D. Reed, C. Swanson, L. Hastings, Z. Hymanson, M. Healey, S. Siegel, S. Cantrell, and B. Herbold. 2012. Using conceptual models and decision-support tools to guide ecosystem restoration planning and adaptive management: an example from the Sacramento-San Joaquin Delta, California. San Francisco Estuary and Watershed Science 10:1–15.

d’Orgeix, C.A., T. Mathies, B.L. Ellison, K.L. Johnson, I.V. Monagan, and T.A. Young. 2013. Northern Mexican gartersnakes, Thamnophis eques megalops, feeding on Spea multiplicata in an ephemeral pond. Herpetological Review 44(2):213–215.

Emmons, I. 2017. Ecology of the federally threatened northern Mexican gartersnakes in north-central Arizona. M.S. thesis. Northern Arizona University, Flagstaff, Arizona.

Emmons, I. and E. Nowak. 2013. Northern Mexican Gartersnake Surveys, 2012 Interim Report. Northern Arizona University, Colorado Plateau Research Station, Flagstaff, Arizona. 56 p.

Emmons, I.D., E.M. Nowak, and K.K. Layuger. 2016a. Prey availability and foraging events of the northern Mexican gartersnake (Thamnophis eques megalops) in North-Central Arizona. Herpetological Review 47(4):555–561.

L-4 Literature Cited

Emmons, I.D., E.M. Nowak, and T.C. Theimer. 2016b. Thamnophis eques megalops (northern Mexican gartersnake). Predation. Herpetological Review 47:485–486.

Etzel, K.E., T.C. Theimer, M.J. Johnson, and J.A. Holmes. 2014. Variation in prey delivered to common black-hawk (Buteogallus anthracinus) nests in Arizona drainage basins. Journal of Raptor Research 48(1):54–60.

Fischenich, J.C. 2008. The Application of Conceptual Models to Ecosystem Restoration. Technical Note EMRRP-EBA-01. U.S. Army Engineer Research and Development Center, Vicksburg, Mississippi.

Fitch, H.S. 2003. Tail loss in gartersnakes. Herpetological Review 34(3):212– 213.

Ford, N.B. and R.A. Seigel. 1989a. Phenotypic plasticity in reproductive traits: evidence from a viviparous snake. Ecology 70:1768–1774.

———. 1989b. Relationships among body size, clutch size, and egg size in three species of oviparous snakes. Herpetologica 45:75–83.

_____. 1994. An experimental study of the trade-offs between age and size at maturity: effects of energy availability. Functional Ecology 8:91–96.

Fullerton, A.H., K.M. Burnett, E.A. Steel, R.L. Flitcroft, G.R. Pess, B.E. Feist, C.E. Torgersen, D.J. Miller, and B.L.L. Sanderson. 2010. Hydrological connectivity for riverine fish: measurement challenges and research opportunities. Freshwater Biology 55:2215–2237.

Gregory, P.T. 1977. Life-history parameters of the red-sided garter snake (Thamnophis sirtalis parietalis) in an extreme environment, the Interlake region of Manitoba. National Museum of Canada Zoology publication 13:1– 44.

Gregory, P.T. and K.M. Skebo. 1998. Trade-offs between reproductive traits and the influence of food intake during pregnancy in the garter snake, Thamnophis elegans. The American Naturalist 151(5):477–486.

Gregory, P.T. and K.W. Stewart. 1975. Long-distance dispersal and feeding strategy of the red-sided garter snake (Thamnophis sirtalis parietalis) in the Interlake of Manitoba. Canadian Journal of Zoology 53:238–245.

Halpert, A.P., W.R. Garstka, and D. Crews. 1982. Sperm transport and storage and its relation to the annual cycle of the female red-sided garter snake, Thamnophis sirtalis parietalis. Journal of Morphology 174(2):149–159.

L-5 Northern Mexican Gartersnake (Thamnophis eques megalops) (NMGS) Basic Conceptual Ecological Model for the Lower Colorado River

Heller, S. and M. Halpern. 1981. Laboratory observations on the conspecific and congeneric scent trailing in garter snakes (Thamnophis). Behavioral and Neural Biology 33(3):372–377.

Holycross A.T., W. Burger, E. Nigro, and T. Brennan. 2006. Surveys for Thamnophis eques and Thamnophis rufipunctatus in the Gila River Watershed of Arizona and New Mexico, submitted to the Arizona Game and Fish Department. Arizona State University, Tempe, Arizona.

Huey, R.B., C.R. Peterson, S.J. Arnold, and W.P. Porter. 1989. Hot rocks and not-so-hot rocks: retreat-site selection by garter snakes and its thermal consequences. Ecology 70(4):931944.

Jacobson, R.B., M.L. Annis, M.E. Colvin, D.A. James, T.L. Welker, and M.J. Parsley. 2016. Missouri River Scaphirhynchus albus (Pallid Sturgeon) Effects Analysis—Integrative Report 2016. Scientific Investigations Report 2016–5064. U.S. Geological Survey, Reston, Virginia.

King, R.B. and J.R. Turmo. 1997. The effects of ecdysis on feeding frequency and behavior of the common garter snake (Thamnophis sirtalis). Journal of Herpetology 31(2):310–312.

Kondolf, G.M., J.G. Williams, T.C. Horner, and D. Milan. 2008. Assessing physical quality of spawning habitat. Pages 249–274 in D.A Sear and P. DeVries (editors). Salmonid Spawning Habitat in Rivers: Physical Controls, Biological Responses, and Approaches. American Fisheries Society Symposium 65. American Fisheries Society, Bethesda, Maryland.

Krueper, D.J. 1993. Effects of land use practices on western riparian ecosystems. Pages 321–330 in D.M. Finch and P.W. Stangel (editors). Status and Management of Neotropical Migratory Birds. U.S. Department of Agriculture, U.S. Forest Service, Rocky Mountain Research Station GTR- RM-229.

Krueper, D.J., J. Bart, and T.D. Rich. 2003. Response of vegetation and breeding birds to the removal of cattle on the San Pedro River, Arizona. Conservation Biology 17:607–615.

Leavitt, D.J., K.O. Sullivan, J.D. Miller, and T.B. Cotten. 2017. Ecology and Habitat Requirements of Lowland Leopard Frogs and Colorado River Toads, 2015 Annual Report. Submitted to the Lower Colorado River Multi-Species Conservation Program, Boulder City, Nevada.

L-6 Literature Cited

Levick, L., J. Fonseca, D. Goodrich, M. Hernandez, D. Semmens, J. Stromberg, R. Leidy, M. Scianni, D. P. Guertin, M. Tluczek, and W. Kepner. 2008. The Ecological and Hydrological Significance of Ephemeral and Intermittent Streams in the Arid and Semi-arid American Southwest. Environmental Protection Agency, EPA/600/R-08/134, Washington, D.C.

Lind, A.J. 1990. Ontogenetic changes in the foraging behavior, habitat use and food habits of the western aquatic gartersnake, Thamnophis couchii, at Hudygurdy Creek, Del Norte County, California. M.S. thesis. Humboldt State University, Arcata, California.

Lougheed, V.L., B. Crosbie, and P. Chow-Fraser. 1998. Predictions on the effect of common carp (Cyprinus carpio) exclusion on water quality, zooplankton, and submergent macrophytes in a Great Lakes wetland. Canadian Journal of Fisheries and Aquatic Sciences 55:1189–1197.

Lower Colorado River Multi-Species Conservation Program (LCR MSCP). 2004. Lower Colorado River Multi-Species Conservation Program, Volume II: Habitat Conservation Plan. Final. December 17 (J&S 00450.00). Sacramento, California.

_____. 2014. Final Implementation Report, Fiscal Year 2015 Work Plan and Budget, Fiscal Year 2013 Accomplishment Report. Bureau of Reclamation, Boulder City, Nevada.

_____. 2017. Final Implementation Report, Fiscal Year 2018 Work Plan and Budget, Fiscal Year 2016 Accomplishment Report. Bureau of Reclamation, Boulder City, Nevada.

_____. 2018a. Final Implementation Report, Fiscal Year 2019 Work Plan and Budget, Fiscal Year 2017 Accomplishment Report. Bureau of Reclamation, Boulder City, Nevada.

_____. 2018b. Five-Year Monitoring and Research Priorities for the Lower Colorado Multi-Species Conservation Program, 2018–2022. Bureau of Reclamation, Boulder City, Nevada.

Macías-García, C.M. and H. Drummond. 1988. Seasonal and ontogenetic variation in the diet of the Mexican gartersnake Thamnophis eques in Lake Tecocomilco, Hidalgo. Journal of Herpetology 22(2):129–134.

