Click here for group photo

2019 Box Turtle Conservation Workshop
May 17-19, 2019
Summit Environmental Education and Conference Center
Haw River State Park, North Carolina

Long-term Studies on Florida Box Turtles: Lessons from an Old Turtle

C. Kenneth Dodd, Jr.

Florida Museum of Natural History, University of Florida, Gainesville, FL

From 1991-2006, I conducted a population study of Florida Box Turtle (Terrapene bauri) on Egmont Key off the west coast of Florida at the mouth of Tampa Bay. Although often considered “common,” little was known about box turtles in Florida prior to 1990. In this talk, I summarize information on the population, much of which is now currently available in the published scientific literature. As one might imagine, spending a long time in the field and analyzing a variety of natural history and demographic information has resulted in much reflection about the history of the island and its turtles, the interpretation of results and the extent of their inference to other populations, and to the foibles of dealing with various land management agencies through time. It is these reflections I offer as my lessons from an old turtle: Long-term studies are absolutely essential to understand all aspects of the biology of any species, especially those with long lifespans; Field studies are site specific in time – they offer only a snapshot in the long evolution of any taxon or community; Populations and habitats are not static – everything changes; Know the history of the land you are working on; Your studies may lead to unexpected results – this is not a bad thing!;Your studies should lead to more questions; If you think you know all the answers, you need to go back into the field; Just because someone works for a land management agency does not mean they are interested in conservation beyond the requirements of their position. Beware the “agency” mentality – trust actions, not words; No habitat or species is truly legislatively protected in perpetuity; Habitats are subject to natural and human-caused disturbances – manage and conserve for potential, not the status quo; “Common” species are important and they may not be common; Don’t be adverse to studying natural history; Beware of reliance on models rather than empirical data – publishing in “high impact” journals may not lead to useful results in the long run; Spend as much time as you can outdoors; Enjoy yourself!


Consistency in Box Turtle Behavioral Variation Across Developmental Stages and Populations

Bradley E. Carlson1, Sasha Tetzlaff2, Jennifer Mook33, Natalie Hyslop3, Sarah Carter4, Austin Hulbert4, Jeanine Refsnider4, and John H. Roe5

1Wabash College, Crawfordsville, IN, 2University of Illinois Urbana-Champaign, Urbana, IL, 3University of North Georgia, Gainesville, GA, 4University of Toledo, Toledo, OH, 5University of North Carolina at Pembroke, Pembroke, NC

Populations of wild animals consist of individuals that vary consistently in their behavioral tendencies (“personality”). This raises questions about whether the environment drives development of behavioral variation. Moreover, it may benefit conservation efforts to understand the origins of behaviors that affect fitness and ecological interactions. Eastern Box Turtles (Terrapene c. carolina) exhibit individually distinct levels of boldness (indicated by the duration of hiding in the shell after standardized handling and confinement). These differences in boldness have been associated with components of survival including predation risk and thermoregulation. To evaluate causes of this variation we first compared the boldness of captive juvenile turtles raised from hatching (n = 32 turtles from 11 clutches). Individuals raised identically still varied in behavior, and there was no overall difference in boldness between turtles from contrasting rearing conditions (simplified laboratory housing versus enriched environments). This suggests that variation in boldness appears early in development, regardless of the environment. Second, we compared the boldness of 237 wild adult turtles from nine populations across five states from September 2017 to October 2018: one population in Georgia (n = 24 turtles), two populations in Indiana (n = 53), one population in Michigan (n = 63), two populations in North Carolina (n = 38), and three populations in Ohio (n = 59). Despite diverse environmental conditions, the distribution of boldness scores was strikingly similar across populations (average emergence times for each population = 141 ± 39is [mean ± 1 SD]), indicating negligible behavioral population differentiation. Further, there was no evidence that temperature, individual human encounter history, sex, age, or body size affected boldness scores. Together, these results suggest that a wide range of boldness phenotypes is intrinsic to box turtle populations. This variation could be adaptive and thus considered an important part of population health and/or it may be unavoidable due to developmental processes.


Applying Dendrochronology Techniques to Develop Long-Term Growth Chronologies in the Eastern Box Turtle

Ashley E. Graham and Kiyoshi Sasaki

Department of Biology, Winthrop University,,,

Climate change is a widely hypothesized cause for decline in wildlife species. However, our understanding of climate change impacts is limited by a lack of long-term data to establish a baseline from which to assess biological change. Monitoring individuals and populations for decades to yield such data is often not feasible. Instead, growth rings recorded in hard parts of organisms serve as biological archives and can provide such data quickly. Cross-dating is a technique used in tree-ring analysis that accounts for false and missing annual rings and allows each annual growth ring to be assigned correctly to its year of formation. Resulting is a long-term growth chronology with the annual resolution necessary for climate-growth analyses. We are investigating whether cross-dating can be adapted to Woodland Box Turtles (Terrapene carolina) in the South Carolina Piedmont. Like trees, terrestrial turtles usually develop a growth ring each year, but may also form more than one ring or no rings during some years. For cross-dating to be valid, the following assumptions must hold true: (1) growth rings are laid down annually, and (2) common environmental conditions induce synchronous growth patterns among all or most individuals. We examined macrophotographs of Woodland Box Turtle scutes taken in 2018. Assuming that the assumptions for cross-dating hold true for box turtles, we identified and assigned the calendar year of formation to the growth rings. By monitoring radio-tracked box turtles over the next two years, we will validate the assumptions of cross-dating for Woodland Box Turtles. This study will demonstrate whether cross-dating can be used to obtain decadal growth data in the Woodland Box Turtle. If successful, growth chronologies will be used to examine their growth response to the climate and weather variabilities.


