Images

Karoline Weston is a 2025 graduate of the MS Program in Environmental Studies at Antioch University, New England.

Thesis Committee:

Lisabeth Willey, Ph.D., Thesis Advisor

Michael Akresh, Ph.D.

Henry Patrick Roberts, Ph.D.

Keywords

eastern box turtle, population decline, invasive vegetation, quantitative methodology

Document Type

Thesis

Publication Date

2026

Abstract

Eastern box turtles are facing population declines across their range due to habitat loss, vehicular strikes, and illegal collection. A lesser-explored aspect of eastern box turtle conservation is the role of non-native, invasive vegetation in shaping fine scale habitat use. Invasive vegetation can alter habitat structure and quality, yet its role in eastern box turtle behavior remains unstudied. To evaluate this relationship, I tracked 19 eastern box turtles via radio telemetry throughout the 2022 active season in a Massachusetts wildlife management area. My study addressed six questions: (1) Do eastern box turtles exhibit a preference for microhabitats containing invasive vegetation? (2) Does the amount (percent cover) of invasive vegetation affect selection? (3) Does selection vary across the active season? (4) Do males and females differ in invasive vegetation selection? (5) Do environmental variables—air temperature, substrate temperature, shrub cover, and canopy cover—influence the likelihood of turtles selecting microhabitats with invasive vegetation? and (6) Are these patterns consistent across individuals? To answer these questions, I utilized a combination of modeling techniques including paired logistic regression analysis and two sets of Generalized Linear Mixed Models (GLMMs) with individual turtle as a random effect. Paired logistic regression revealed no overall preference for invasive vegetation presence at turtle-selected versus paired random locations, with a 50% probability of turtles using microhabitats dominated by invasive plants. An AICc-based model comparison of my GLMMs revealed that models including percentage of canopy cover and percentage of shrub cover best explained the presence of invasive plants at turtle-selected locations. Top model results revealed that turtles were more likely to be found in areas with invasive vegetation when shrub density was high and canopy cover was low. Sensitivity analysis of the top models showed that these patterns were consistent after removing three turtles with unusual microhabitat selection patterns. My second set of GLMMs revealed that, across all turtles, turtles were more likely to select sites as invasive shrub cover became denser. Models for males and females yielded similar results, with the effect being stronger in males (slope = 1.07, 95% CI: 0.80 to 1.33) than in females (slope = 0.585, 95% CI: 0.379 to 0.790). Seasonality was not a significant predictor in any of the three models. Together, my findings suggest that invasive vegetation alone does not influence microhabitat selection by eastern box turtles. GLMM results indicated that plant density, rather than simple presence or absence, is an important factor influencing turtle microhabitat use. However, this may reflect structural or microclimatic associations rather than true selection for invasive vegetation. The neutrality of eastern box turtles toward invasive vegetation presence challenges assumptions that these plants are inherently avoided or preferred and underscores the need for further research. Conservation strategies should consider how canopy management and invasive plant control interact to shape microhabitats used by turtles. Further investigation is needed to identify more specific influences and long-term trends in habitat selection at finer scales. This dissertation is available in open access at AURA (https://aura.antioch.edu)

Download

Share

COinS