Mac Nally, R., E. Fleishman, and D.D. Murphy. 2004. Influence of temporal scale of sampling on detection of relationships between invasive plants and the diversity patterns of plants and butterflies. Conservation Biology 18:1525–1532.

L-7 Northern Mexican Gartersnake (Thamnophis eques megalops) (NMGS) Basic Conceptual Ecological Model for the Lower Colorado River

Mann, R.M., R.V. Hyne, C.B. Choung, and S.P. Wilson. 2009. Amphibians and agricultural chemicals: review of the risks in a complex environment. Environmental Pollution 157:2903–2927.

McDonald, D. B. and H. Caswell. 1993. Matrix methods for avian demography. Pages 139–185 in D.M. Power (editor). Current Ornithology. Plenum Press, New York, New York.

Meixner, T. and P. Wohlgemuth. 2004. Wildfire impacts on water quality. Southwest Hydrology 3(5):22–25.

Meyer, R. 2005. Species: Atriplex lentiformis. Fire Effects Information System. U.S. Department of Agriculture, Forest Service, Rocky Mountain Research Station, Fire Sciences Laboratory, Fort Collins, Colorado. https://www.fs.fed.us/database/feis/plants/shrub/atrlen/all.html

Meyer, J.L., D.L. Strayer, J.B. Wallace, S.L. Eggert, G.S. Helfman, and N.E. Leonard. 2007. The contribution of headwater streams to biodiversity in river networks. Journal of the American Water Resources Association 43:86–103.

Morafka, D.J., E.K. Spangenberg, and V.A. Lance. 2000. Herpetological Monographs 14:353–370.

Munes, R., R.N. Reed, B. Falk, A.L. Hall, and A.T. Holycross. 2016. Thamnophis marcianus (checkered gartersnake). Herpetological Review 47(4):631–632.

Murphy, D.D. and P.S. Weiland. 2011. The route to best science in implementation of the Endangered Species Act’s consultation mandate: the benefits of structured effects analysis. Environmental Management 47:161– 72.

_____. 2014. Science and structured decision making: fulfilling the promise of adaptive management for imperiled species. Journal of Environmental Studies and Sciences 4:200–207.

Myatt, J. 2016. Battling an invasion of watersnakes. USFWS Pacific Southwest Region. Online newsroom. https://www.fws.gov/cno/newsroom/featured/2016/Invasive_Nerodia/

Myrand, J.M., T.A. Sprague, K. Anderson, and E.M. Nowak. 2021. Thamnophis equis megalops (Northern Mexican gartersnake). Reproduction. Herpetological Review 52(1):173–174.

L-8 Literature Cited

NatureServe. 2020. Mexican gartersnake. NatureServe Explorer: An online encyclopedia of Life. http://www.natureserve.org/explorer

NAU (see Northern Arizona University).

Nelson, K.J. and P.T. Gregory. 2000. Activity patterns of garter snakes, Thamnophis sirtalis, in relation to weather conditions at a fish hatchery on Vancouver Island, British Columbia. Journal of Herpetology 34(1):32–40.

Neskey, J. 2018. Blog: Feral swine eradication in Havasu National Wildlife Refuge: protecting endangered species from feral swine damage. https://www.usda.gov/media/blog/2018/04/17/feral-swine-eradication- havasu-national-wildlife-refuge-protecting-endangered

New Mexico Department of Game and Fish. 2013. Mexican gartersnake. Biota Information System of New Mexico, New Mexico Department of Game & Fish. http://www.bison-m.org/booklet.aspx?id=030265

New Mexico Environment Department. 2014. Wildfire Impacts on Surface Water Quality. https://www.env.nm.gov/surface-water-quality/wildfire-impacts-on-surface- water-quality/

Northern Arizona University (NAU). 2020. Northern Mexican Gartersnake Project. Flagstaff, Arizona. https://in.nau.edu/Gartersnake-Research-Project/Northern-Mexican- Gartersnake/

Nowak, E.M. and V.L. Boyarski. 2012. Thamnophis eques megalops (northern Mexican gartersnake). Reproduction: litter size. Herpetological Review 43(2): 351–352.

Nowak, E.M., M. Liszewski, and I. Emmons. 2011. Surveys for Northern Mexican Gartersnakes in Tavasci Marsh (Tuzigoot National Monument), unpublished final report. U.S. Geological Survey, Colorado Plateau Research Station, Flagstaff, Arizona.

Nowak, E.M, V. Boyarski, S. Nichols, and T. Greene. 2014. Thamnophis eques megalops (northern Mexican gartersnakes): maternal transmission of parasites. Herpetological Review 45(1):148–149.

L-9 Northern Mexican Gartersnake (Thamnophis eques megalops) (NMGS) Basic Conceptual Ecological Model for the Lower Colorado River

Nowak, E.M., I. Emmons, J. Myrand, C. Klovanish, and R. Bergamini. 2019. 2015–2017 Surveys and Telemetry of Northern Mexican Gartersnakes in Lower Tonto Creek. Unpublished final report to the Salt River Project. Colorado Plateau Research Station, Northern Arizona University, Flagstaff, Arizona.

O‘Donnell, R.P. and S.J. Arnold. 2005. Evidence for selection on thermoregulation: effects of temperature on embryo mortality in the garter snake Thamnophis elegans. Copeia 2005(4):929–933.

O’Donnell, R.P., R. Mixan, S.L. Arnett-Romero, and M.F. Ingraldi. 2019. Surveys for Threatened and Endangered Species and their Potential Breeding Habitat: Yellow-billed Cuckoo and Mexican Garter Snake Surveys on the Bill Williams River and Burro Creek, Arizona. Arizona Game and Fish Department, Wildlife Contracts Branch, Phoenix, Arizona.

Osgood, D.W. 1970. Thermoregulation in water snakes studied by telemetry. Copeia 1970(3):568–571.

Paroz, Y., J. Monzingo, and D. Propst. 2009. Inventory of the East, Middle, and West Forks of the Gila River. Submitted to the Bureau of Reclamation and U.S. Fish and Wildlife Service. New Mexico Department of Game and Fish and the Gila National Forest. 29 p.

Porter, K.R. 1972. Herpetology. W.B. Saunders Company, Philadelphia, Pennsylvania.

Pringle, C. 2003. What is hydrologic connectivity and why is it ecologically important? Hydrological Processes 17:2685–2689.

Roelke, D.L., J.P. Grover, B.W. Brooks, J. Glass, D. Buzan, G.M. Southard, L. Fries, G.M. Gable, L. Schwierzke-Wade, M. Byrd, and J. Nelson. 2011. A decade of fish-killing Prymnesium parvum blooms in Texas: roles of inflow and salinity. Journal of Plankton Research 33:243–253.

Ronning, C. 2020. Lower Colorado River Multispecies Conservation Program, Boulder City, Nevada, personal communication.

_____. 2021. Lower Colorado River Multispecies Conservation Program, Boulder City, Nevada, personal communication.

Rosen, P.C. 1991. Comparative field study of thermal preferenda in gartersnakes (Thamnophis). Journal of Herpetology 25(3):301–312.

L-10 Literature Cited

Rosen, P.C. and C. Schwalbe. 1988. Status of the Mexican and Narrow-headed Gartersnakes (Thamnophis eques megalops and Thamnophis rufipunctatus rufipunctatus) in Arizona. Unpublished report to the U.S. Fish and Wildlife Service. Arizona Game and Fish Department, Phoenix, Arizona.

Rosen, P.C., J. Wallace, and C. Schwalbe. 2001. Resurvey of the Mexican Gartersnake (Thamnophis eques) in Southeastern Arizona. Final report for the Arizona Game and Fish Department Heritage Program. University of Arizona School of Renewable Natural Resources, Tucson, Arizona.

Rosenfeld, J. 2003. Assessing the habitat requirements of stream fishes: an overview and evaluation of different approaches. Transactions of the American Fisheries Society 132:953–968.

Rosenfeld, J.S. and T. Hatfield. 2006. Information needs for assessing critical habitat of freshwater fish. Canadian Journal of Fisheries and Aquatic Sciences 63:683–698.

Sabin, L.B. 2018. Northern Mexican Gartersnake (Thamnophis eques megalops) Species Profile. Lower Colorado Region, Lower Colorado River Multi- Species Conservation Program, Boulder City, Nevada.

Sabin, B. 2020. Lower Colorado River Multispecies Conservation Program, Boulder City, Nevada, personal communication.