Box Turtle Commerce: Understanding the Domestic and International Trade in the Genus Terrapene

Jordan Gray

Turtle Survival Alliance, Charleston, South Carolina, 29406 USA,

The globalization of the trade in wildlife has increased species demand and quantity, adding to the multitude of anthropogenic pressures already impacting wild populations. Although protected by local, regional, national, and international legislature, agreements, and treaties, a dynamic global marketplace utilizing new and evolving market technologies continues to circumvent these conservation measures. Among the herpetofauna traded in this marketplace, Testudines represent the most commonly kept Order. Of the 469 terminal taxa of chelonian, representing 94 genera, the North American box turtle complex, genus Terrapene, remains a commonly kept turtle in the United States and internationally. The trade in all 11 terminal member taxa of this genus is governed by the Convention on International Trade in Endangered Species of Wild Fauna and Flora (CITES), protected from commercial collection and trade in Mexico and Canada, and protected from commercial collection in 37 of the 39 states in which they reside in the United States of America. Despite this, a domestic and international market place both exists and continues to find new growth for all 11 taxa. This dynamic and growing trade in the member taxa of the genus poses an additional extirpation threat to local and regional populations, places an increased burden on wildlife agencies and their cooperating organizations and private contractors, and undermines protective environmental legislation. New stop-gap and long-term solutions must be devised, coordinated, and implemented to ensure the survival of these taxa at the local, regional, and national population levels.


A Reassessment of the Florida Box Turtles on Egmont Key

Michael T. Jones1,2,3, Lisabeth L. Willey1,2, Jonathan D. Mays1,4,  and C. Kenneth Dodd, Jr.5 

1American Turtle Observatory, New Salem, MA, 2 Department of Environmental Studies, Antioch University New England, Keene, NH, 3 University of Massachusetts, Amherst, MA, 4 Florida Fish and Wildlife Conservation Commission, Gainesville, FL, 5 Department of Wildlife Ecology and Conservation, University of Florida, Gainesville, FL

The Florida Box Turtle (Terrapene bauri) population on Egmont Key, Hillsborough County, Florida, was the subject of long-term ecological studies from 1991–2006 by C. K. Dodd, Jr. and colleagues. Dodd et al. documented T. bauris population size and structure, reproductive parameters, demographic trends, habitat use, and activity patterns. Among North American box turtle populations, the Egmont Key population was unusually dense, demographically robust, and resilient, with an increasing population trajectory. The Egmont Key box turtle population was estimated to be 780 turtles in 1991 and approximately 1480 turtles in 2006 (the last year the population was assessed prior to this study). The Egmont Key population differed from many other well-studied populations, which tend to demonstrate population decline over time. In 2017–2018, we conducted three extended surveys of the island to reassess the condition of the box turtle population. In March 2017, November-December 2017, and March 2018, we detected 348 box turtles, of which 33 were alive and 315 were dead. The 33 live box turtles were captured 37 times, and included only four turtles marked previously by Dodd et al. (12.1%). Of the 315 dead box turtles, 83 (26.4%) had been marked by Dodd between 1991–2006, 119 (37.8%) had not been previously notched, and 113 (35.9%) were too badly damaged or deteriorated to ascertain whether they had been previously notched. To evaluate the effects of a 26.1-ha wildfire that occurred in July 2016, we implemented standardized sampling protocols including point-centered 1/4 hectare plots, a cross-island transect, and GPS-referenced meandering visual encounter surveys. Randomized, time-constrained surveys within the 2016 wildfire area revealed an average of 38.5 dead box turtles per hectare, from which we extrapolate approximately 1,004.9 (95% CI=786–1223) observable, dead box turtles across the entire extent of the 2016 wildfire. Clearly, not all of these died in the wildfire, but we note that the extrapolated number of dead turtles within the 2016 wildfire area represents nearly 68% of the last population estimate in 2006. A randomized, east-west transect across the extent of the 2016 wildfire produced comparable results. Although there are many sources of uncertainty, the 2016 wildfire apparently resulted in a significant mortality event. Of 259 dead box turtles found during this targeted wild fire assessment, 148 were too badly burned or deteriorated to estimate the likely cause of death, but at least 65 were judged to have died coincident with the 2016 wildfire. Another 43 appeared to have been killed by a mammalian or avian predator. Simple log linear estimators indicate a current (March 2017–March 2018) population size of about 238 animals, but we note that low detection rates and elevated rates of recent mortality indicate that we may violate key assumptions of closed-population modeling. Our preliminary assessment indicates that a once-robust and regionally-significant box turtle population has been substantially reduced in numbers, and that mortality rates from chronic and stochastic sources are higher than they were during the initial years of study. 


