Author ORCID Identifier

https://orcid.org/0009-0007-8184-6209

Semester

Spring

Date of Graduation

2024

Document Type

Thesis

Degree Type

MS

College

Davis College of Agriculture, Natural Resources and Design

Department

Division of Forestry and Natural Resources

Committee Chair

Christopher M. Lituma

Committee Co-Chair

Donald J. Brown

Committee Member

Michael T. Jones

Abstract

To better understand wood turtle (Glyptemys insculpta) population trends, two standardized population survey protocols were recently developed for use by state monitoring programs. The protocols were independently developed, tested, and deployed for population monitoring in the eastern and upper midwestern United States. It was unknown if the sampling designs result in similar detection success or if data from the two protocols are compatible for broad-scale analyses. Additionally, neither monitoring protocol includes a sampling design that allows for the separation of detection (p) into availability (pa), and detection given availability (pd). Wood turtles typically remain near streams year-round, but they spend a large portion of their time moving throughout the riparian and upland habitat during the active months, which includes moving in and out of the survey area. If pa is non-random and pd or pa are influenced by demographic factors such as sex or age class, then abundance estimates derived from standardize protocol data may be biased.

In Chapter 1, I first discuss the ecology of wood turtles, including distribution and habitat associations. I then provide information on wood turtle population survey techniques, focusing on the two standardized protocols referred to as the Eastern Protocol and Midwestern Protocol. I also discuss abundance modeling and the factors that can impact modeling accuracy. Lastly, I state the goals of this thesis research and provide a summary of the chapters.

In Chapter 2, we used survey results from a modified wood turtle sampling design that included a double observer approach within surveys. This allowed us to separately estimate pd and pa,to determine if pa is random or Markovian and what covariates influence pd or pa. We modified the wood turtle survey protocol used in the upper Midwest to include a double-pass design, allowing us to estimate pa and pd using a robust design capture-recapture model. Our results indicated that pa was non-random and that pd increased with turtle carapace length. Our study suggests that model assumptions for current wood turtle population models may be violated, likely resulting in an overestimation of abundance. The results of Chapter 2 improve our understanding of potential biases associated with the Midwestern protocol, and thus the modifications needed to obtain accurate population estimates.

In Chapter 3, we compared p and expected count estimates between constrained and unconstrained survey designs for aquatic-focused and terrestrial-focused surveys at wood turtle monitoring sites from across the distribution of the species within the United States. We estimated adult wood turtle p and expected count for aquatic-focused and terrestrial-focused surveys using N-mixture models and generalized linear mixed models, respectively. Our study supports that constrained and unconstrained surveys can be used interchangeably for population monitoring if survey type is accounted for in the population models. However, our findings indicated that, within sampling periods at monitoring sites, restricting all survey replicates to only aquatic or only terrestrial surveys is optimal for use in population abundance and trend models. This information can be used to guide the creation of a range-wide protocol that maximizes flexibility in survey design while maintaining data congruency across space.

Download

Share

COinS