_____. 2021. Lower Colorado River Multispecies Conservation Program, Boulder City, Nevada, personal communication.

Sallenave, R. 2010. Toxic Golden Algae (Prymnesium parvum). Circular 647. Cooperative Extension Service, New Mexico State University, Las Cruces, New Mexico. http://aces.nmsu.edu/pubs/_circulars/cr-647/welcome.html

Shine, R., M.M. Olson, I.T. Moore, M.P. LeMaster, and R.T. Mason. 1999. Why do male snakes have longer tails than females? Proceedings of the Royal Society of London 266:2147–2151.

Smalling, K.L., R. Reeves, E. Muths, M. Vandever, W.A. Battaglin, M.L. Hladik, and C.L. Pierce. 2015. Pesticide concentrations in frog tissue and wetland habitats in a landscape dominated by agriculture. Science of the Total Environment 502:80–90.

L-11 Northern Mexican Gartersnake (Thamnophis eques megalops) (NMGS) Basic Conceptual Ecological Model for the Lower Colorado River

Sprague, T.A. and H.L. Bateman. 2018. Influence of seasonality and gestation on habitat selection by northern Mexican gartersnakes (Thamnophis eques megalops). PLOS One 13(1):e0191829. https://doi.org/10.1371/journal.pone.0191829

Sredl, M.J. 2018. Rana yavapaiensis (Platz and Frost, 1984) Lowland Leopard Frog. AmphibiaWeb. University of California, Berkeley, California. http://amphibiaweb.org

Sredl, M.J., E.P. Collins, and J.M. Howland. 1997. Mark-recapture of Arizona leopard frogs. Pages 1–20 in M.J. Sredl (editor). Ranid Frog Conservation and Management, Technical Report Number 121. Arizona Game and Fish Department, Nongame and Endangered Wildlife Program, Phoenix, Arizona.

Tattersall, G.J., W.K. Milsom, A.S. Abe, S.P. Brito, and D.V. Andrade. 2004. The thermogenesis of digestion in rattlesnakes. The Journal of Experimental Biology 207:579–585.

Tecle, A. and D. Neary. 2015. Water quality impacts of forest fires. Journal of Pollution Effects and Control 3:3. https://doi.org/10.4172/2375-4397.1000140

Todd, B.D. and X.A. Glaudas. 2013. Appendix I. Handling live amphibians and reptiles. Pages 214–219 in G.J. Graeter, K.A. Buhlmann, L.R. Wilkinson, and J.W. Gibbons (editors). Inventory and Monitoring: Recommended Techniques for Reptiles and Amphibians. Technical Publication IM-1. Partners in Amphibian and Reptile Conservation, Birmingham, Alabama.

Unrine, J. M., C. H. Jagoe, W. A. Hopkins, and H. A. Brant. 2004. Adverse effects of environmentally relevant dietary mercury exposure in larvae of the southern leopard frog (Rana sphenocephala) larvae. Environmental Toxicology and Chemistry 23:2964–2970.

U.S. Department of Agriculture. 2019. Wildlife Damage Management – Food habitats of feral hogs. Extension Service website. https://wildlife-damage-management.extension.org/food-habits-of-feral- hogs/

U.S. Environmental Protection Agency. 2016. Aquatic Life Ambient Water Quality Criterion for Selenium-Freshwater 2016. U.S. Environmental Protection Agency, Office of Water, EPA 822-R-16-006, Washington, D.C.

L-12 Literature Cited

U.S. Department of the Interior. 2016. Draft Havasu National Wildlife Refuge Feral Swine Eradication Plan. U.S. Department of the Interior, U.S. Fish and Wildlife Service, Havasu National Wildlife Refuge, Needles, California. https://www.fws.gov/uploadedFiles/Plan_FeralSwineEradicationDRAFT_9- 19-2016.pdf

U.S. Fish and Wildlife Service (USFWS). 2002. Southwestern Willow Flycatcher Recovery Plan, Appendix G: Management of Livestock Grazing in the Recovery of the Southwestern Willow Flycatcher. U.S. Fish and Wildlife Service, Albuquerque, New Mexico.

_____. 2013. Endangered and threatened wildlife and plants; threatened status for the northern Mexican gartersnake and narrow-headed gartersnake; proposed rules. 50 CFR 17. Federal Register 78(132):41499–41547.

_____. 2014. Endangered and threatened wildlife and plants; threatened status for the northern Mexican gartersnake and narrow-headed gartersnake; final rule. 50 CFR 17. Federal Register 79(130):38678–38746.

_____. 2020. Designation of critical habitat for the northern Mexican gartersnake and narrow headed gartersnake, proposed rules. Federal Register 85(82):23608.

USFWS (see U.S. Fish and Wildlife Service).

Welsh, H.H., Jr., C.A. Wheeler, and A.J. Lind. 2010. Spatial ecology of the Oregon gartersnake, Thamnophis atratus hydrophilus, in a free-flowing stream environment. Copeia 2010(1):75–85.

Westerling, A.L., H.G. Hidalgo, D.R. Cayan, and T.W. Swetnam. 2006. Warming and earlier spring increase western U.S. forest wildfire activity. Science 313:940–943.

Wildhaber, M.L., A.J. DeLonay, D.M. Papoulias, D.L. Galat, R.B. Jacobson, D.G. Simpkins, P.J. Baaten, C.E. Korschgen, and M.J. Mac. 2007. A Conceptual Life-History Model for Pallid and Shovelnose Sturgeon. Circular 1315. U.S. Geological Survey, Reston, Virginia.

_____. 2011. Identifying structural elements needed for development of a predictive life-history model for pallid and shovelnose sturgeons. Journal of Applied Ichthyology 27:462–469.

Willis, L., S.T. Threlkeld, and C.C. Carpenter. 1982. Tail loss patterns in Thamnophis (Reptilia: Colubridae) and the probable fate of injured animals. Copeia 1982(1):98–101.

L-13 Northern Mexican Gartersnake (Thamnophis eques megalops) (NMGS) Basic Conceptual Ecological Model for the Lower Colorado River

Wolff, B.A., B.M. Johnson, A.R. Breton, P.J. Martinez, and D.L. Winkelman. 2012. Origins of invasive piscivores determined from the strontium isotope ratio (87Sr/86Sr) of otoliths. Canadian Journal of Fisheries and Aquatic Sciences 69:724–739.

Wood, D.A., I.D. Emmons, E.M. Nowak, B.L. Christman, A.T. Holycross, R.D. Jennings, and A.G. Vandergast. 2018. Conservation Genomics of the Mogollon Narrow-headed Gartersnake (Thamnophis rufipunctatus) and Northern Mexican Gartersnake (Thamnophis eques megalops). Open- File Report 2018-1141. U.S. Geological Survey, Reston, Virginia. 47 p. https://doi.org/10.3133/ofr20181141

Woodin, W.H., III. 1950. Note on Arizona species of Thamnophis. Herpetologica 6(2):39–40.

Young, M.E. and V.L. Boyarski. 2012. Thamnophis eques megalops; diet and mortality. Herpetological Review 43(3):498.

_____. 2013. Thamnophis eques megalops; predation. Herpetological Review 44(1):158–159.

L-14

ACKNOWLEDGMENTS

The authors gratefully acknowledge the input and advice from Beth Sabin and Jenny Smith, biologists with the LCR MSCP; Carolyn Ronning, Wildlife Group Manager (LCR MSCP); Becky Blasius, Adaptive Management Specialist, (LCR MSCP); and Jimmy Knowles, Adaptive Management Group Manager (LCR MSCP).

We would also like to acknowledge John Swett, Program Manager (LCR MSCP), for his leadership and support of this modeling effort that will guide and inform the work of the program well into the future.

A-1

ATTACHMENT 1

Species Conceptual Ecological Model Methodology for the Lower Colorado River Multi-Species Conservation Program

OVERVIEW OF METHODOLOGY

The conceptual ecological models (CEMs) for species covered by the Lower Colorado River Multi-Species Conservation Program (LCR MSCP) Habitat Conservation Plan expand on a methodology developed by the Sacramento-San Joaquin Delta Ecosystem Restoration Program (ERP): https://www.dfg.ca.gov/ERP/conceptual_models.asp. The ERP is jointly implemented by the California Department of Fish and Wildlife, U.S. Fish and Wildlife Service, and National Marine Fisheries Service. The Bureau of Reclamation participates in this program.

The ERP methodology incorporates common best practices for constructing CEMs for individual species (DiGennaro et al. 2012; Fischenich 2008;Wildhaber et al. 2007, 2011). It has the following key features:

• It focuses on the major life stages or events through which each species passes and the output(s) of each life stage or event. Outputs typically consist of survivorship or the production of offspring.

