Large-scale collaboration reveals landscape-level effects of land-use on turtle demography

Abstract

Freshwater turtles and tortoises are declining worldwide and currently represent one of the most imperiled major vertebrate groups. Identifying the conditions that promote long-term viable populations is a critical conservation need. However, for most species, there is relatively little or no empirical information about the factors influencing population demographics. Large-scale population monitoring efforts necessary to acquire such information remain rare due to the logistic challenges associated with low and variable detectability, which generally preclude large monitoring initiatives by any single entity. The development of collaborative population monitoring programs represents one potential strategy for overcoming these challenges. Our goal was to leverage partnerships to identify the potential factors and relevant scales affecting wood turtle (Glyptemys insculpta) population demographics. Through a large-scale collaborative multi-institutional monitoring effort, we conducted 983 spring stream surveys at 293 sites across the northeastern United States. Wood turtle abundance was negatively associated with agriculture (300 m and 5500 m) and road traffic (5500 m) and positively associated with mature forest (5500 m). Juvenile proportion displayed strong negative relationships with stream gradient and imperviousness (300 m). Sex ratios were more male-skewed with higher mature forest cover (90 m) and road density (5500 m) and less undeveloped land (300 m). These findings suggest that effective conservation of demographically robust turtle populations will require consideration of multiple spatial scales. Landscape-level conservation may be particularly important for ensuring long-term viable populations. This study highlights the valuable role that collaboration across institutions and jurisdictions can play in the conservation of cryptic taxa.