Terrapene coahuila: Benefiting an Endangered Species Through Captive Success 

Chris Leone

Owner, Garden State Tortoise, Galloway, NJ, Director of Animal Husbandry, theTurtleRoom, P.O. Box 521, Litiz, PA 17543,

The Coahuilan or Aquatic Box Turtle (Terrapene coahuila) remains a highly endangered chelonian with a severely limited natural range. Little is understood about its ecology and breeding statistics. As an advocate for this species under responsible captive conditions, I have been successful in breeding them for a number of years. I have compiled data concerning reproduction tips and gender outcome, activity patterns, diet preferences and seasonal behavioral changes. In addition to this I am an active member in the AZA Species Survival Plan Program for T. coahuila and a source of information for the pilot breeding program in Mexico. At my facility, several Terrapene species that are kept and bred come from government confiscations, zoos, and the private sector. By working hands on with T. coahuila and other box turtle species on a day to day basis, the information I have compiled may be beneficial to researchers in the field. Some specifics may be nearly impossible to detect in a wild setting without the capability of being around the turtles constantly. T. coahuila has proven to be a rather prolific species in captivity despite its grim status in nature, for example. The social dynamic of the animals appears to have a heavy influence on the overall success a potential breeding colony may or may not yield, and conspecific compatibility may play a factor in this. Through trial and error and participating in the species survival transfer plan, I have witnessed both declines and increases in the viability of eggs along with male breeding engagements. Additional particulars including how the turtles deal with water levels and quality, preferences within a captive habitat and comparability to other Terrapene behaviors have all been recorded. With T. coahuila being a turtle so subjected to isolation and impending habitat loss, it is perhaps highly serviceable for captive experiences to be shared on a more prominent level. In the end, these findings and observations at such a consistent degree may be instrumental in both conservation and preservation outcomes for the species. 


A Significant Eastern Box Turtle Population Near the Northern Edge of the Species Range in Massachusetts

Julie Lisk and Bryan Windmiller

Zoo New England’s Grassroots Wildlife Conservation, Concord, MA,, 978-844-4471,

The northernmost range of the Eastern Box Turtle (Terrapene carolina carolina) is in New Hampshire, just north of the Massachusetts boundary. Within Massachusetts, Eastern Box Turtle populations are largely concentrated in the southeast and Connecticut River Valley. Although there have been scattered reports of individual Eastern Box Turtles in northeastern Massachusetts, sustainable populations were thought to be potentially absent from this area. However, in September 2016, we observed a male Eastern Box Turtle crossing a road in northwestern Middlesex County, and fitted him with a radio-tracker. We actively tracked him through the remainder of the season to see if he was part of an active, robust local population. In 2017 and 2018, we continued our weekly observations and conducted surveys within areas of suitable habitat nearby. We discovered 18 additional individuals (15 adults and three juveniles). We also located and protected box turtle nests (four in 2017, and seven in 2018). Our findings have extended the known range of an apparently sustainable Eastern Box Turtle population into northern Middlesex County. Currently, there are no regulated populations in northern Middlesex County, and sightings of single individuals of this state-listed species do not trigger an environmental review under the Massachusetts Endangered Species Act. To effectively manage and protect populations at their northern range limit, we must determine the full extent of populations and their associated habitat requirements. We are currently radio-tracking 15 adult Eastern Box Turtles, and will increase our survey effort in 2019. We will present preliminary data on home ranges, nesting behavior, habitat associations, and the co-occurrence of males and females of Eastern Box Turtles close to the northernmost edge of their published range.


The Maryland Zoo Box Turtle Monitoring and Citizen Science Program: Moving 20 Years of Research to the Next Level

Katharine Mantzouris

The Maryland Zoo in Baltimore,, (910) 624-4728

The Maryland Zoo has been monitoring and collecting data on free-ranging Eastern Box Turtles (Terrapene carolina carolina) on zoo campus for over 20 years. What started as a small, outside project led by a university researcher, has grown into an all-inclusive, zoo-wide monitoring effort. Recently this program has undergone a revitalization effort in order to be brought up to speed with today’s technology and employ the latest scientific techniques of studying turtles while expanding our reach across the zoo campus. An updated data collection sheet was initiated to include more relevant data and became a more functional data collection tool. Informational posters were designed and put into communal staff areas to serve as a reminder to call when a box turtle is found. A protocol was written to train staff on standardized handling of turtles and data collection and to mitigate the risk of cross-contamination with the zoo’s collection animals.  These efforts have resulted in an increase in box turtle sightings and have drummed up appreciation and awareness of box turtles from our staff and volunteers. In 2018, we pilot tested our next step in the program which is tracking turtles with radio transmitters. We attached two turtles with transmitters and we look forward to this next step in this program which includes using this technology as a teaching tool for staff and youth volunteers. We will also be using it as a new means to research the behavior, movement, and habitat use of the zoo’s wild box turtle population.  


