Category: News

By Tyjaha Steele

Kiersten Nelson (left) and Stacey Lance (right) monitor and record data during the release of a head-started gopher frog metamorph into an artificial burrow. (Photo courtesy of DOE)

Across the southeastern United States, conservationists are working to restore longleaf pine ecosystems, landscapes once common across the region but now greatly reduced from their historic extent. Many plants and animals depend on these fire-maintained habitats, including the Carolina gopher frog, a species that spends much of its life hidden underground and is rarely seen outside a brief breeding season.

At the Savannah River Site (SRS) in South Carolina, researchers from the University of Georgia’s Savannah River Ecology Laboratory (SREL), the USDA Forest Service–Savannah River, and other conservation partners are working together to better understand and protect the species. Led by senior research scientist Stacey Lance, the effort combines habitat restoration, population monitoring, and conservation research to support one of South Carolina’s remaining Carolina gopher frog populations.

The need for that work has become increasingly urgent. Historically, Carolina gopher frogs occurred in three distinct metapopulations at the Savannah River Site and used at least 17 wetlands for breeding. Today, only one of those metapopulations remains, and breeding activity has been documented in just a handful of wetlands in recent years. Researchers have also observed declining genetic diversity within the remaining population, raising concerns about the species’ long-term persistence.

At the same time, restoration efforts are continuing to expand across the landscape. The USDA Forest Service–Savannah River has developed a comprehensive habitat restoration plan aimed at improving more than 3,000 acres of gopher frog habitat through prescribed fire, forest thinning, wetland restoration, and improvements to habitat connectivity.

Together, these efforts are designed to restore the open-canopy wetlands and longleaf pine habitats that gopher frogs depend upon. Yet one challenge remains: understanding how frogs use the landscape beyond the breeding season.

Finding What The Naked Eye Can’t See

DJ, a Belgian Malinois dog trained in scent tracking, sits patiently while waiting to begin his work. (Photo courtesy of Tyjaha Steele)

Carolina gopher frogs breed in isolated seasonal wetlands, but adults spend most of the year in surrounding upland habitats where they shelter in underground refuges and dense vegetation. Because they are so difficult to detect, researchers often know far more about where frogs breed than where they spend the majority of their lives.

To help address that challenge, Lance partnered with wildlife biologist Dr. Karen DeMatteo and her conservation detection dog, DJ. Unlike traditional surveys that rely on visual observations or breeding-season activity, DJ is trained to locate gopher frogs using scent. Working through upland habitats, he searches for frogs hidden beneath vegetation or occupying underground refuges that would otherwise be difficult, and sometimes impossible, for researchers to locate.

The partnership began in 2023, when DeMatteo and DJ traveled to the Savannah River Site to help researchers locate recently metamorphosed gopher frogs released as part of a headstarting program. Those early efforts helped expose DJ to the scent of the species under a variety of conditions and demonstrated that conservation detection dogs could play an important role in amphibian monitoring.

A follow-up effort expanded the work to adult frogs, focusing on upland habitats and winter refuges, one of the least understood portions of the species’ life cycle. Trained using positive reinforcement, DJ performs a passive alert when he detects the scent of a gopher frog, stopping and staring at DeMatteo until he receives his reward. For researchers, each alert provides valuable information about where frogs are living across the landscape.

A Common Goal

While DJ’s role often draws attention, the project represents something much larger than a single dog or survey effort. It reflects years of collaboration among researchers, wildlife specialists, conservation agencies, and land managers working toward the same goal: ensuring the long-term survival of the Carolina gopher frog.

Information gathered through detection dog surveys helps researchers better understand how far frogs travel from breeding wetlands, which habitat features they rely upon, and whether restored areas are being used. Those findings can then inform future management decisions, helping partners refine restoration efforts and prioritize areas most important to the species.

Each frog located provides another piece of a much larger puzzle. Together, those pieces help researchers evaluate habitat restoration, improve conservation planning, and build a clearer picture of what Carolina gopher frogs need to survive.

