More than half of the chelonians that inhabit our planet are currently threatened with extinction. Risks to chelonians include many factors, such as capture for use in medicine, capture for food, capture for the pet trade, habitat loss, and climate change (Stanford et al. 2020). In particular, climate change and habitat loss could drive populations of freshwater turtles to decline rapidly across North America (Stanford et al. 2020). As poikilothermic species, turtles are physiologically dependent on the ambient temperatures of the environments they inhabit. This climatic dependency is why chelonians, and other poikilotherms, are at a high risk of facing large climate-change-related declines, and it is predicted that the bioclimatic niches of chelonians will change drastically (Ihlow et al. 2012). Additionally, many chelonians occupy wetland ecosystems, which are being degraded at a higher rate than other ecosystems (Millenium Ecosystem Assessment 2005). Another potential risk to freshwater turtles is habitat loss and fragmentation. Infrastructure development, land conversion, water withdrawal, eutrophication, and pollution are some of the largest causes of observed habitat loss and degradation for turtle species (Millenium Ecosystem Assessment 2005). The combination of these risks faced by chelonians is a serious cause for concern. Understanding the behavior, natural history, and ecology of understudied species is of utmost importance for land management.

Chicken turtles (Deirochelys reticularia) are similar to many other chelonian species at risk and are assumed to be declining throughout their range (Ryberg et al. 2017). A lack of historical surveys, detailed demographic studies, and fine-scale microhabitat and spatial research of this species creates a limited understanding of their natural history, behavior, and ecology (McKnight et al. 2022). Chicken turtles typically inhabit locations that undergo large seasonal changes in water levels and are generally absent from large permanent bodies of water (Buhlmann et al. 2009). Although chicken turtles are very resilient to varying levels of water within their environment, it is unknown how increasing temperatures and more frequent and intense drought, storms, and heat waves will impact their habitat. Moreover, they have many life history characteristics that are distinct when compared to other freshwater turtle species. Chicken turtles have a short lifespan and mature after approximately 5–6 years (Congdon et al. 2022). They have a thin shell that makes them susceptible to predation from other animals such as otters, mink, raccoons, coyotes, foxes, and herons (Jackson 1988). Chicken turtles are mostly carnivorous, feeding upon large amounts of aquatic invertebrates such as crayfish (Jackson 1996; Demuth and Buhlmann 1997; McKnight et al. 2015). Nesting habits within the eastern range of D. reticularia are also different compared to other chelonians; they typically nest from August to March, often with a gap between these nesting seasons due to cold temperatures (Gibbons 1969; Iverson 1977; Gibbons and Greene 1978; Jackson 1988). Additionally, chicken turtles can retain eggs for a prolonged period compared to most other chelonians. This strategy may aid in survival within the unpredictable wetland ecosystems that they inhabit, allowing for nesting to occur during the most favorable periods (Cagle and Tihen 1948; Buhlmann 1995). Although other factors of chicken turtle life history have been studied, fine-scale microhabitat requirements for nest site selection and adult survival remain ambiguous.

The geographic distribution of D. reticularia spans the coastal plains of the southeastern United States. They are absent from the Piedmont Mountains and exist as far north as North Carolina with a disjunct population in Virginia, south to Florida, and west as far as eastern Texas (Buhlmann et al. 2008). Chicken turtles also inhabit barrier island ecosystems, which are at a high risk of change due to anthropogenic impacts from an increase in human populations. Climate change has the potential to significantly impact habitat suitability for many barrier island species; rising sea levels and major weather events continue to increase the destabilization and erosion of coastal shorelines. Flooding and hurricanes continue to increase within barrier island systems (Culver et al. 2007), and these ecosystems are also at a high risk of drought (Bellis 1995). These drought patterns have been listed as reasons for concern by Mayes and List (1988), whereas traditionally they have been thought of as a natural pattern necessary to maintaining these systems (Hillestad et al. 1975). As the impacts of climate change continue to increase, these coastal wetland systems will undergo significant change (Day et al. 2008). In species distribution models presented by Ihlow et al. (2012), the southeastern United States is predicted to undergo a dramatic decrease in biodiversity, with the highest impacts occurring in coastal areas. The degradation and loss of habitat to coastal wetland buffers is often overlooked and has the potential to play a large role in the degradation of turtle habitat in these ecosystems (Quesnelle et al. 2013; Howell and Seigel 2019). Wetland buffers and terrestrial habitat also remain extremely important to chicken turtles because of their use for terrestrial habitat in migration between ponds (Gibbons and Greene 1978; Buhlmann and Gibbons 2001). Research into this highly sensitive and easily disrupted system, as well as a turtle species with unusual life history characteristics, can provide invaluable insight that pertains to the future of chelonians in general.