Confirmation of Successful Brumation of Eastern Box Turtles (Terrapene carolina carolina): A Comparison of Resident Versus Translocated Through Incidental Encounters at the Maryland Zoo In Baltimore

Katharine Mantzouris

The Maryland Zoo in Baltimore,, (910) 624-4728

The Maryland Zoo in Baltimore has been monitoring the on-site, free-ranging Eastern Box Turtle (Terrapene carolina carolina) population for more than 20 years. While in the process of revitalizing this long-term monitoring project, we re-examined our paper and digital records, and it was noted that 38 turtles in the monitoring program had been translocated from another area in the state. These animals were released, free-ranging, onto suitable habitat within the zoo’s property. Translocating turtles has been shown to have negative effects on a turtle’s survivability as they tend to cross more hazardous situations such as roads and expend more energy attempting to return to their original territory or trying to find a new home range. Based on the historical data available, we sought to compare the survivability of the translocated turtles compared to the resident turtles in the program by comparing how many turtles in each group (resident and translocated) were confirmed to have survived at least one overwintering period. We predicted that certain features of the Maryland Zoo such as perimeter fencing and minimal vehicular traffic would increase the survival of the translocated turtles so that it is comparable to the resident turtles in the same area. It was found that there was no significant difference in two groups using this criterion. These findings were encouraging and likely due to the mitigation of common box turtle hazards such as vehicular strikes, predators, and habitat threats.


Introgression and Selection in Terrapene From the Southeastern United States

Bradley T. Martin1, Marlis R. Douglas1, Tyler K. Chafin1, John S. Placyk Jr. 2, Roger D. Birkhead3, and Michael E. Douglas1

1Department of Biological Sciences, University of Arkansas, Fayetteville, AR 72701,, Phone: (479) 502-3613, 2Department of Biology, The University of Texas at Tyler, Tyler, TX 75799, 3Alabama Science in Motion, Auburn University, Auburn, AL 36849

Hybridization can trigger a variety of phenomena, some with evolutionary ramifications (e.g., adaptive potential), while others may lead to conservation-related issues (e.g., outbreeding depression). For example, hybrid sterility wastes reproductive effort and can contribute to population declines. On the contrary, hybrids that can produce viable offspring (i.e., introgressive hybridization) represent adaptive potential and are subject to environmental and evolutionary pressures (i.e., selection). Importantly, introgression can occur differentially across the genome, with some loci experiencing elevated levels of admixture and others being resistant to it. Until recently empirical and statistical identification of genome-wide introgression in natural populations has been limited, but advances in sequencing technologies now make such studies feasible. Hybridization has been described in Terrapene from the southeastern United States, but not yet fully explored at the genome-wide scale. To assess introgression and selection in southeastern Terrapene taxa (T. carolina carolina, T. c. major, and T. mexicana triunguis) we sequenced variation across thousands of genome-wide SNPs (Single Nucleotide Polymorphisms) via reduced-representation Next-Generation-Sequencing techniques. Admixture and population structure were assessed for pure and putatively hybrid individuals (N=394) using the software ADMIXTURE, and individuals with mixed ancestry were statistically evaluated using HyDe (Hybrid Detection). We then used NewHybrids to classify hybrids into parental, F1, F2, backcross, and backcross2 generations. Finally, the SNP data were aligned to the Terrapene m. triunguis transcriptome, and genomic clines (i.e., locus-specific ancestry compared against the genome-wide average) were generated between each pair-wise comparison of taxa to identify loci putatively under selection or experiencing hybrid vigor/dysfunction. From four discrete populations (K=4) that included T. c. carolina, T. m. triunguis, and two distinct T. c. major populations from Florida and Mississippi, we identified 157 individuals that were statistically corroborated with HyDe (P<0.05) as admixed. Almost all introgressed individuals were assigned by NewHybrids to the F2 and backcross generations, and admixture patterns generally reflected spatial clines, possibly indicating exogenous selection for introgressed genotypes in intermediate habitats. Ten unique loci under selection (P<0.001) aligned to the transcriptome, with the majority potentially associated with temperature tolerance, anoxia, and temperature-dependent sex determination. These data shed light on both gene flow and reproductive isolation across the genomes of three of the four Terrapene taxa inhabiting the southeastern United States. Furthermore, having a better understanding of the dynamics of hybridization and introgression occurring in the region will help facilitate effective conservation management strategies.


Autecological Study of Gulf Coast Box Turtles (Terrapene carolina major) in the Florida Panhandle Reveals Unique Spatial and Behavioral Characteristics

Jessica R. Meck*1,2, Michael T. Jones2, Lisabeth L. Willey1,2, and Jonathan D. Mays2,3

1Department of Environmental Studies, Antioch University New England, Keene, New Hampshire 03431, 2American Turtle Observatory, 90 Whitaker Road, New Salem, Massachusetts 01355, 3Florida Fish and Wildlife Conservation Commission, Gainesville, Florida 32601, *Corresponding author, e-mail:; phone: 814-515-7089

Terrapene carolina major (Gulf Coast Box Turtle), similar to other T. carolina lineages, is vulnerable to the consequences of the rapidly growing anthropogenic pressures on the landscape, but no intensive studies have been conducted or published to provide information on how this lineage compares to its conspecifics. Therefore, we conducted a radio-telemetry study on 21 adult T. c. major in the Florida panhandle in 2016 and 2017 to evaluate seasonal home range, habitat use and selection, and behavior. We calculated summer home range size using minimum convex polygons (MCP) and fixed kernel density estimators (KDE). Habitat use was evaluated using Geographic Information Systems (GIS) to determine proportions of different habitat types within individual’s 100% MCPs. Habitat selection was determined by generating 100 random points within each individual’s 100% MCPs to compare to the telemetry locations using logistic regression models. Contrary to most T. carolina studies, females had significantly larger home ranges than males. 2016 female home ranges averaged eight times larger than male’s, and 11 times larger than male’s in 2017. Males, however, demonstrated significantly higher home range fidelity than females and displayed frequent fighting behavior with other males both years, suggesting potential territoriality characteristics. Logistic regression models indicated males utilized forested wetlands significantly more than wet coniferous forests where as females utilized wet coniferous forests significantly more than forested wetlands. Approximately 25% of total radio-location observations for both sexes were aquatic environments. While the phylogeny of this lineage remains unclear, we recommend T. c. major be considered a unique, distinct conspecific of T. carolina when evaluating conservation decisions based on the ecological characteristics we observed.