By combining ecological research, habitat restoration, and innovative monitoring tools, partners across the Savannah River Site are working to ensure that Carolina gopher frogs remain part of the longleaf pine ecosystem for generations to come.

A dark and empty road stretches ahead at night, illuminated by vehicle headlights (Photo courtesy of Carson Pakula).

Roadways are often studied from the perspective of wildlife, where animals cross, when they move, and how they respond to traffic. These patterns have helped identify high-risk areas and inform strategies to reduce collisions. But every encounter on the road involves two participants. While animal behavior has been widely examined, the role of the driver, how hazards are detected, interpreted, and responded to in real time, has received far less attention, despite being a critical factor in whether a collision occurs. 

Researchers from the University of Georgia’s Savannah River Ecology Laboratory examined this overlooked side of the interaction by focusing on how driver behavior influences animal-vehicle collisions. The study, led by Carson Pakula, who conducted the research as a doctoral student and is now a postdoctoral research associate, drew from 118 published studies to better understand how drivers notice, interpret, and respond to wildlife on roadways. 

Rather than looking only at where collisions happen, the team focused on what drivers are actually doing in the moments leading up to them. They described collisions as a sequence in which a driver must first notice the animal, recognize it as a risk, and then react in time to avoid hitting it. When any step in that process breaks down, a collision can still occur, even when both the driver and the animal are behaving as expected.

“Most of the research so far has focused on animal-based behaviors with studies identifying the decisions animals must make to avoid a collision,” says Pakula. “As drivers are involved in most wildlife-vehicle collisions, I was interested in adapting this approach to identify what factors influences specific driver behaviors during an interaction. “

Across the research, vehicle speed emerged as one of the most consistent factors influencing collision risk. Faster speeds reduce the time available for drivers to react and increase the distance needed to stop, and these effects become even more pronounced at night. Headlights illuminate only a limited distance ahead, which means drivers may be traveling faster than they can safely see, especially on unfamiliar roads. 

However, speed does not always affect collisions in the same way across different road types. High-speed roads may discourage animals from crossing, while moderate-speed roads can see more frequent crossings. These patterns reflect how closely driver behavior and animal behavior are linked, with each influencing the other. 

The study also examined how warning systems, including standard roadside signs and animal detection technology, affect driver behavior. Traditional wildlife crossing signs often have limited impact because drivers become used to seeing them without encountering animals. In contrast, signs that activate when animals are present show more promise, as they can increase driver awareness and encourage slower speeds, although their long-term effectiveness remains uncertain. 

Road conditions and surrounding environments also influence how well drivers can detect animals. Curves, dense roadside vegetation, and low-light conditions can reduce visibility, while open roads may encourage faster speeds or reduced attention. Animal characteristics further shape these interactions, as larger animals are generally easier to detect, whereas smaller or less visible species may go unnoticed until it is too late. 

Despite decades of research, few studies directly measure how drivers respond to animals in real-world conditions. Much of what is currently understood comes from indirect evidence rather than observations of driver decision-making as it occurs.

“One of the biggest challenges with studying driver reactions in real time is the logistical difficulty of observing driver behavior in truly realistic conditions. While driving simulators and decoy animals allow for researchers to test how well drivers can detect animals, these approaches lack real world complexities which can which may limit how well the findings translate to actual wildlife encounters,” explains Pakula. “Capturing realistic driver reactions require natural, unanticipated encounters with wildlife, which is time and resource intensive. We still know relatively little about how driver behave during wildlife-vehicle encounters, particularly how different headlights impact how well drivers see animals and what evasive maneuvers drivers take during an encounter with an animal.”

These findings suggest that reducing animal-vehicle collisions will require a better understanding of both driver and animal behavior. While new technologies may help improve detection and response, their effectiveness will depend on how well they align with the way drivers perceive and react to risk. 

Ultimately, reducing collisions will require more than a single solution, and progress will depend on combining insights from engineering, wildlife science, and human behavior. Each encounter on the road is shaped by both the driver and the animal, and understanding that interaction is the first step to improving safety. 