• It identifies the major drivers that affect the likelihood (rate) of each output. Drivers are physical, chemical, or biological factors—both natural and anthropogenic—that affect output rates and, therefore, control the viability of the species in a given ecosystem.

• It characterizes these interrelationships using a “driver-linkage-outcomes” approach. Outcomes are the output rates. Linkages are cause-effect relationships between drivers and outcomes.

• It characterizes each causal linkage along four dimensions: (1) the character and direction of the effect, (2) the magnitude of the effect, (3) the predictability (consistency) of the effect, and (4) the certainty of present scientific understanding of the effect (DiGennaro et al. 2012).

The CEM methodology used for species covered by the LCR MSCP Habitat Conservation Plan species expands this ERP methodology. Specifically, the present methodology incorporates the recommendations and examples of Burke et al. (2009), Kondolf et al. (2008), and Wildhaber et al. (2007, 2011) for a more hierarchical approach and adds explicit demographic notation for the characterization of life-stage outcomes (McDonald and Caswell 1993). This expanded approach provides greater detail on causal linkages and outcomes. The expansion specifically calls for identifying four types of model components for each life stage, and the causal linkages among them as follows:

Attachment 1 – Page 1 Northern Mexican Gartersnake (Thamnophis eques megalops) (NMGS) Basic Conceptual Ecological Model for the Lower Colorado River

• Life-stage outcomes are outcomes of an individual life stage, including the recruitment of individuals to the next succeeding life stage (e.g., juvenile to adult). For some life stages, the outcomes, alternatively or additionally, may include the survival of individuals to an older age class within the same life stage or the production of offspring. The rates of life-stage outcomes depend on the rates of the critical biological activities and processes for that life stage.

• Critical biological activities and processes are activities in which a species engages and the biological processes that must take place during each life stage that significantly affect life-stage outcomes. They include activities and processes that may benefit or degrade life-stage outcomes. Examples of critical activities and processes include mating, foraging, avoiding predators, avoiding other specific hazards, gamete production, egg maturation, leaf production, and seed germination. Critical activities and processes are “rate” variables. Taken together, the rate (intensity) of these activities and processes determine the rates of different life-stage outcomes.

• Habitat elements are specific habitat conditions that significantly ensure, allow, or interfere with critical biological activities and processes. The full suite of natural habitat elements constitutes the natural habitat template for a given life stage. Human activities may introduce habitat elements not present in the natural habitat template. Defining a habitat element may involve estimating the specific ranges of quantifiable properties of that element whenever the state of knowledge supports such estimates. These properties concern the abundance, spatial and temporal distributions, and other qualities of the habitat element, that significantly affect the ways in which it ensures, allows, or interferes with critical biological activities and processes.

• Controlling factors are environmental conditions and dynamics—both natural and anthropogenic—that determine the quality, abundance, and spatial and temporal distributions of one or more habitat elements. In some instances, a controlling factor alternatively or additionally may directly affect a critical biological activity or process. Controlling factors are also called “drivers.” A hierarchy of controlling factors will exist, affecting the system at different temporal and spatial scales. Long-term dynamics of climate and geology define the domain of this hierarchy (Burke et al. 2009). For example, the availability of suitable nest sites for a riparian nesting bird may depend on factors such as canopy closure, community type, humidity, and intermediate structure which, in turn, may depend on factors such as water storage-delivery system design and operation (dam design, reservoir morphology, and dam operations) which, in turn, is shaped by watershed geology, vegetation, climate, land use, and

Attachment 1 – Page 2 Attachment 1 – Species Conceptual Ecological Model Methodology for the Lower Colorado River Multi-Species Conservation Program

water demand. The LCR MSCP conceptual ecological models focus on controlling factors that are within the scope of potential human manipulation, including management actions directed toward the species of interest.

The present CEM methodology also explicitly defines a “life stage” as a biologically distinct portion of the life cycle of a species. The individuals in each life stage undergo distinct developments in body form and function; engage in distinct types behaviors, including reproduction; use different sets of habitats or the same habitats in different ways; interact differently with their larger ecosystems; and/or experience different types and sources of stress. A single life stage may include multiple age classes. A CEM focused on life stages is not a demographic model per se (McDonald and Caswell 1993). Instead, it is a complementary model focused on the ecological factors (drivers) that shape population dynamics.

This expanded approach permits the consideration of six possible types of causal relationships, on which management actions may focus, for each life stage of a species:

(1) The effect of one controlling factor on another

(2) The effect of a controlling factor on the abundance, spatial and temporal distributions, and other qualities of a habitat element

(3) The effect of the abundance, spatial and temporal distributions, and other qualities of one habitat element on those of another

(4) The effect of the abundance, spatial and temporal distributions, and other qualities of a habitat element on a critical biological activity or process

(5) The effect of one critical biological activity or process on another

(6) The effect of a critical biological activity or process on a specific life- stage outcome

Each controlling factor may affect the abundance, spatial and temporal distributions, and other qualities of more than one habitat element, and several controlling factors may affect the abundance, spatial or temporal distributions, or other qualities of each habitat element. Similarly, the abundance, spatial and temporal distributions, and other qualities of each habitat element may affect more than one biological activity or process, and the abundances, spatial or temporal distributions, or other qualities of several habitat elements may affect

Attachment 1 – Page 3 Northern Mexican Gartersnake (Thamnophis eques megalops) (NMGS) Basic Conceptual Ecological Model for the Lower Colorado River

each biological activity or process. Finally, the rate of each critical biological activity or process may contribute to the rates of more than one life-stage outcome.

Integrating this information across all life stages for a species provides a detailed picture of: (1) what is known, with what certainty, and the sources of this information; (2) critical areas of uncertain or conflicting science that demand resolution to better guide LCR MSCP management planning and action, (3) crucial attributes to use to monitor system conditions and predict the effects of experiments, management actions, and other potential agents of change, and (4) how managers may expect the characteristics of a resource to change as a result of changes to controlling factors, including changes in management actions.

Conceptual Ecological Models as Hypotheses

The CEM for each species produced with this methodology constitutes a collection of hypotheses for that species. These hypotheses concern: (1) the species’ life history; (2) the species’ habitat requirements and constraints; (3) the factors that control the quality, abundance, and spatial and temporal distributions of these habitat conditions; and (4) the causal relationships among these. Knowledge about these model components and relationships may vary, ranging from well settled to very tentative. Such variation in the certainty of current knowledge always arises as a consequence of variation in the types and amount of evidence available and in the ecological assumptions applied by different experts.

Wherever possible, the information assembled for the LCR MSCP species CEMs documents the degree of certainty of current knowledge concerning each component and linkage in the model. This certainty is indicated by the quality, abundance, and consistency of the available evidence and by the degree of agreement/disagreement among the experts. Differences in the interpretations or arguments offered by different experts may be represented as alternative hypotheses. Categorizing the degree of agreement/disagreement concerning the components and linkages in a CEM makes it easier to identify topics of greater uncertainty or controversy.

Characterizing Causal Relationships

A causal relationship exists when a change in one condition or property of a system results in a change in some other condition or property. A change in the

Attachment 1 – Page 4 Attachment 1 – Species Conceptual Ecological Model Methodology for the Lower Colorado River Multi-Species Conservation Program

first condition is said to cause a change in the second condition. The present CEM methodology includes methods for assessing causal relationships (links) along four dimensions (attributes) adapted from the ERP methodology (DiGennaro et al. 2012):

(1) The character and direction of the effect

(2) The magnitude of the effect

(3) The predictability (consistency) of the effect

(4) The certainty of present scientific understanding of the effect

The present and ERP methodologies for assessing causal linkages differ in three ways. First, the ERP methodology assesses these four attributes for the cumulative effect of the entire causal chain leading up to each outcome. However, the LCR MSCP methodology recognizes six different types of causal linkages as described above. This added level of detail and complexity makes it difficult, in a single step, to assess the cumulative effects of all causal relationships that lead up to any one individual causal link. For example, in the present methodology, the effect of a given critical biological activity or process on a particular life-stage outcome may depend on the effects of several habitat elements on that critical biological activity or process which, in turn, may depend on the effects of several controlling factors. For this reason, the present methodology assesses the four attributes separately for each causal link by itself rather than attempting to assess cumulative effects of all causal linkages leading to the linkage of interest. The present methodology assesses cumulative effects instead through analyses of the data assembled on all individual linkages. The analyses are made possible by assembling the data on all individual linkages in a spreadsheet as described below.