Previous research has investigated population dynamics (Gibbons 1969; Dinkelacker and Hilzinger 2014; Hanscom et al. 2020), spatial ecology (Ryberg et al. 2017; Bowers et al. 2021), and physiology (Gibbons 1969; Buhlmann et al. 2009; Dinkelacker and Hilzinger 2014; McKnight et al. 2018; Congdon et al. 2022) of chicken turtles throughout different portions of their range; however, most of this research has been conducted on mainland populations. When compared to other turtle species such as pond sliders (Trachemys scripta) and snapping turtles (Chelydra serpentina), research on life history and habitat requirements of chicken turtle populations has been extremely limited, especially within barrier island ecosystems. Recently, Ryberg et al. (2017) examined the macrohabitat suitability of the Western chicken turtle subspecies (D. r. maria) across their entire geographic range with a species distribution modeling approach developed by Labay et al. (2011). According to their analyses, southeast Texas and southwest Louisiana had the most suitable habitats, which were characterized by low elevations, consistent rainfall, and proximity to wetlands. Despite this research, understanding is lacking of what microhabitat factors are driving chicken turtle presence.

We investigated the fine-scale microhabitat use of chicken turtles using a unique barrier island interdunal pond ecosystem on the coast of North Carolina. Our goal was to determine which microhabitat variables are associated with chicken turtle occupancy within this ecosystem. This approach will provide land managers insight into conserving or developing ponds that promote chicken turtle occupancy.

METHODS

RESULTS

We found 24 chicken turtles (9 males, 9 females, and 6 juveniles) present in 6 out of 20 ponds (ranging from 1 to 12 unique individuals per pond) across the study period. Substantial variation was seen across all pond characteristics (Table 1). The data in Table 1 were used to develop the candidate models to understand the influence of abiotic factors on chicken turtle presence. Of the candidate models, the model that garnered most of the weight included canopy cover, water depth, and salinity (Table 2). Across all models, chicken turtles used ponds with less canopy cover (p < 0.01) and were found in ponds with shallower water depths except the model that garnered the least amount of weight that included only canopy cover and water depth (p < 0.05; Table 3, Fig. 1). Across all models in which salinity was included, chicken turtles were found in ponds with little to no salinity (p < 0.10; Table 3).

View Figure

• Download as Powerpoint
• Download Figure

Table 1. Environmental variables averaged across 3 sampling periods for ponds within Nags Head Woods Ecological Preserve where we trapped for chicken turtles (D. reticularia). Ponds where chicken turtles were found are in bold.

View Fullscreen

Table 2. Akaike Information Criterion (AIC) candidate models representing habitat occupancy by Deirochelys reticularia at NHWEP. We report the number of parameters (K), second-order information criterion (AICc), Akaike weights (AICcWt), and chi-squared goodness-of-fit (GoF).

View Fullscreen

Table 3. Poisson regression model results, parameter coefficients, and p-values for each variable in describing the occupancy of D. reticularia in the study area.

View Fullscreen

Ponds where D. reticularia were present tended to have near-average values for most environmental variables (Fig. 2), but negative relationships of occupancy were found with increase in canopy cover, water depth, and salinity (Table 3). The NMDS ordination of pond measurements, using Euclidean distances and 2 dimensions, resulted in a stress value of 0.11, indicating a good fit for the ordination. The PERMANOVA results showed that the presence of chicken turtles explained approximately 8.02% of the variation in pond characteristics (R² = 0.08, F = 1.57, p = 0.16); thus, ponds in which D. reticularia were found were similar in composition to all other ponds regarding all abiotic factors considered (Fig. 2).

View Figure

• Download as Powerpoint
• Download Figure

DISCUSSION

Over the course of the study, we were able to determine which environmental variables were likely to predict the presence of chicken turtles across the 20 ponds located at NHWEP. Chicken turtle presence was strongly correlated with low canopy cover, shallow pond depth, and little to no salinity. The 20 ponds sampled within our study showed substantial variation in canopy cover percentage (from 14% to 78%; Table 1); however, chicken turtles were present only in ponds that ranged in canopy cover from 14% to 36%, and interestingly, the pond within our study site that had the highest abundance of chicken turtles (n = 12) had the lowest average canopy cover of 14% across the study period. No previous study has examined how canopy cover of ponds influences the presence of chicken turtles, but in a study on the upland terrestrial estivation site selection of chicken turtles by McKnight et al. (2020), it was suggested that chicken turtles prefer estivation sites with very high percentage of canopy cover (50% to 100%). Within their study, they observed that many turtles estivated within sites that contained high amounts of canopy cover. It is yet to be determined if canopy cover preferences change from times of estivation to the active season, so these differences in data and observations between our studies are worth noting. Buhlmann and Gibbons (2001) also observed preference chicken turtles of closed canopy forests while seeking upland terrestrial refugia. The authors found that when seeking such refugia, chicken turtles preferred the more closed canopy of the mixed forest, rather than the more open canopy found within the nearby pine plantation. They also observed some use of terrestrial turtle refugia within open canopy areas such as clear cuts from D. reticularia. Though our results show lower percentages of canopy cover preferences than McKnight et al. (2019), these differences are most likely due to different microhabitat preferences during estivation.