The full study, Evaluating causes of animal-vehicle collisions through the lens of driver behavior, was published in Accident Analysis and Prevention and authored by Carson J. Pakula, Olin E. Rhodes Jr., and Travis L. DeVault. 

Courtney Werner is seen banding and collecting blood samples from one of the birds captured throughout this study. (Photo courtesy of Travis DeVault)

Parasitic infections in wildlife species are influenced by more than just their exposure to parasites and their vectors in the wild. The environments animals live in, including the quality of their soil, water, and habitat, can affect whether infections occur and how they spread through populations. Chemical elements and contaminants present in the environment may subtly alter an animal’s condition, changing interactions among hosts, parasites, and the insects that transmit them. 

In an attempt to better understand these relationships, researchers working at the University of Georgia’s Savannah River Ecology Laboratory (SREL) studied parasite infections in wild birds at the Savannah River Site (SRS) in Aiken, South Carolina, where relatively undisturbed habitats and areas affected by past industrial activity occur side by side. This study was led by former master’s student Courtney Werner, who was co-advised by Olin E. Rhodes Jr., Director of SREL and UGA Athletic Association Professor of Applied Ecology in the Odum School of Ecology, and Travis DeVault, Associate Director for Research and Senior Research Scientist at SREL. The team focused on haemosporidian parasites, a group of avian blood parasites related to the organism that causes malaria, and examined whether environmental contamination influenced rates of infection. 

In natural systems, disease depends on more than exposure alone: birds serve as hosts, mosquitoes transfer infections between individuals, and parasites rely on both to complete their life cycle. Because contaminants can affect nutrition and immune defenses, the researchers examined whether exposure to contaminants was associated with differences in parasitic infection patterns, rather than only with direct harm to birds. 

Over the course of one breeding season, the team captured 329 birds representing 31 species across six wetland and streamside habitats on the SRS, and collected blood samples to measure contaminant exposure and test for parasite infections. Mosquitoes, who are vectors of parasite transmission, were also sampled to determine whether differences in bird infections were driven by changes in the parasite infection rates of the mosquito vectors  or by changes in immunity to the parasites within the birds themselves. 

The study compared relatively uncontaminated areas with locations influenced by legacy industrial activities, including coal-combustion waste and nuclear-related contamination, and researchers measured several trace elements and a radionuclide in the birds. These included zinc, copper, mercury, lead, arsenic, and selenium, and cesium-137. Among the contaminants examined, selenium showed the clearest relationship with infection patterns. 

“One of the clearest effects that we observed within birds was the relationship between selenium, a trace element commonly found in coal combustion waste, and a parasite that is commonly found in the blood of birds,” says Rhodes. “Birds with concentrations of selenium above a certain level, just did not have the parasite, despite the fact that many other birds in those same areas were infected.” 

Birds living in areas affected by coal-combustion waste had higher selenium concentrations in their blood, and those elevated levels were associated with fewer infections from one parasite group known as Leucocytozoon. Individuals with selenium levels above a certain threshold showed no infection by that parasite, although the same pattern was not observed for other parasites, such as avian malaria (Plasmodium) or Haemoproteus, suggesting that contaminants influenced specific host–parasite relationships rather than all infections equally. 

When researchers examined mosquitoes, however, infection rates did not differ between contaminated and reference sites. Instead, they followed seasonal patterns, increasing during the breeding season when dormant infections in birds can re-emerge and spread. These results indicated the contaminant was affecting the birds’ ability to resist infection rather than altering parasite transmission. 

“These results suggest that some environmental contaminants, such as selenium, may influence individual host immunity more than they disrupt broad parasite transmission cycles within vector populations,” explains Daniel Peach,  an assistant professor from SREL and the Department of Infectious Diseases in the College of Veterinary Medicine. 

Although fewer infections might appear positive, selenium exposure can also affect reproduction, making the overall effect more complicated. Birds can transfer contaminants into their eggs, and higher selenium concentrations may reduce hatching success even while certain parasite infections in adult birds decline, creating a tradeoff between disease resistance and reproductive health.

