New Published Research: Landscape Structure and Turtle Communities


Common Snapping Turtle crossing a road in southern Georgia. – Houston Chandler

I have written extensively about the work that we have done as part of a large collaboration to create a Conservation and Management Plan for Spotted Turtles in the eastern portion of their range. The results of multiple years of fieldwork are now starting to make their way into the published literature. I wrote in June about our analysis of Spotted Turtle sex ratios and climate change. This month, I take a look at a recently published paper that leverages the large dataset built over several field seasons to examine how landscape characteristics influence abundance in freshwater turtle communities.

Spotted Turtles captured as part of recent status assessment surveys. – Houston Chandler

Landscape configuration is hypothesized to play an important role in the ecology of animal populations. However, in many groups (especially reptiles), we know relatively little about how landscape properties actually impact basic population biology. There are lots of hypotheses but little data to test these hypotheses, which generally requires large datasets collected over large geographic areas. A luxury that most studies do not have. Landscape configuration can be thought of as having two distinct properties: 1) the composition or diversity of habitat types and 2) the spatial arrangement of these different habitats. Freshwater turtles present an interesting case study for these ideas because they frequently make terrestrial movements out of wetland systems that likely depend on both landscape configuration and individual species’ biology. Furthermore, turtles are known to be sensitive to anthropogenic land use change, suggesting that humans are likely changing how populations respond to landscape configuration. Understanding these processes would allow us to make better management decisions for turtles at broad spatial scales.

Conceptual models outlining hypotheses of how freshwater turtle abundance may be impacted by landscape configuration. – Patrick Roberts

To explore these hypotheses, we used the combined sampling dataset from the recent Spotted Turtle status assessments. Data were collected in 15 states and the District of Columbia. These data included enough capture events to explore trends for six species: Blanding’s Turtles (Emydoidea blandingii), Common Snapping Turtles (Chelydra serpentine), Eastern Mud Turtles (Kinosternon subrubrum), Painted Turtles (Chrysemys picta), Spotted Turtles (Clemmys guttata), and Striped Mud Turtles (Kinosternon baurii). Six additional species were captured during surveys but were excluded from the analyses because of low sample size. We combined turtle trapping data with a variety of available environmental layers describing various aspects of the landscape. We quantified various types of anthropogenic land use hypothesized to impact turtle populations (e.g., amount of impervious surfaces, road density, and proportion of cultivated cropland). We also summarized various wetland-specific variables that quantified the diversity, amount, and configuration of wetlands across the study area. We combined all of these data sources into models examining how the abundance varied across landscapes.

In general, turtle abundance increased with increasing diversity of wetland types within the surrounding landscape. However, the pattern of these relationships was variable across species. Spotted Turtle abundance was positively related to wetland ephemerality, while Painted Turtle abundance was negatively related to ephemerality. Overall, these results tended to support our hypotheses for most species and provide evidence that promoting wetland diversity across the landscape would benefit turtle populations.

Relationships between turtle abundance and wetland diversity for six turtle species. Gray areas are 95% confidence intervals. – Patrick Roberts

Results were complicated by strong interactive relationships between the spatial arrangement of wetlands and anthropogenic land use (again, there was variability among species). For example, Spotted Turtle abundance was negative impacted by hay and pasture habitats at low values of wetland aggregation (i.e., isolated wetlands), but this negative relationship disappeared in areas with high wetland aggregation (i.e., lots of clustered wetlands). These interactions reinforce that species relationships with surrounding landscapes are complicated and that multiple factors can influence whether landscape changes have positive or negative effects on turtle populations. Such differences may also help to explain why studies of similar phenomena sometimes find different results (context matters).

Road density had a strong effect on four turtle species. However, this effect was negative only for Spotted Turtles and Eastern Mud Turtles. Surprisingly, road density had positive effects on Striped Mud Turtles and Common Snapping Turtles, although the error associated with these predictions was high at high road densities. Overall, these results suggest that the effect of roads on turtles is species-specific, and that caution is needed when making broad conclusions about potential threats to turtle populations. Species- and site-specific data are often needed to make strong conclusions.

Effects of road density on the abundance of four turtle species. Gray areas are 95% confidence intervals. – Patrick Roberts

This study leveraged a large, multi-state database to make inferences about the effects of landscape configuration on freshwater turtle abundance. Such opportunities are rare, and this was the first study to analytically show that wetland diversity generally promoted turtle abundance. Our results provide evidence that wetland loss could negatively impact turtles even if those wetlands do not provide core habitat. Furthermore, relationships across species and landscape characteristics were complex, and the effects of land use on turtle populations is context-dependent. This is an important consideration when making conservation and management decisions at a local scale. One solution will not fit all circumstances. Overall, our results indicate that conservation of wetland habitats should generally shift away from focusing on individual wetlands to multiscale landscape-level wetland conservation projects.

The research was published in the leading conservation journal, Biological Conservation. You can read the full publication here.