Previous studies have consistently commented on the importance of water depth to the life history of D. reticularia. Shallow ponds such as interdunal wetlands are zones where chicken turtles have been studied previously (Buhlmann 1995). Within some studies, it has been suggested that chicken turtles prefer shallow water depths (Buhlmann et al. 2009; McKnight et al. 2022). Our research confirms that chicken turtles prefer shallow ponds, because ponds of different depths were available at our study site, but chicken turtles were found only in shallower ponds. At NHWEP, average water depth varied across the site from 66 to 255 cm, and ponds where chicken turtles were present averaged only depths of 67–100 cm. Ewert et al. (2006) mentions that chicken turtles prefer ponds with depths less than 50 cm, and association with water deeper than 200 cm is rare. Results from Ewert et al. remain consistent with observations of shallow water depth preferences; however, the data collected by our study differed slightly, with turtles being observed in ponds that were deeper than 50 cm. Studies that prioritize understanding species habitat preferences in relation to water depth are becoming crucial as the effects of climate change continue to disrupt coastal ecosystems. Our research observed a narrow range of preference for water depth, and this narrow range could aid in the determination of current suitable habitat for this species, as well as future habitat that has been modified because of various factors such as flooding and severe weather events.

Examining how changes in salinity may affect chicken turtle presence has not been investigated before; however, our study site presented a unique opportunity to quantify whether chicken turtles tolerated salinity or preferred ponds that contain low amounts (i.e., negligible amounts) of salinity. The barrier island ecosystem within the NHWEP contains ponds of varying salinity due to the coastal nature of the system and ponds that are interconnected to the Roanoke Sound, a body of water that contains diluted saltwater due to river outflow (Hanscom et al. 2020). Thus, within coastal zones of the geographic distribution of D. reticularia, pond preferences in relation to salinity become especially informative. Chicken turtles were observed only in ponds with a salinity that was negligible (0.1 ppt; Table 1). Although little variation was seen in salinity across all 20 ponds, 2 ponds would be considered brackish water (pond 5 = 20.1 ppt and pond 13 = 2.1 ppt; all other ponds were 0.1 ppt). We found that chicken turtles would not occupy those ponds with higher salinity despite that they were near ponds where chicken turtles were found and that other freshwater turtle species were found occupying those ponds. This salinity preference may seem obvious because of the freshwater nature of these turtles; however, it has large implications in terms of habitat suitability for this species within coastal regions and on barrier islands specifically. Salinity tolerance of freshwater turtles is variable between species (Dunson and Seidel 1986; Kinneary 2021). For instance, the common snapping turtle has some tolerance to salinity up to 35% seawater but is then unable to hyposmoregulate when salinity levels get too high and subsequently loses body mass (Kinneary 2021). Additionally, Pseudemys nelsoni and Trachemys decussata have a higher observed tolerance to saltwater than other true freshwater turtle species (i.e., Kinosternon bauri and Apalone ferox), but still had losses in body mass when in saltwater environments (Dunson and Seidel 1986). Terrapins are the only nonmarine turtle species that have a lachrymal gland, which helps with regulation of salt within their bodies (Schmidt-Nielsen and Fange 1958; Agha et al. 2018). With sea levels projected to rise in the future, this poses large implications for coastal habitats. Although chicken turtles have been observed within brackish systems by Neill (1958), their biological constraints regarding saltwater are unknown. Rising sea levels have the potential to create shifts in the salinity of ponds within coastal systems. Agha et al. (2018) suggest that coastal freshwater turtle species, particularly those from the families Chelidae, Emydidae, and Trionychidae, could lose more than 10% of their current range because of rising sea levels and saltwater inundation. Future research should examine the biological constraints of chicken turtles in relation to saltwater experimentally, providing more specific details on the physiological limitations of the effect of saltwater inundation in the freshwater habitats of this species.

Chicken turtles are a chelonian species with a host of unique life history characteristics, which translates into microhabitat preferences that are characteristically unique as well. We determined that when a suite of environmental variation is available for chicken turtles to occupy, they prefer ponds that are shallower (maximum pond depths of less than 150 cm), have an open canopy (less than 51% canopy cover), and are not inundated with saltwater (salinity levels around 0.1). Because of their atypical ecology for a freshwater turtle species, it is imperative to continue to monitor and establish baseline ecological characteristics to aid in their conservation. Understanding the microhabitat preferences of chicken turtles is critical to future management decisions and habitat conservation pertaining to this species, and especially populations that exist in barrier island ecosystems and coastal habitats, as these populations are at a high risk of population declines due to many factors such as rising sea levels, habitat fragmentation, climate change, habitat disruption due to hurricane overwash, and saltwater inundation.