Second, the present methodology explicitly divides link magnitude into three separate subattributes and provides a specific methodology for integrating their rankings into an overall ranking for link magnitude: (1) link intensity, (2) link spatial scale, and (3) link temporal scale. In contrast, the ERP methodology treats spatial and temporal scale together and does not separately evaluate link intensity. The present methodology defines link intensity as the relative strength of the effect of the causal node on the affected node at the places and times where the effect occurs. Link spatial scale is the relative spatial extent of the effect of the causal node on the affected node. Link temporal scale is the relative temporal extent of the effect of the causal node on the affected node. The present methodology defines link magnitude as the average of the separate rankings of link intensity, spatial scale, and temporal scale as described below.

Attachment 1 – Page 5 Northern Mexican Gartersnake (Thamnophis eques megalops) (NMGS) Basic Conceptual Ecological Model for the Lower Colorado River

Third, the ERP methodology addresses a single, large landscape, while the present methodology needed the flexibility to generate models applicable to a variety of spatial scopes. For example, the present methodology needed to support modeling of a single restoration site, the LCR main stem and floodplain, or the entire Lower Colorado River Basin. Consequently, the present methodology assesses the spatial scale of cause-effect relationships only relative to the spatial scope of the model.

The LCR MSCP conceptual ecological model methodology thus defines the four attributes for a causal link as follows:

• Link character – This attribute categorizes a causal relationship as positive, negative, involving a threshold response, or “complex.” “Positive” means that an increase in the causal node results in an increase in the affected node, while a decrease in the causal node results in a decrease in the affected node. “Negative” means that an increase in the causal node results in a decrease in the affected element, while a decrease in the causal node results in an increase in the affected node. Thus, “positive” or “negative” here do not mean that a relationship is beneficial or detrimental. The terms instead provide information analogous to the sign of a correlation coefficient. “Threshold” means that a change in the causal agent must cross some value before producing an effect. “Complex” means that there is more going on than a simple positive, negative, or threshold effect. In addition, this attribute categorizes a causal relationship as uni- or bi-directional. Bi-directional relationships involve a reciprocal relationship in which each node affects the other.

• Link magnitude – This attribute refers to “… the degree to which a linkage controls the outcome relative to other drivers” (DiGennaro et al. 2012). Magnitude takes into account the spatial and temporal scale of the causal relationship as well as the strength (intensity) of the relationship in individual locations. The present methodology provides separate ratings for the intensity, spatial scale, and temporal scale of each link, as defined above, and assesses overall link magnitude by averaging these three elements. Just as the terms for link character provide information analogous to the sign of a correlation coefficient, the terms for link magnitude provide information analogous to the size of a correlation coefficient. Tables 1-1 through 1-4 present the rating framework for link magnitude.

• Link predictability – This attribute refers to “… the degree to which the current understanding of the system can be used to predict the role of the driver in influencing the outcome. Predictability … captures variability …[and recognizes that] effects may vary so much that properly measuring and statistically characterizing inputs to the model are difficult”

Attachment 1 – Page 6 Attachment 1 – Species Conceptual Ecological Model Methodology for the Lower Colorado River Multi-Species Conservation Program

(DiGennaro et al. 2012). A causal relationship may be unpredictable because of natural variability in the system or because its effects depend on the interaction of other factors with independent sources for their own variability. Just as the terms for link character provide information analogous to the sign of a correlation coefficient, the terms for link predictability provide information analogous to the size of the range of error for a correlation coefficient. Table 1-5 presents the scoring framework for link predictability.

• Link understanding refers to the degree of agreement represented in the scientific literature and among experts in understanding how each driver is linked to each outcome. Table 1-6 presents the scoring framework for understanding. Link predictability and understanding are independent attributes. A link may be considered highly predictable but poorly understood or poorly predictable but well understood.

Conceptual Ecological Model Documentation

The documentation for each CEM provides information in three forms: (1) a narrative report, (2) causal diagrams showing the model components and their causal linkages for each life stage, and (3) a spreadsheet that is used to record the detailed information (e.g., linkage attribute ratings) for each causal linkage. The spreadsheet and diagrams, built using Microsoft Excel and Microsoft Visio, respectively, are linked so that the diagrams provide a fully synchronized summary of the information in the spreadsheet. This linkage between the two applications, supported by software scripts developed in association with these CEMs, allows users to generate a “master” diagram for each life stage from the data in the spreadsheet and, crucially, to query the CEM spreadsheet for each life stage and generate diagrams that selectively display query results concerning that life stage.

The narrative report for each species presents the definitions and rationales for the life stages/events and their outcomes identified for the species’ life history; the critical biological activities and processes identified for each life stage; the habitat elements identified as supporting or impeding each critical biological activity or process for each life stage; the controlling factors identified as affecting the abundance, spatial and temporal distributions, and other qualities of the habitat elements for each life stage; and the causal linkages among these model components.

The narrative report includes causal diagrams (aka “influence diagrams”) for each life stage. These diagrams show the individual components, or nodes, of the model for that stage (life-stage outcomes, critical biological activities and

Attachment 1 – Page 7 Northern Mexican Gartersnake (Thamnophis eques megalops) (NMGS) Basic Conceptual Ecological Model for the Lower Colorado River

processes, habitat elements, and controlling factors) and their causal relationships. The causal relationships (causal links) are represented by arrows indicating which nodes are linked and the directions of the causal relationships. The attributes of each causal link are represented by varying line thickness, line color, and other visual properties as shown on figure 1-1. The diagram conventions mostly follow those in the ERP methodology (DiGennaro et al. 2012).

The spreadsheet for each CEM contains a separate worksheet for each life stage. Each row in the worksheet for a life stage represents a single causal link. Table 1-7 lists the fields (columns) recorded for each causal link.

Link Attribute Ratings, Spreadsheet Fields, and Diagram Conventions

Table 1-1.—Criteria for rating the relative intensity of a causal relationship – one of three variables in the rating of link magnitude (after DiGennaro et al. 2012, Table 2)

Link intensity – the relative strength of the effect of the causal node on the affected node at the places and times where the effect occurs. High Even a relatively small change in the causal node will result in a relatively large change in the affected node at the places and times where the effect occurs. Medium A relatively large change in the causal node will result in a relatively large change in the affected node; a relatively moderate change in the causal node will result in no more than a relatively moderate change in the affected node; and a relatively small change in the causal node will result in no more than a relatively small change in the affected node at the places and times where the effect occurs. Low Even a relatively large change in the causal node will result in only a relatively small change in the affected node at the places and times where the effect occurs. Unknown Insufficient information exists to rate link intensity.

Attachment 1 – Page 8 Attachment 1 – Species Conceptual Ecological Model Methodology for the Lower Colorado River Multi-Species Conservation Program

Table 1-2.—Criteria for rating the relative spatial scale of a cause-effect relationship – one of three variables in the rating of link magnitude (after DiGennaro et al. 2012, Table 1)

Link spatial scale – the relative spatial extent of the effect of the causal node on the affected node. The rating takes into account the spatial scale of the cause and its effect. Large Even a relatively small change in the causal node will result in a change in the affected node across a large fraction of the spatial scope of the model. Medium A relatively large change in the causal node will result in a change in the affected node across a large fraction of the spatial scope of the model; a relatively moderate change in the causal node will result in a change in the affected node across no more than a moderate fraction of the spatial scope of the model; and a relatively small change in the causal node will result in a change in the affected node across no more than a small fraction of the spatial scope of the model. Small Even a relatively large change in the causal node will result in a change in the affected node across only a small fraction of the spatial scope of the model. Unknown Insufficient information exists to rate link spatial scale.

Attachment 1 – Page 9 Northern Mexican Gartersnake (Thamnophis eques megalops) (NMGS) Basic Conceptual Ecological Model for the Lower Colorado River

Table 1-3.—Criteria for rating the relative temporal scale of a cause-effect relationship – one of three variables in the rating of link magnitude (after DiGennaro et al. 2012, Table 1)

Link temporal scale – the relative temporal extent of the effect of the causal node on the affected node. The rating takes into account the temporal scale of the cause and its effect. Large Even a relatively small change in the causal node will result in a change in the affected node that persists or recurs over a relatively large span of time—decades or longer—even without specific intervention to sustain the effect. Medium A relatively large change in the causal node will result in a change in the affected node that persists or recurs over a relatively large span of time— decades or longer—even without specific intervention to sustain the effect; a relatively moderate change in the causal node will result in a change in the affected node that persists or recurs over only a relatively moderate span of time—one or two decades—without specific intervention to sustain the effect; a relatively small change in the causal node will result in a change in the affected node that persists or recurs over only a relatively short span of time—less than a decade—without specific intervention to sustain the effect. Small Even a relatively large change in the causal node will result in a change in the affected node that persists or recurs over only a relatively short span of time—less than a decade—without specific intervention to sustain the effect. Unknown Insufficient information exists to rate link temporal scale.