Juvenile Ornate Box Turtles (Terrapene ornata ornata) Prefer Safety Over Thermoregulation.

Abigail A. Neyer1 and Gwendolyn C. Bachman2

1University of North Georgia,, 513-607-8758, 2University of Nebraska—Lincoln

Juvenile organisms must make decisions regarding their growth, development, and survival that can greatly affect their overall fitness. In ectothermic species, juveniles may be expected to seek habitats that allow them to maintain preferred body temperatures which could reflect temperatures that are optimal for foraging, growth, or digestion. However, the areas that allow juveniles to maintain preferred body temperatures may not provide protection from predators. Juvenile turtles could face high mortality due to their small body sizes and undeveloped shells that provide little protection; therefore, juvenile turtles may face trade-offs between maintaining preferred body temperatures and avoiding predation risks. Little is known about how the behavioral decisions that juvenile turtles make affect their growth and survival, and yet these decisions could have large impacts on their future reproduction. In this study, we first placed juvenile (< 1 year old) Ornate Box Turtles (Terrapene ornata ornata) in a thermal gradient to determine their preferred body temperature which could reflect temperatures that are optimal for physiological processes such as growth. We then tested juvenile turtles in an experiment that offered individuals an opportunity to either maintain a preferred body temperature or remain in protective cover that was kept at a temperature below preferred. Our data suggests that juvenile turtles selected areas that provide protective cover even though this resulted in the turtles having body temperatures almost 10°C below preferred. However, our data also suggests that larger juvenile turtles were more likely to leave the provided shelter and maintain higher body temperatures compared to smaller turtles. Turtles are long-lived animals and so promoting long-term survival may be more critical than maintaining a preferred body temperature during the juvenile life stage. However, the effects that this strategy has on their growth and future reproduction remain unclear. These results begin to address the decisions that a long-lived organism may make as a juvenile when faced with trade-offs between maximizing growth and maximizing survival.


Urbanization May Contribute to Box Turtle Decline

Jack Nguyen, Ashley E. Graham and Kiyoshi Sasaki

Department of Biology, Winthrop University,,,,

Urbanization is a major threat to biodiversity. However, the underlying mechanisms through which it affects the persistence of wildlife are not well understood. One possibility is that urbanization-induced abiotic and biotic changes reduce overall food availability. The Eastern Box Turtle (Terrapene carolina) is listed as vulnerable by the IUCN. To determine whether urbanization associated reductions in food availability contribute to the decline of the Eastern Box Turtle, we measured the abundance and body condition of turtles along a gradient of impervious surface cover across 11 sites in the rapidly urbanizing South Carolina Piedmont region. Timed visual searches were performed at each site to locate turtles. The body mass, carapace width and length as well as the sex was recorded for each individual. Body condition was evaluated based on scaled mass index. Data from the first field season showed that both male and female abundance declined significantly as impervious surface cover increased. Urban males displayed significantly worse body condition; however urban females exhibited comparable body condition to rural females. The apparent lack of body condition decline in urban females suggests that resource availability is not a direct cause of observed population decline. The poorer body condition of males may be attributable to the increased energetic costs of searching for scare females in urban sites. Therefore, we propose that the observed population decline is not driven by reduced habitat quality per se, but primarily by external sources of mortality such as vehicle strikes, urban predation, or poaching. This upcoming summer we will collect additional data to evaluate this hypothesis.

Cause-specific Mortality in Urban and Rural Box Turtles (Terrapene carolina triunguis) in Missouri

Jamie L. Palmer, MS

Saint Louis Zoo Institute for Conservation Medicine,

Anthropogenic disturbances such as pollution, habitat fragmentation, vehicle/lawnmower strikes, and collection for the pet trade threaten the long-term survival of land turtles, including North American box turtles.  Many of these impacts are amplified in urban areas, and may contribute to population decreases for native turtle species.  Although few studies have been conducted on box turtle ecology and health in Missouri, these and other anthropogenic disturbances are believed to threaten the survival of the two resident box turtle species: the Three-toed Box Turtle (Terrapene carolina triunguis) and the Ornate Box Turtle (Terrapene ornata ornata).

We study two populations of Three-toed Box Turtles: one in Forest Park (FP), an urban park in the city of St. Louis, and the other at a biological field station, Tyson Research Center (TRC).  These two geographically separate populations have provided a means for us to compare turtles in a human-dominated landscape to those in an area of limited human influence.  Within this project, beginning in spring 2012, we have documented cause-specific mortality of VHF radio-tagged box turtles at FP (n= 25) and TRC (n=15). We hypothesized higher mortality at FP, the urban site, given its mosaic of land use with small, fragmented patches of suitable habitat and increased human presence and anthropogenic disturbance.

