C. Sheldon Davis, R. Wes Flynn, Cara N. Love, Stacey L. Lance, and David E. Scott
Savannah River Ecology Laboratory 2016 Summer Research Experience
In recent decades, infectious diseases have contributed to population declines in several taxa of wildlife across the globe. While many of these diseases have been around, they have recently become more apparent and cause for conservation concern across the United States with increased die-off events (i.e. viruses in bees, white nose syndrome in bats, and chytrid fungus and ranavirus in reptiles, amphibians, and fish).
Multiple Stressors?
Stressors are biological, chemical, or physical effects that negatively influence the success of an organism or an ecosystem. These die-off events could be attributed to the effects of multiple stressors, rather than any single stressor. Amphibians specifically are experiencing global declines due to a variety of natural and anthropogenic stressors. The causes of many declines are complex; however emerging infectious diseases, including ranavirus (RV), and contaminants have both been implicated. Heavy metals are common environmental contaminants due to human activities and can negatively affect growth, development, and survival at levels commonly found in the environment. RV is an emerging infectious disease implicated in die-offs globally, but how these commonly occurring stressors interact is still largely unknown.
To examine how metals could influence susceptibility and interact with disease in larval amphibians, we exposed southern toad (Anaxyrus terrestris) and eastern narrowmouth toad (Gastrophryne carolinensis) larvae to environmentally relevant levels of copper (Cu) in combination with a local RV strain. To assess potential lethal and sublethal impacts on larvae we measured, survivorship, RV loads, and growth rates.
Our objectives were to 1) determine if chronic copper (Cu) exposure influences susceptibility to ranavirus (RV) in larval southern and eastern narrowmouth toads, and 2) to attempted to identify any sub-lethal effects of Cu-RV interactions on larval individuals treated with both copper and ranavirus.
We predicted that chronic copper exposure will increase susceptibility to ranavirus. We used ranavirus as our disease of interest since it is known to cause mortality events in larval amphibians. As previously mentioned, heavy metals are also stressors for amphibians; our second stressor of interest is the heavy metal copper. Copper, like other heavy metals, is known to negatively affect larval growth and development in amphibians, even at minute concentrations.
Methods
Our study species were the southern toad (Anaxyrus terrestris) (left above) and the eastern narrowmouth toad (Gastrophryne carolinensis) (right above). These species are commonly found in and around both our reference and metal contaminated sites, and have both been used in metal contaminant studies before. The southern toad is closely related to several other species of toad across the U.S., and can live in a wide range of habitat types. The eastern narrowmouth toad is also a native Anuran species found across the Southeast, and is known to be more susceptible to contaminants. Eastern narrowmouth toads additionally have not been involved in many (if any) studies regarding disease susceptibility.
Our 4 study sites consisted of wetlands on the Department of Energy, Savannah River Site (SRS): a nearly 200,000-acre National Environmental Research Park in Aiken and Barnwell, South Carolina. The SRS is mainly upland pine and hardwood forest, and around twenty percent of the Site is made up of river swamp, bottomland hardwood, streams, and Carolina bays. Two wetlands on site that have elevated levels of copper and other heavy metals from ash plume and industrial runoff served as our metal contaminant sites. Two additional non-metal contaminated wetlands served as reference sites.
Adult toads were captured in drift fences at the metal contaminated and reference wetlands on the SRS, and then bred within wetland populations in the lab. Adults were paired in individual containers and injected with human chorionic gonatotrophin to induce mating.
Upon fertilization, viable embryos were then taken and placed in several copper treatments (0μg Cu/ L (control), 5μg Cu/ L, 10μg Cu/ L, and 15μg Cu/ L) for the duration of the experiment (3 weeks). Southern toads were subjected to 0 and 15μg Cu/ L treatments given their greater tolerance to copper contamination, while the more contaminant-sensitive eastern narrowmouth toads were subjected to 0, 5, and 10μg Cu/ L. All water was first filtered and treated by volume in carboys that were labeled and tested for accuracy following treatment.
Factorial Design
Southern toad: 4 populations x 2 [Cu] x 2 disease x 14 replicates = 224 units
Eastern narrowmouth toad: 4 populations x 3 [Cu] x 2 disease x 14 replicates = 336 units
All water was changed twice per week, and all larvae were fed throughout. Mortality was recorded daily, and each individual larva was assigned an ID (indicating treatment and parental wetland) and photographed from one week after fertilization during water changes until the conclusion of the study in a well using a scale ruler and a fixed-height camera. Photos were analyzed at the conclusion of the study using ImageJ software to calculate individual growth rates through development.
Given that ranavirus has been documented on the Savannah River Site, and is also found across the eastern U.S., it was therefore an appropriate disease for our study system. Two weeks into development, half of the larvae were exposed to ranavirus, and half were exposed to virus culture media (which did not contain ranavirus) for 24 hours; the purpose of exposing to culture media was to make sure there was nothing about the culture itself that could have affected them. After exposure, all individuals were returned to their respective copper treatments for the remaining week of the study. Containers of individuals exposed to ranavirus were marked as such and carefully stored and handled away from the non-exposed treatment containers for the duration of the experiment. Water samples were then taken from non-exposed containers and also tested for ranavirus intermittently to determine if cross-contamination between RV and non-RV individual containers occurred at any point after exposure. At the end of the study, we collected the larvae and extracted their DNA to quantify viral loads (total percentage of viral DNA accumulated within the host) for each larvae using real-time, quantitative polymerase chain reaction (qPCR).
