Date Approved

8-28-2025

Graduate Degree Type

Thesis

Degree Name

Biology (M.S.)

Degree Program

Biology

First Advisor

Jennifer Moore

Second Advisor

Eric McCluskey

Third Advisor

Megan Woller-Skar

Academic Year

2024/2025

Abstract

Turtles worldwide are declining from habitat loss and fragmentation, predation, and illegal collection, and their slow life histories and sensitivity to environmental change make reliable abundance and demographic data critical for conservation. Wood turtles (Glyptemys insculpta) are declining across much of their range, yet baseline information is lacking in Michigan’s Lower Peninsula. To fill that gap, we performed standardized visual mark–recapture surveys at 27 one-kilometer river reaches in two distinct watershed systems during springs of 2022 and 2023. Using hierarchical N-mixture models with negative binomial mixtures to account for imperfect detection, we estimated site-level latent abundance and evaluated how survey effort and habitat variables influenced detectability and abundance. We also characterized population composition via sex ratios and age-class structure. Per-survey detection probability was low (~5– 6%) but improved with increased effort. Northern river sites consistently hosted higher abundances than southern ones, and juvenile‐to‐adult representation was markedly greater in the north both years (2022: 1:0.5 north vs. 1:0.4 south; 2023: 1:0.6 north vs. 1:0.1 south), suggesting stronger recent recruitment or more persistent demographic stability, likely reflecting higher habitat quality, greater nesting habitat availability, and lower disturbance, while southern sites occur at the species’ range margin where populations often face additional environmental stressors and reduced resilience. Site-level latent abundance estimates were highly variable within and between watersheds—ranging from roughly 4–50 turtles per site in 2022 (mean ≈17, northern ≈23 vs. southern ≈9) and 2–40 in 2023 (mean ≈14, northern ≈20 vs. southern ≈6)— indicating concentrated, relatively robust populations in the north alongside much lower and potentially more vulnerable abundances in the south. This pattern highlights the need to prioritize protection of northern strongholds while closely monitoring and supporting recruitment in southern reaches. Relative abundance rankings of sites, evaluated using Pearson’s correlation, were largely maintained between consecutive springs, indicating short-term temporal consistency. These results provide a geographically resolved baseline and demographic context for wood turtle populations in the region. By pairing detection-corrected abundance estimation with demographic indicators, this study offers practical guidance for allocating monitoring effort, maintaining population viability, and directing targeted interventions in more vulnerable areas.

Available for download on Sunday, September 27, 2026

Share

COinS