Table 1-4.—Criteria for rating the overall relative link magnitude of a cause-effect relationship based on link intensity, spatial scale, and temporal scale

Link magnitude – the overall relative magnitude of the effect of the causal node on the affected node based on the numerical average for link intensity, spatial scale, and temporal scale. (Calculated by assigning a numerical value of 3 to “High” or “Large,” 2 to “Medium,” 1 to “Low” or “Small,” and not counting missing or “Unknown” ratings.) High Numerical average ≥ 2.67 Medium Numerical average ≥ 1.67 but < 2.67 Low Numerical average < 1.67 Unknown No subattribute is rated High/Large, Medium, or Low/Small, but at least one subattribute is rated Unknown.

Attachment 1 – Page 10 Attachment 1 – Species Conceptual Ecological Model Methodology for the Lower Colorado River Multi-Species Conservation Program

Table 1-5.—Criteria for rating the relative predictability of a cause-effect relationship (after DiGennaro et al. 2012, Table 3)

Link predictability – the statistical likelihood that a given causal agent will produce the effect of interest. High Magnitude of effect is largely unaffected by random variation or by variability in other ecosystem dynamics or external factors. Medium Magnitude of effect is moderately affected by random variation or by variability in other ecosystem processes or external factors. Low Magnitude of effect is strongly affected by random variation or by variability in other ecosystem processes or external factors. Unknown Insufficient information exists to rate link predictability.

Table 1-6.—Criteria for rating the relative understanding of a cause-effect relationship (after DiGennaro et al. 2012, Table 3)

Understanding – the degree of agreement in the literature and among experts on the magnitude and predictability of the cause-effect relationship of interest. High Understanding of the relationship is subject to little or no disagreement or uncertainty in peer-reviewed studies from within the ecosystem of concern or in scientific reasoning among experts familiar with the ecosystem. Understanding may also rest on well-accepted scientific principles and/or studies in highly analogous systems. Medium Understanding of the relationship is subject to moderate disagreement or uncertainty in peer-reviewed studies from within the ecosystem of concern and in scientific reasoning among experts familiar with the ecosystem. Low Understanding of the relationship is subject to wide disagreement, uncertainty, or lack of evidence in peer-reviewed studies from within the ecosystem of concern and in scientific reasoning among experts familiar with the ecosystem. Unknown (The “Low” rank includes this condition).

Attachment 1 – Page 11 Northern Mexican Gartersnake (Thamnophis eques megalops) (NMGS) Basic Conceptual Ecological Model for the Lower Colorado River

Table 1-7.—Organization of the worksheet for each life stage Col. Label Content A Species Identifies the species being modeled by four-letter code B Link# Contains a unique identification number for each causal link C Life Stage Identifies the life stage affected by the link D Causal Node Type Identifies whether the causal node for the link is a controlling factor, habitat element, critical biological activity or process, or life-stage outcome E Causal Node Identifies the causal node in the link F Effect Node Type Identifies whether the effect node for the link is a controlling factor, habitat element, critical biological activity or process, or life-stage outcome G Effect Node Identifies the effect node in the link H Link Reason States the rationale for including the link in the conceptual ecological model, including citations as appropriate I Link Character Type Identifies the character of the link based on standard definitions J Link Character Direction Identifies whether the link is uni- or bi-directional K Link Character Reason States the rationale for the entries for Link Character Type and Link Character Direction, including citations as appropriate L Link Intensity Shows the rating of link intensity based on the definitions in table 1-1 M Link Spatial Scale Shows the rating of link spatial scale based on the definitions in table 1-2 N Link Temporal Scale Shows the rating of link temporal scale based on the definitions in table 1-3 O Link Average Magnitude Shows the numerical average rating of link intensity, spatial scale, and temporal scale based on the definitions in table 1-4 P Link Magnitude Rank Shows the overall rating of link magnitude based on the Link Average Magnitude, grouped following the criteria in table 1-4 Q Link Magnitude Reason States the rationale for the ratings for link intensity, spatial scale, and temporal scale, with citations as appropriate R Link Predictability Rank Shows the rating of link predictability based on the definitions in table 1-5 S Link Predictability Reason States the rationale for the rating of link predictability, with citations as appropriate T Link Understanding Rank Shows the rating of link understanding based on the definitions in table 1-6 U Link Understanding Reason States the rationale for the rating of link predictability, including comments on alternative interpretations and publications/experts associated with different interpretations when feasible, with citations as appropriate V Management Questions Briefly notes questions that appear to arise from the preceding entries for the link, focused on critical gaps or uncertainties in knowledge concerning management actions and options, with reasoning, including the estimate of relative importance when possible W Research Questions Brief notes that appear to arise from the preceding entries for the link, focused on critical gaps or uncertainties in basic scientific knowledge, with reasoning, including the estimate of relative importance when possible X Other Comments Provides additional notes on investigator concerns, uncertainties, and questions. Y Update Status Provides information on the history of editing the information on this link for updates carried out after completion of an initial version

Attachment 1 – Page 12 Attachment 1 – Species Conceptual Ecological Model Methodology for the Lower Colorado River Multi-Species Conservation Program

Link Magnitude (line thickness) Controlling Fac tor High – thick line Medium – medium line

Link# Low – thin line Unknown – very thin line

Habitat Element Link Understanding (line color) High – black line

Link# Medium – blue line Low – red line

Critical Biological Activity or Process Link Predictability (link label color) High – black text Link# Medium – blue text Low – red text

Life-Stage Outcome Unknown – grey text

Figure 1-1.—Conventions for displaying cause and effect nodes, linkages, link magnitude, link understanding, and link predictability.

Attachment 1 – Page 13 Attachment 1 – Species Conceptual Ecological Model Methodology for the Lower Colorado River Multi-Species Conservation Program

LITERATURE CITED IN ATTACHMENT 1

Burke, M., K. Jorde, and J.M. Buffington. 2009. Application of a hierarchical framework for assessing environmental impacts of dam operation: changes in streamflow, bed mobility and recruitment of riparian trees in a western North American river. Journal of Environmental Management 90:S224– S236.

DiGennaro, B., D. Reed, C. Swanson, L. Hastings, Z. Hymanson, M. Healey, S. Siegel, S. Cantrell, and B. Herbold. 2012. Using conceptual models and decision-support tools to guide ecosystem restoration planning and adaptive management: an example from the Sacramento–San Joaquin Delta, California. San Francisco Estuary and Watershed Science 10(3):1–15. http://escholarship.org/uc/item/3j95x7vt

Fischenich, J.C. 2008. The application of conceptual models to ecosystem restoration. Technical Note ERDC/EBA TN-08-1. U.S. Army Corps of Engineers, Engineer Research and Development Center (ERDC), Ecosystem Management and Restoration Research Program (EMRRP), Vicksburg, Mississippi. February 2008.

Kondolf, G.M., J.G. Williams, T.C. Horner, and D. Milan. 2008. Assessing physical quality of spawning habitat. Pages 249–274 in D.A. Sear and P. DeVries (editors). Salmonid Spawning Habitat in Rivers: Physical Controls, Biological Responses, and Approaches. American Fisheries Society Symposium 65. American Fisheries Society, Bethesda, Maryland.

McDonald, D.B. and H. Caswell. 1993. Matrix methods for avian demography. Pages 139–185 in D.M. Power (editor). Current Ornithology. Plenum Press: New York, New York.

Wildhaber, M.L., A.J. DeLonay, D.M. Papoulias, D.L. Galat, R.B. Jacobson, D.G. Simpkins, P.J. Baaten, C.E. Korschgen, and M.J. Mac. 2007. A Conceptual Life-History Model for Pallid and Shovelnose Sturgeon. Circular 1315. U.S. Geological Survey, Reston, Virginia.

_____. 2011. Identifying structural elements needed for development of a predictive life-history model for pallid and shovelnose sturgeons. Journal of Applied Ichthyology 27:462–469

Attachment 1 – Page 15

ATTACHMENT 2

Northern Mexican Gartersnake (Thamnophis eques megalops) (NMGS) Habitat Data

Table 2-1.—Northern Mexican gartersnake (Thamnophis eques megalops) habitat data (Note: The data presented in this table reflect those available in the literature at the time this model was developed. These data have not been validated.)