Over an eight-year period, overall mortality rate in VHF (very high frequency) tagged urban box turtles was 61% (14 out of 23), consisting of seven males and seven females.  During this same period, the mortality rate for the 15 turtles with VHF tags at the rural site was 27% (4 out of 15), with three females and one male. The difference in mortality rate is statistically significant between sites (p= 0.0133), but no significant difference was found between sex at either site or collectively. We divided cause of mortality into five categories: predation, “winter kill,” lawnmower trauma, trauma other than lawnmower, and unknown.  In this study, “winter kill” turtles reported here are those found dead on the surface either during the months of brumation or directly after brumation. The leading cause of death at FP was “winter kill” (43%). This may be a reflection of poor brumation site availability or site disturbance and anthropogenic impacts that impact box turtle winter survival. This supports our hypothesis that box turtles face a higher threat of mortality in an urban setting, and we hope these data may help with management recommendations that may help ensure box turtle survival in an urban park.  


Incorporating Biotic Interactions into Box Turtle Density Estimates on Long Island, NY 

Lisa Prowant

Department of Ecology and Evolution, Stony Brook University,,  (913) 909-6468

Knowing species density and occupancy patterns is crucial for effective conservation planning. Biotic interaction can alter these patterns yet are rarely included in occupancy and distribution models. One reason for this is that most models are unable to differentiate biotic interactions from unmeasured environmental variables. Additionally, distribution models must assume that biotic interactions are constant in space and time.  To overcome these challenges, we need fine-grained data on biotic interactions along environmental gradients over large spatial extents. The aim of this research is to explore approaches for incorporating species interactions into conservation plans for the Eastern Box Turtle in New York. I measured box turtle densities using systematic surveys at 30 locations across Long Island, NY in May-July 2016 and 2017. I estimated box turtle predator densities at each of these locations using wildlife cameras. I incorporated these data, along with information on human landscape impacts, landcover and land use, into Bayesian n-mixture models. These models allowed me to estimate true turtle densities while accounting for imperfect and variable detection probabilities. This research will help to determine the status and possible causes of decline of box turtles on Long Island. Understanding where box turtles are able to maintain high densities, even in a highly disturbed and fragmented landscape will inform future conservation efforts for this species. 


Initial Assessment of Terrapene carolina Status and Trends in North Carolina Using Ten Years of Data From The Box Turtle Connection

John H. Roe1, Ann B. Somers2, Ashley Lavere3, and Gabrielle Graeter4

1University of North Carolina Pembroke, 2University of North Carolina Greensboro, 3University of Georgia, 4North Carolina Wildlife Resources Commission

Turtles are among the most imperiled vertebrate groups worldwide, with nearly half of all species listed as vulnerable to extinction.  The Eastern Box Turtle (Terrapene carolina) is a widespread and once common species throughout the eastern United States, but a number of localized long-term studies have identified population declines over several decades.  Now more than half of states in which box turtles occur list them as a species of conservation priority.  Using data collected as part of The Box Turtle Connection (BTC), we analyzed population trends, vital rates, and demographics at 39 protected sites across the Coastal Plain, Piedmont, and Blue Ridge physiographic regions of North Carolina over a 10-year period.  The BTC has recorded 3,761 captures of 3,082 individual turtles through 2017.  Data were sufficient to estimate annual capture and survivorship probabilities stratified by demographic groups (juvenile, adult male or female) for 21 populations.  Population densities ranged from 0.15 – 6.0 turtles/ha, and population sizes were generally stable over time within sites.  Annual capture probabilities ranged from 0.01 – 0.23 among sites, and annual survivorship (adjusted for temporary emigration) was 0.86, 0.90, and 0.95 in Piedmont, Blue Ridge, and Coastal Plain populations.  Population sex ratios were biased towards males at many sites, and skewed towards adults.  Surrounding land cover and use was not a good predictor of any examined population metric.  These data will provide useful benchmarks for comparison to other sites across the range of T. carolina and will allow comparisons to the same site over time as the BTC continues.  Land managers and policy makers can use these data to update the status of T. carolina and to make localized decisions on conservation and management to maintain stable or growing turtle populations.


The Box Turtle Connection Creates Leaders in Citizen Science

Sara Steffen1, Ashley Lavere2, and Ann Somers3

1Piedmont Wildlife Center,2University of Georgia, 3University of North Carolina Greensboro,

The Box Turtle Connection, a 100-year citizen science study of Eastern Box Turtles (Terrapene carolina carolina), was created to address the need for long-term data on Eastern Box Turtle populations in North Carolina. Box turtles have been identified as a species of greatest conservation need by the state. Collaborations formed among educators, state agencies, and environmental organizations to establish over 30 project sites across North Carolina where volunteer Project Leaders (PLs) collect long-term box turtle morphometric and health data. The project hosts one of the largest databases known for box turtles of any species, with sites covering all physiographic provinces in the state. The project was designed to run on a shoestring budget and has earned the NC Wildlife Diversity Program over 1 million dollars in matching funds for state wildlife grants, while requiring minimal effort on the part of the leaders.