The findings suggest that environmental contamination can reshape disease dynamics by influencing interactions among hosts, parasites, and vectors, showing that wildlife health is closely linked to ecosystem conditions. Understanding these relationships helps scientists better predict how species respond to human-altered environments and reveals that pollution may affect wildlife not only through toxicity, but also through changes in infection dynamics. 

The full study, Use of Contaminated Habitat and Associated Selenium Uptake Mediate Haemosporidian Parasite Infections in Wild Passerine Birds, was published in Ecology and Evolution. Authors include Courtney S. Werner, Mary Chapman, Daniel A. H. Peach, Travis L. DeVault, and Olin E. Rhodes Jr. 

By Tyjaha Steele

Photo of a wild pig farrowing nest constructed of palmettos within a pine forest. (Photo courtesy of Jim Beasley)

For many animals, where offspring are born can influence their survival from the very first few moments of life. Shelter, nearby resources, and protection from disturbance all play a role in reproductive success. For invasive wild pigs, understanding where females choose to give birth, known as farrowing sites, can provide insight into their reproductive behavior and help inform more effective management strategies. 

To better understand how wild pigs select these nesting locations, researchers from the University of Georgia’s Savannah River Ecology Laboratory (SREL) and the Warnell School of Forestry and Natural Resources examined farrowing site selection at the Savannah River Site (SRS) in South Carolina. The work began as part of former graduate student Sarah Chinn’s, Ph.D. research, during which she led the field component of the study, and was later expanded upon by Travis Stoakley, a current SREL graduate student, who led the analysis and writing. The study focused on 24 mature female wild pigs monitored using GPS collars and internal transmitters that signaled when birth occurred.  

“Within the southern U.S. wild pigs give birth throughout the entire year, with many sows giving birth twice in a given year,” explains Beasley, a researcher and professor from the University of Georgia’s Savannah River Ecology Laboratory (SREL) and Warnell School of Forestry and Natural Resources. “This incredible reproductive capacity is one of the contributing factors to their ability to rapidly invade new areas and one of the biggest challenges to controlling their populations.” 

Former doctoral student Sarah Chinn taking measurements on a piglet captured at a farrowing nest (Photo courtesy of Jim Beasley)

Wild pigs are among the most widespread and costly invasive mammals in the United States, causing billions of dollars in damage each year. Their high reproductive capacity allows populations to grow rapidly, making it difficult for land managers to slow or reverse their spread. While previous research has clarified when wild pigs reproduce, far less is known about the environmental features that influence where females choose to farrow. 

Once researchers identified farrowing sites, they conducted detailed field surveys to document both fine-scale and broad-scale environmental characteristics. At each site, the team measured vegetation structure, light levels, temperature, canopy cover, and proximity to water. These features were then compared to nearby random locations to determine whether pigs were selecting specific conditions rather than using habitat at random.  

The results suggest that female wild pigs consistently selected farrowing sites with dense and diverse understory vegetation and close access to water. All observed farrowing sites were located within a short distance of a water source, such as streams, ditches, or ephemeral pools. These areas may provide important thermal relief, hydration, and reduced travel demands during a time when females are less mobile and piglets are most vulnerable. 

“Wild pigs are poor thermoregulators because they have few functional sweat glands, so they can’t easily cool themselves through evaporation and must rely on panting or behavioral modifications like seeking shade and water to cool down,” says Beasley. “Newborn piglets also have limited mobility their first few days, so having close access to water near the nest is likely important for both sow and piglet survival especially during hot summer months.” 

Photo of a wild pig farrowing nest constructed of palmettos within a pine forest. (Photo courtesy of Jim Beasley)

Researchers found that fine-scale understory vegetation was a stronger indicator of farrowing site selection than broader forest type. Farrowing sites consistently occurred in areas with dense ground-level cover, even when the surrounding forest classification varied. When the team examined broader land cover patterns using satellite-based data, distance to water emerged as the only strong predictor of farrowing site selection, while other landscape features such as forest type or proximity to roads were used roughly in proportion to their availability across the study area. This finding highlights an important challenge for wildlife managers: features that matter most to reproducing wild pigs often occur at a scale too fine to be detected using commonly available land cover datasets. 