Habitat element Value or range Location Reference No values in literature, only mention of Range-wide U.S. Fish and Wildlife potential effects. Service (USFWS) 2014 Trampling of riparian vegetation. Range-wide USFWS 2014 Biological monitoring activities, minnow Range-wide; USFWS 2014; Lower trapping, drowning; new guidelines and Lower Colorado Colorado River Multi- Anthropogenic protocols. River (LCR) Species Conservation disturbance Program 2004; C. Ronning 2020; B. Sabin 2020 Handling of live animals, porous skin and Range-wide Todd and Glaudas 2013 chemicals. Road traffic. Range-wide USFWS 2014 American bullfrogs (Lithobates catesbeiana) Range-wide Rosen and Schwalbe 1988; and crayfish (Oronectes virilis and USFWS 2014, 2020 Procambarus clarki) compete with NMGS. Competitors may include other gartersnake Arizona Myatt 2016; O’Donnell et species (e.g., Marcy’s checkered gartersnake al. 2019; Rosen and Competitors [T. marcianus marcianus] and black-necked Schwalbe 1988; B. Sabin gartersnakes [T. cyrtopsis]) as well as other 2021 snake species (e.g., non-native southern watersnake [Nerodia fasciata]). Feral hogs (Sus scrofa) may compete with LCR C. Ronning 2020 NMGS for food resources. Original rule mentioned salinity < 5 parts per Range-wide USFWS 2014, 2020 thousand and ph > 5.6 to support native amphibian prey. However, they have a broader diet than originally suspected; some prey species may be more tolerant.

Environmental Selenium, mercury, and ammonia all reported Range-wide; LCR Sabin 2018; USFWS 2014 contaminants as possible environmental contaminants in NMGS habitat. Ammonia detected in Alamo Lake upstream of Bill Williams River Ajami et al. 2005 NMGS habitat on the Bill Williams River. Optimal water quality – absent of pollutants, Range-wide USFWS 2020 or of low levels to prevent NMGS recruitment.

Attachment 2 – Page 1 Northern Mexican Gartersnake (Thamnophis eques megalops) (NMGS) Basic Conceptual Ecological Model for the Lower Colorado River

Table 2-1.—Northern Mexican gartersnake (Thamnophis eques megalops) habitat data (Note: The data presented in this table reflect those available in the literature at the time this model was developed. These data have not been validated.)

Habitat element Value or range Location Reference Seem to prefer native frogs, especially (Rana LCR Rosen and Schwalbe 1988; yavapaiensis) and Chiricahua leopard frogs USFWS 2014 (R. chirichuensis) and juveniles and adults of native fish species. Lack of native food reduced NMGS San Bernardino Rosen and Schwalbe 1988 population. National Wildlife Refuge, southeast Arizona Will occasionally eat invertebrates (leeches, Range-wide d’Orgeix et al. 2013; earthworms) and other vertebrate taxa Emmons and Nowak 2016 (e.g., small mammals, salamanders, toads, in Northern Arizona lizards). University (NAU) 2020; USFWS 2014 Food availability If necessary, NMGS will eat non-native Range-wide Emmons and Nowak 2013, species, including larval and juvenile American 2016 in NAU 2020; bullfrogs, crayfish, mosquitofish (Gambusia Emmons et al. 2016a; affinis) and other soft-rayed fish. USFWS 2020 NMGS have eaten Chinese mystery snails Bubbling Ponds Fish Young and Boyarski 2012a (Cipangopaludina chinensis). Hatchery, Arizona in USFWS 2014 Birth may coincide with influx of prey in a Lake Tecocomulco, d’Orgeix et al. 2013; specific area (e.g., earthworms and leeches in Mexico; Appleton- Macías-García and Mexico, New Mexican spadefoot toad [Spea Whittell Research Drummond 1988 multiplicate] breeding sites, and larval Ranch, Santa Cruz Ambystoma tigrinum. County, Arizona Have eaten spiny-rayed fishes. Bubbling Ponds Fish Emmons et al. 2016a Hatchery

Attachment 2 – Page 2 Attachment 2 – Northern Mexican Gartersnake (Thamnophis eques megalops) Habitat Data

Table 2-1.—Northern Mexican gartersnake (Thamnophis eques megalops) habitat data (Note: The data presented in this table reflect those available in the literature at the time this model was developed. These data have not been validated.)

Habitat element Value or range Location Reference Genetic No values in the literature. diversity Hydrologic regime that supports perennially U.S. populations USFWS 2020 or spatially intermittent streams, including those with ephemeral reaches. Critical habitats include lentic habitats such as U.S. populations USFWS 2014, 2020 stock tanks, cienegas, and springs; also slow- moving streams with in-stream pools, off channel pools, and backwater habitat. Natural flow regime that allows for periodic Range-wide USFWS 2020 flooding and water movement through stream network. Mean distance to water ranged from 3.9 to Tonto Creek, Arizona Nowak et al. 2019 312.5 feet (ft) (1.2 to 95.5 meters [m]). Distance from brumation sites to water Bubbling Ponds Fish Boyarski et al. 2015 in ranged from 1.6 to 558 ft (0.49 to 170.1 m). Hatchery, Arizona Emmons et al. 2016a; Hydrologic Emmons and Nowak 2016 regime in NAU 2020; Myrand et al. 2017 in Emmons 2017. Distance from brumation sites to water Verde River, Arizona Emmons and Nowak 2016 averaged 129 ft (39.3 m). in NAU 2020 Brumation sites for 14 NMGS ranged from Tonto River, Arizona Nowak et al. 2019 2 to 1,257 ft (0.61 to 383.1 m) from water’s edge. Brumation sites for seven NMGS were Bubbling Ponds Fish Boyarski et al. 2015 in recorded within 230 ft (7 m) of ponds; one Hatchery, Arizona Emmons et al. 2016a snake overwintered 1,115 ft (350 m) from a pond. Most NMGS detections occur in the water or North-central Emmons and Nowak 2016 in adjacent meadows or woodlands within the Arizona in NAU 2020 floodplain.

Attachment 2 – Page 3 Northern Mexican Gartersnake (Thamnophis eques megalops) (NMGS) Basic Conceptual Ecological Model for the Lower Colorado River

Table 2-1.—Northern Mexican gartersnake (Thamnophis eques megalops) habitat data (Note: The data presented in this table reflect those available in the literature at the time this model was developed. These data have not been validated.)

Habitat element Value or range Location Reference Snake fungal disease (SFD) (Ophidiomyces Northern Arizona NAU 2020 ophiodiicola) was found in northern Arizona in 2019.

Infectious A number of parasites have been found in Bubbling Ponds Fish Boyarski 2008 in USFWS agents NMGS, including Spirometra sp. tapeworms, a Hatchery, Arizona 2014; nematode parasite, Macdonaldius sp., and Nowak et al. 2014; Sabin 13 species of helminth (= 11) and trematode 2018 and references (= 2) parasites. It is not known if they carry therein any infectious agents. For NMGS, may include American bullfrogs, Range-wide USFWS 2014, 2020; crayfish, raptors – red-tailed hawk (Buteo Emmons et al. 2016b jamaicensis), and others. Potential predators: common black hawk Arizona Etzel et al. 2014 (common (Buteogallus anthracinus) and river otter black hawk); Nowak et al. Predators (Lontra canadensis). 2011 (river otter) Feral hogs may prey on NMGS. Arizona U.S. Department of Agriculture 2019; C. Ronning 2020 Non-native fishes prey on NMGS. Range-wide USFWS 2014 No specific soil measurements in the literature. Definition includes organic or inorganic Range-wide USFWS 2020 structural complexity, including boulders, rocks, downed trees or logs, small mammal Substrate burrows, or leaf litter. Terrestrial substrate needs to be friable, Range-wide something that small animals can burrow into, as NMGS use their burrows as hibernacula.

Attachment 2 – Page 4 Attachment 2 – Northern Mexican Gartersnake (Thamnophis eques megalops) Habitat Data

Table 2-1.—Northern Mexican gartersnake (Thamnophis eques megalops) habitat data (Note: The data presented in this table reflect those available in the literature at the time this model was developed. These data have not been validated.)