PLs include private landowners, state employees, educators, and others, each making a long-term commitment to the project and work at a single site collecting data and entering it into the statewide database. They agree to attend training on how to handle box turtles, collect and input data, and promote citizen science by educating and involving community members. Biennial meetings are held to train new leaders and to keep trained PLs excited and engaged, updated on the status of the project, and to keep their skills fresh. In an effort to preserve the quality and standardization of data across multiple sites, a handbook was created and has since been revised, as an easy reference for protocols in project management, data collection, and community engagement. Now being translated into Spanish, this book will assist Mexican colleagues in standardizing data collection and training new scientists to learn more about the disappearing box turtles in their region, while engaging Mexican citizens in turtle conservation.

As the project grew, some PLs became engaged at a deeper level, extending the project to a wider audience. Project leaders are collaborating with outside organizations, such as WildTrack, developing educational programs and activities that captivate and educate the public, and building relationships with community members, such as John Rucker and his “turtle dogs”, to further foster box turtle research. Through the efforts of our PLs and community members, we have established and continue to add to a dataset with over 4400 entries that can be used to determine trends in populations. These data can then be used to inform and persuade legislation into developing management practices that will offer protection to this and other species. Arguably more important, we educate citizens about box turtle conservation at public events, school programs, and via digital and written materials. Our project is a great example for other state institutions or organizations that want to collect regional scientific data for wildlife conservation, with limited resources.


Perspectives From Out West on Variability of the Life and Times of Terrapene ornata.

Marty Tuegel

Independent Researcher, Albuquerque, NM,, 520-444-3330

Terrapene ornata ranges from the Mississippi River in the east, west to southcentral Arizona; and into Chihuahua and Sonora, Mexico north to Wisconsin.   In 1996, I started a field research project on what was known at the time as the Western or Desert Box Turtle (Terrapene ornata luteola) in the San Pedro River Valley of Arizona.  The study was of a population that occupied Sonoran semi-desert grassland with mesquite intrusions at the oak encinal ecotone, on the east slopes of the Huachuca Mountains. This project lasted for 13 consecutive field seasons, using primarily broad area searches and radio telemetry.  During the study 85 individuals were captured and marked, including 13 juveniles / subadults, with a total of 1306 recaptures.  The active season was timed with the onset of summer rains, resulting in a prolonged period of inactivity from the beginning of the winter through the hot and drier part of the year.  A total of 17 mortalities were recorded, but only one from a motor vehicle.   Most occurred after particularly dry winters and multiple years of drought.  Home ranges were variable, but generally larger than found in two other populations in Arizona and New Mexico.  Movements towards the periphery of home ranges were observed to vary between the sexes and predictably for individuals.  These life history characteristics are in response to the same challenges as seen for this species on the west end of its range, but this site’s history is unique when compared to most grassland in Arizona or New Mexico, providing a differing view of what is typical.


Current Status, Recategorization and Conservation Efforts of the Coahuilan Box Turtle (Terrapene coahuila)

Sara Valenzuela-Ceballos1, Gamaliel Castañeda-Gaytán, Ernesto Becerra-López, Nohemi Villa-Vázquez, Bruno Rodríguez-López, Miguel Borja-Jiménez, and Craig Stanford.

1Biology Department, University of Juarez in the State of Durango, Mexico

The Coahuilan Box Turtle (Terrapene coahuila) is endemic to the valley of Cuatro Ciénegas in northern Mexico. Due to its habits that resemble those of the more aquatic emydids, the species may be considered an aquatic species, the only one of the genus. Recent evaluation of the species’ status in the wild led to a proposal to change its categorization under Mexican law from “threatened” to “in risk of extinction.” The status assessment for T. coahuila was made using the four criteria of Mexico’s Risk Assessment Method (MER, Spanish acronym). They include: a) Range of the taxon’s distribution in Mexico; b) Habitat status with respect to the how well it supports the taxon; c) Intrinsic biological vulnerability of the taxon and; d) Impact of human activity on the taxon. T. coahuila has experienced a reduction in range with just 5.30 km2 of suitable box turtle habitat (estimated using bioclimatic factors, presence of water bodies of appropriate depth) remaining within the valley. This represents only 0.6% of the Natural Reserve of Cuatro Ciénegas. The habitat has been lost in the past few decades due to massive groundwater extraction for agriculture, tourism, and irrigation infrastructure. The photographic record shows a decrease in size and desiccation of the “pozas,” where box turtles live, with practically all ponds undergoing modification to some degree. Some of the ponds have completely disappeared, which has impacted in the entire biota, including the Coahuilan Box Turtle. This species has shown a severe population decline over the last 40 years. In 1974 Brown estimated a population density of 148.2 ind/ha (14,820 ind/km2), meanwhile Castañeda et al. (2011, 2012, 2013, 2014, 2015) reported a density of 3.1 ind /ha (3,100 ind/km2). The current population structure shows a male bias in the wild with a mean of 61% males. Also, just 20% of the population is immature. This percentage could be indicative of low recruitment. The change in the current legal status would bring stiffer penalties for illegal trade. The recategorization could also bring about a major change in attention paid to the species’ vulnerability and greater economic support from the government. The recategorization has been approved by the Mexican government and it will be published in the next revision of the country’s imperiled species list (NOM-059). However, a mere change in the turtle’s classification will not be enough to ensure the recovery of the population, especially when the intrinsic biology of the species (long maturation period, high egg and juvenile mortality, and low recruitment into the adult breeding population) increases the difficulty of creating a successful conservation program. To build up the population, approval was gained last year in Cuatro Ciénegas to establish a captive head-start program for the Coahuilan Box Turtle. The first step (construction of the infrastructure) was completed last December. The current effort for a population recovery program now includes several organizations, local people, and academics to build a legal and strong cooperative program for long-term success.