“Sole reliance on remote sensing data can often lead to missing the important fine-scale cues that help us understand wildlife behaviors. Satellite data that generalize the dominant vegetation type of an area wouldn’t capture the diverse plant communities or understory vegetation composition in forested areas that are captured by boots-on-the-ground field surveys,” states Stoakley. “So while ecology research and wildlife management increasingly rely upon remote sensing technologies to inform our inferences, there is no substitute for good old-fashioned ground truthing.”  

This study suggests that wild pigs do not rely on a single habitat type when selecting farrowing sites. Instead, females appear to prioritize areas that provide nearby water and dense understory cover, regardless of the surrounding forest type. This could help explain the species’ ability to thrive in a wide range of environments, and complicates efforts to control their populations. 

By identifying the environmental features associated with farrowing sites, the research offers practical insight for wildlife managers working to detect and disrupt reproduction during key periods. Knowing where wild pigs are most likely to give birth can help guide targeted monitoring and removal efforts, particularly in areas where eradication or population reduction is a priority. 

The full study, Multi-scale predictors of farrowing site selection of wild pigs (Sus scrofa), was published in Applied Animal Behaviour Science. Authors include Travis E. Stoakley, Sarah M. Chinn, David A. Keiter, Linda S. Lee, and James C. Beasley. 

By Tyjaha Steele

Pictured above is a photograph of New Ellenton Bay (Photo courtesy of Amanda Hurst).

In wetlands, small changes in water levels and vegetation can have a significant impact on how elements move through the environment. Carolina bays are shallow, isolated, oval-shaped wetlands found throughout the southeastern United States that naturally fill and dry over time. Mercury, a naturally occurring element introduced through both natural processes and human activities, may remain stored in wetland sediments or be transformed into methylmercury, a form that more readily accumulates in plants and animals and moves through food webs. These pathways are shaped by local conditions, which vary widely among isolated wetlands in their water levels and vegetation structure.  

To better understand how these factors influence mercury cycling, researchers from the University of Georgia’s Savannah River Ecology Laboratory (SREL) examined mercury dynamics in Carolina Bay wetlands at the Savannah River Site (SRS) in South Carolina. Led by Xiaoyu Xu, an associate research scientist at SREL, the study focused on ten Carolina Bay wetlands that differ in hydroperiod, the length of time water remains in a wetland each year. Some bays dry out seasonally (short-hydroperiod bays), while others remain flooded for extended periods (long-hydroperiod bays).  

“In Carolina bay wetlands, mercury behavior is largely dictated by how long a wetland stays underwater, a cycle known as its hydroperiod. Short-hydroperiod wetlands, which are only seasonally flooded, typically act as mercury ‘sponges’ because their dense tree canopies capture mercury from the atmosphere and drop it into the soil via falling leaves,” explains Xu. “However, long-hydroperiod wetlands, those that stay flooded for most of the year, create the perfect environment for bacteria to transform inorganic mercury into methylmercury.” 

Researchers collected surface sediment samples from each wetland during both wet and dry seasons and measured concentrations of total mercury and methylmercury. Sampling across seasons and wetland types allowed the research team to compare the amount of mercury accumulated in sediments and its rate of conversion into its more toxic form under changing environmental conditions. 

Cara Love, one of the co-authors listed on the paper, is seen collecting a sample at a wetland edge (Photo courtesy of David E. Scott).

The study suggests that wetlands with shorter hydroperiods, those that dry out part of the year, accumulated higher levels of total mercury, particularly when flooded. This pattern was associated with greater tree canopy cover, which can influence how mercury enters wetland sediments through natural processes such as leaf uptake and litterfall. 

“The density of the tree canopy controls how much mercury enters the wetland in the first place and the leaves absorb mercury from the atmosphere and then deliver it directly to the ground,” says Xu. “The shade from a dense canopy protects the soil from sunlight, which would otherwise help ‘burn’ the mercury back into the atmosphere when the wetland dries out.” 