Habitat element Value or range Location Reference Includes perennial streams, but can be Range-wide USFWS 2020 spatially intermittent, with ephemeral reaches Surface water acting as corridors. connectivity Best if free of non-native species Range-wide USFWS 2020 (e.g., American bullfrogs, crayfish, spiny-rayed fishes) that prey on NMGS. NMGS most active between temperature of Arizona Rosen 1991 Temperature 22 to 33 degrees Celsius (71–91 degrees Fahrenheit). NMGS require clear water to locate prey, Range-wide Sabin 2018; USFWS 2014 Turbidity although able to use chemical scents as well. Can use a variety of vegetation types but Range-wide USFWS 2020 require dense vegetation or other natural structural components that provide cover. NMGS use shallow, dense marshes – not Havasu Wildlife Bourne 2020 found in areas of open water. Management Area, Arizona Typical aquatic vegetation along streams and Arizona Rosen and Schwalbe 1988 in and adjacent to cienegas in Arizona that support NMGS includes knot grass (Paspalem distichum), spike rush (Eleocharis spp.), bulrush (Scirpus spp.), and cattails (Typha spp.). Deer grass (Muhlenbergia spp.), sacaton (Sporobolus Vegetation spp.), velvet mesquite (Prosopis velutina) with structure Fremont cottonwood (Populus fremontii), willow (Salix spp.), and seep willow (Baccharis salicifolis) are also commonly present. NMGS will also use bank vegetation with introduced species such as Bermudagrass (Cynadon dactylon). Sometimes found in areas with less vegetative Verde River, Arizona Emmons and Nowak 2016 cover using deeper water. in NAU 2020 Vegetation of brumation sites: males selected Bubbling Pond Fish Sprague and Bateman areas with higher amounts of vegetation and Hatchery, Arizona 2018 shrubs; females used burrows with higher canopy cover ≥ 3.3 ft (1 m).

Attachment 2 – Page 5 Attachment 2 – Northern Mexican Gartersnake (Thamnophis eques megalops) Habitat Data

LITERATURE CITED IN ATTACHMENT 2

Ajami, H., D.P. Guertin, L.R. Levick, and K. Uhlman. 2005. NEMO Watershed Based Plan Bill Williams Watershed. University of Arizona, Tucson, Arizona.

Bourne, J. 2020. Surveys for the northern Mexican gartersnake (Thamnophis eques megalops) at Havasu National Wildlife Refuge. Colorado River Terrestrial and Riparian meeting, McLaughlin, Nevada.

d’Orgeix, C.A., T. Mathies, B.L. Ellison, K.L. Johnson, I.V. Monagan, and T.A. Young. 2013. Northern Mexican gartersnakes, Thamnophis eques megalops, feeding on Spea multiplicata in an ephemeral pond. Herpetological Review 44(2):213–215.

Emmons, I.D. 2017. Ecology of the federally threatened northern Mexican gartersnakes in north-central Arizona. M.S. thesis. Northern Arizona University, Flagstaff, Arizona.

Emmons, I. and E. Nowak. 2013. Northern Mexican Gartersnake Surveys, 2012 (Interim Report). Northern Arizona University, Colorado Plateau Research Station, Flagstaff, Arizona. 56 p.

Emmons, I.D., E.M. Nowak, and K.K. Layuger. 2016a. Prey availability and foraging events of the northern Mexican gartersnake (Thamnophis eques megalops) in North-Central Arizona. Herpetological Review 47(4):555–561.

Emmons, I.D., E.M. Nowak, and T.C. Theimer. 2016b. Thamnophis eques megalops (northern Mexican gartersnake). Predation. Herpetological Review 47:485–486.

Etzel, K.E., T.C. Theimer, M.J. Johnson, and J.A. Holmes. 2014. Variation in prey delivered to common black-hawk (Buteogallus anthracinus) nests in Arizona drainage basins. Journal of Raptor Research 48(1):54–60.

Lower Colorado River Multi-Species Conservation Program (LCR MSCP). 2004. Lower Colorado River Multi-Species Conservation Program, Volume II: Habitat Conservation Plan. Final. December 17 (J&S 00450.00). Sacramento, California.

Macías-García, C.M. and H. Drummond. 1988. Seasonal and ontogenetic variation in the diet of the Mexican gartersnake Thamnophis eques in Lake Tecocomilco, Hidalgo. Journal of Herpetology 22(2):129–134.

Attachment 2 – Page 7 Northern Mexican Gartersnake (Thamnophis eques megalops) (NMGS) Basic Conceptual Ecological Model for the Lower Colorado River

Myatt, J. 2016. Battling an invasion of watersnakes. USFWS Pacific Southwest Region. Online newsroom. https://www.fws.gov/cno/newsroom/featured/2016/Invasive_Nerodia/

Northern Arizona University. 2020. Northern Mexican Gartersnake Project. Flagstaff, Arizona. https://in.nau.edu/Gartersnake-Research-Project/Northern-Mexican- Gartersnake/

Nowak, E.M., M. Liszewski, and I. Emmons. 2011. Surveys for Northern Mexican Gartersnakes in Tavasci Marsh (Tuzigoot National Monument), unpublished final report. U.S. Geological Survey, Colorado Plateau Research Station, Flagstaff, Arizona.

Nowak, E.M, V. Boyarski, S. Nichols, and T. Greene. 2014. Thamnophis eques megalops (northern Mexican gartersnakes): maternal transmission of parasites. Herpetological Review 45(1):148–149.

Nowak, E.M., I. Emmons, J. Myrand, C. Klovanish, and R. Bergamini. 2019. 2015–2017 surveys and telemetry of Northern Mexican Gartersnakes in Lower Tonto Creek. Unpublished final report to the Salt River Project. Colorado Plateau Research Station, Northern Arizona University, Flagstaff, Arizona.

O’Donnell, R.P., R. Mixan, S.L. Arnett-Romero, and M.F. Ingraldi. 2019. Surveys for Threatened and Endangered Species and their Potential Breeding Habitat: Yellow-billed Cuckoo and Mexican Garter Snake Surveys on the Bill Williams River and Burro Creek, Arizona. Arizona Game and Fish Department, Wildlife Contracts Branch, Phoenix, Arizona.

Ronning, C. 2020. Lower Colorado River Multispecies Conservation Program, Boulder City, Nevada, personal communication.

Rosen, P.C. 1991. Comparative field study of thermal preferenda in gartersnakes (Thamnophis). Journal of Herpetology 25(3):301–312.

Rosen, P.C. and C. Schwalbe. 1988. Status of the Mexican and Narrow-headed Gartersnakes (Thamnophis eques megalops and Thamnophis rufipunctatus rufipunctatus) in Arizona. Unpublished report to the U.S. Fish and Wildlife Service. Arizona Game and Fish Department, Phoenix, Arizona.

Sabin, L.B. 2018. Northern Mexican Gartersnake (Thamnophis eques megalops) Species Profile. Lower Colorado River Multi-Species Conservation Program, Boulder City, Nevada.

Attachment 2 – Page 8 Attachment 2 – Northern Mexican Gartersnake (Thamnophis eques megalops) Habitat Data

Sabin, B. 2020. Lower Colorado River Multispecies Conservation Program, Boulder City, Nevada, personal communication.

_____. 2021. Lower Colorado River Multispecies Conservation Program, Boulder City, Nevada, personal communication.

Sprague, T.A. and H.L. Bateman. 2018. Influence of seasonality and gestation on habitat selection by northern Mexican gartersnakes (Thamnophis eques megalops). PLOS One 13(1):e0191829. https://doi.org/10.1371/journal.pone.0191829

Todd, B.D. and X.A. Glaudas. 2013. Appendix I: Handling live amphibians and reptiles. Pages 214–219 in G.J. Graeter, K.A. Buhlmann, L.R. Wilkinson, and J.W. Gibbons (editors). Inventory and Monitoring: Recommended Techniques for Reptiles and Amphibians. Technical Publication IM-1. Partners in Amphibian and Reptile Conservation, Birmingham, Alabama.

U.S. Department of Agriculture. 2019. Wildlife Damage Management – Food habitats of feral hogs. Extension Service website. https://wildlife-damage-management.extension.org/food-habits-of-feral- hogs/

U.S. Fish and Wildlife Service (USFWS). 2014. Endangered and threatened wildlife and plants; threatened status for the northern Mexican gartersnake and narrow-headed gartersnake; final rule. 50 CFR 17. Federal Register 79(130):38678–38746.

_____. 2020. Designation of critical habitat for the northern Mexican gartersnake and narrow headed gartersnake, proposed rules. Federal Register 85(82):23608.

USFWS (see U.S. Fish and Wildlife Service).

Young, M.E. and V. Boyarski. 2012. Thamnophis eques megalops; diet and mortality. Herpetological Review 43(3):498.

Attachment 2 – Page 9