A Review of Thread-Trailing Devices for Eastern Box Turtles (Terrapene carolina carolina), and a Comparison to Other Tracking Methods

Jessica M. Vannatta* and Matthew Klukowski

Department of Biology, Middle Tennessee State University, Box 60, Murfreesboro, TN USA

*Corresponding Author: School of Environmental Studies, Tennessee Technological University, Box 5152, Cookeville, TN USA,

Thread-trailing is a tracking technique used to monitor precise turtle movements, but there are positive and negative aspects to using this method. Using thread-trailers is inexpensive but can be a labor-intensive process, and there are often issues with the apparatus itself. We employed thread-trailing devices to track Eastern Box Turtles (Terrapene carolina carolina) in a suburban wetland habitat in middle Tennessee, USA. Unfortunately, we had limited success that was likely due to insufficient attachment and wet climatic conditions at our field site. However, many other researchers have used thread-trailing devices successfully, and this method is often used in conjunction with other tracking methods, such as radio-telemetry or mark-recapture. We will discuss the pros and cons of thread-trailing based on our experiences with this method, compare thread-trailing to other common tracking methods, and make recommendations about the thread-trailing apparatus. This information will aid other researchers in determining if the thread-trailing technique is appropriate for monitoring turtles in their study, and if so, how to best construct the thread-trailing device.


Ecology of the Yucatán Box Turtle (Terrapene yucatana) in Yucatán

 Lisabeth L. Willey1,2*, Michael T. Jones1,3, Eddie G. Nahuat1, Thomas S.B. Akre1,4, Erika Gonzalez1,4, Rodrigo Macip Rios1,5, and Luis Diaz1,6

1American Turtle Observatory, 90 Whitaker Road, New Salem, MA, 2Antioch University New England, 40 Avon St., Keene, NH, 3University of Massachusetts, Amherst, MA, 4Smithsonian Conservation Biology Institute, Front Royal, VA, 5UNAM Morelia, Morelia, Michoacán, Mexico, 6RCARY, Merida, Yucatán, Mexico, *Supported by a Lucille F. Stickel Award

The Yucatán Box Turtle (Terrapene yucatana) is the southernmost representative of the North American box turtles and the subfamily Emydinae, and one of only three Emydine lineages to occur within the tropics. As such, T. yucatana is an enigmatic biogeographic outlier and a significant evolutionarily lineage. The Yucatán Box Turtle has been reported from the northern and central Yucatán Peninsula including most of Yucatán state and the western border of Quintana Roo, south to at least 18.5˚N in southeastern Campeche. Throughout the Yucatán Peninsula, T. yucatana is known primarily from forested and agricultural landscapes. We studied the demography, movements, seasonal ecology, and behavior of the Yucatán Box Turtle in Yucatán, Mexico from July 2014 to December 2018. Our study area encompassed 23 ha of tropical forest and thornscrub in a remote, largely forested area used primarily for cattle grazing. We evaluated home range size, interannual home range fidelity, and habitat use by using radio telemetry to study the movements of 10 females and 7 males for periods ranging from 3 to 5 years. We estimated annual home range size using the 100% minimum convex polygon (MCP) method. Annual home range size averaged 0.84 ha for females and 1.41 ha for males; males use significantly larger areas than females on an annual basis. Turtles were found predominately in upland forested habitats with occasional observations in clearings dominated by columnar cacti, thornscrub, and graminoids. During the entire study period, female turtles used a combined average of 2.39 ha and males used 2.72 ha. The long axis of the MCPs averaged 165 m for females and 194 m for males. The active season varied annually but usually extended from late spring to late autumn. Twenty-one instances of courtship or mating were observed from June to November, with more than half occurring in September or October. Gravid turtles were observed in July, but no nesting was observed. Fighting between males was observed on three occasions between August and November. Turtles were observed feeding on various fruits from June to October. Turtles were primarily dormant in roots, shallow burrows, rocky cavities, and dense vegetation during the winter months. Using loglinear models we estimated the population size within our study area to be 56.1 (49.0–60.3), equivalent to a density of 2.4 turtles/ha. None of the study animals died during the period of observation, although one turtle marked during the study was found dead. We encountered three other turtle species within our study area: Kinosternon scorpioides, Rhinoclemmys areolata, and Trachemys venusta. This study forms the basis of a longer-term study of the annual variation in movements and seasonal ecology, the response of individual turtles to landscape change, and the status of this unique species in tropical Yucatán. As the only fully tropical lineage of the subfamily Emydinae, the Yucatán Box Turtle is a significant evolutionarily lineage that would benefit from large-scale forest conservation programs on the Yucatán Peninsula. However, the species is clearly threatened by urbanization, energy development, deforestation, prescribed fire, climate change, and increasingly rampant collection.