Short-hydroperiod bays primarily functioned as storage sites for total mercury, whereas long-hydroperiod bays showed a greater capacity to convert mercury into methylmercury. The highest methylmercury concentrations were observed in the central areas of long-hydroperiod bays, where prolonged flooding supports conditions favorable to mercury methylation. 

“Methylmercury is most concentrated in the deeper center areas of long-hydroperiod wetlands, which stay underwater for most of the year. These locations maintain waterlogged, oxygen-poor conditions where bacteria can convert inorganic mercury into methylmercury,” says Xu. “Identifying these hotspots allows us to better predict which species are most at risk and how methylmercury might move and magnify through the food web.” 

The study suggests that mercury methylation potential decreased with increasing total mercury concentrations and tree canopy cover, with short-hydroperiod bays functioning primarily as sinks for atmospheric mercury and long-hydroperiod bays favoring methylmercury production. 

The full study, Influence of hydroperiod and canopy cover on mercury accumulation and methylation in Carolina bay wetland sediments in the Southeastern United States, was published in Environmental Research. Authors include Chongyang Qin, David E. Scott, Stacey L. Lance, Demetrius Calloway, Cara N. Love, and Xiaoyu Xu. 

November 18, 2025

Contact: Tyjaha Steele, (803) 508 – 0892 

FOR IMMEDIATE RELEASE

Aiken, SC — The Savannah River Ecology Laboratory (SREL) is proud to announce that Dr. Olin E. “Gene” Rhodes has been selected as the recipient of the 2025 Fred C. Davison Distinguished Scientist Award. He will be formally honored and presented with the award during the 34th Annual Teller Lecture and Banquet in Aiken, South Carolina, on Nov. 21, 2025. 

The award, presented annually by Citizens for Nuclear Technology Awareness, recognizes scientists and engineers whose lifetime scientific contributions have meaningfully shaped the scientific landscape. This year’s selection of Dr. Rhodes highlights his decades-long record of scientific excellence, transformative leadership, and service to the research community. 

Since 2012, Dr. Rhodes has served as Director of SREL and, in 2021, assumed the additional directorship of the University of Georgia Research Institute. He also holds a faculty appointment at the Odum School of Ecology and serves as an Adjunct Professor at the Warnell School of Forestry and Natural Resources.  

Dr. Rhodes completed his B.S. in Biology at Furman University in 1983, earned his M.S. in Wildlife Biology from Clemson University in 1986 and received his Ph.D. in Wildlife Science from Texas Tech University in 1991. His connection to the Savannah River Site (SRS) began early in his career when he conducted master’s research at SREL in the 1980s and returned in the 1990s as a postdoctoral researcher in theoretical population genetics. 

Throughout his scientific career, Dr. Rhodes has authored or co-authored more than 250 publications covering wildlife ecology and genetics, the application of molecular tools in conservation, species reintroduction strategies, wildlife diseases and human-wildlife conflict. Throughout his career, he has also maintained a consistent role in mentoring graduate students and postdoctoral researchers. 

Under Dr. Rhodes’ leadership, SREL has expanded its research publications, modernized facilities using external funding, broadened outreach and education efforts and grown graduate, faculty, and staff populations. These accomplishments complement his professional recognition as a Fellow of the American Association for the Advancement of Science (AAAS), a Fellow of The Wildlife Society, and a member of the Sigma Xi Scientific Research Honor Society. 

This record of scientific achievement and leadership aligns closely with the legacy of Dr. Fred C. Davison, for whom the award is named. Davison’s career was defined by his commitment to encouraging math and science education and doubling graduate enrollment, principles that have guided Dr. Rhodes throughout his tenure at SREL and in his contributions to the broader scientific community. 

The Savannah River Ecology Laboratory, a research unit of the University of Georgia located near Aiken, South Carolina, studies a wide range of ecological research topics, including contaminant transport and ecotoxicology, wildlife ecology, conservation genetics, and ecosystem restoration. For further information, call SREL at 803-508-0892 or e-mail connect-srel@uga.edu.

– 30 –