Data Analysis
Survival analyses and viral load assessments were conducted using Cox proportional hazards models (which examine differences in probability of survival over a given interval of time) and linear regression respectively. Growth rates were analyzed using mixed linear regression, with tadpole ID as a random effect to account for the non-independence of the measurements made on the same animals over time. All analyses were conducted in R version 3.2.4 at alpha values equal to 0.05.
Survival Probabilities
Figure 1a. depicts modeled survival probability over time for all treatments of southern toads using Cox proportional hazard models . On the x-axis, we have days in treatment, and on the y-axis we have survival probability. The black lines show survival probability of larvae exposed that were not exposed to ranavirus, and the red line represents those that were exposed to ranavirus. The broken lines represent larvae that were chronically exposed to copper. We found that there was no significant affect of treatments (control, RV, Cu) nor site type on survival probability over the course of the study.
Similarly we modeled survival probability over time for eastern narrowmouth toads in each treatment (Figure 1b). For eastern narrowmouth toads, parental site type had a significant affect on survival probability (p=0.0007), and the risk of mortality was 25% greater for larvae exposed to ranavirus (p=0.005).
Viral Loads
Here we have the quantified viral loads in southern toads exposed to ranavirus (Figure 2a.). On the x-axis we have the parental wetland type (metal contaminated and reference), and on the y-axis we have viral load. Black dots and triangles represent larvae that were exposed to 0 parts copper, while those in yellow were exposed to 15 μg Cu/ L,. Larvae exposed to both copper and ranavirus had lower mean viral loads than those not exposed to cu, though the differences are not statistically different; neither parental site type nor copper exposure had a significant effect on viral load in southern toad larvae.
We also quantified viral loads in eastern narrowmouth toads exposed to ranavirus (Figure 2b.). Again we have parental site type on the x-axis, and viral load on the y-axis. Color denotes the copper treatment: black being 0 μg Cu/ L, yellow 5 μg Cu/ L, and red 10 μg Cu/ L. While larvae exposed to both copper and ranavirus (yellow and red) exhibited lower mean viral loads that those that were not exposed to copper, the differences are not statistically significant.
Growth Rate
Finally, we analyzed growth rates in southern toads before and after ranavirus exposure. On the x-axis we have RV exposure period (pre- and post-RV exposure), and on the y-axis we have average growth rates for larvae each treatment type. Color denotes the copper treatment (black shapes were exposed to 0 μg Cu/ L, and all yellow shapes exposed to 15 μg Cu/ L), and shape denotes whether or not treatments were exposed to ranavirus (triangles) or virus culture (circles) treatments. We can see that copper had a significant affect on growth rate in all treatments (p<0.0001). Both ranavirus and non-Rv treatment groups had the same growth rates before going into exposure. Growth rates decreased across all treatments, but only ranavirus treated groups saw significant decreases in growth rate (p=0.011).
Unfortunately we were unable to analyze growth rates of eastern narrowmouth toads due to time constraints of the REU internship.
These results highlight the difficulty in predicting interactions between environmental stressors in amphibians, and suggest that RV can negatively impact amphibians even in the absence of increases in mortality. While there was some evidence of lower viral loads in larvae chronically exposed to copper, future studies should be conducted to confirm this observation. Although copper served as a proxy for heavy metal contaminants in this study, the effects other heavy metals and environmental contaminants (i.e. pesticides, dissolved antibiotics) have on disease susceptibility and interactions could also be studied.
Acknowledgments
References
Daszak, P., Cunningham, A. A., & Hyatt, A. D. (2003). Infectious disease and amphibian population declines. Diversity and Distributions, 9(2), 141-150.
Flynn, R., Scott, D. E., Kuhne, W., Soteropoulos, D., & Lance, S. L. (2015). Lethal and sublethal measures of chronic copper toxicity in the eastern narrowmouth toad, Gastrophryne carolinensis. Environmental toxicology and chemistry, 34(3), 575-582.
Forson, D. D., & Storfer, A. (2006). Atrazine increases ranavirus susceptibility in the tiger salamander, Ambystoma tigrinum. Ecological Applications, 16(6), 2325-2332.
Lance, S. L., Flynn, R. W., Erickson, M. R., & Scott, D. E. (2013). Within-and among-population level differences in response to chronic copper exposure in southern toads, Anaxyrus terrestris. Environmental pollution, 177, 135-142.
Redick, M. S., & La Point, T. W. (2004). Effects of sublethal copper exposure on behavior and growth of Rana pipiens tadpoles. Bulletin of environmental contamination and toxicology, 72(4), 706-710.
R Studio. Version 3.2.4 for Windows. R Development Core Team. March 2016.
Rumrill, C. T., Scott, D. E., & Lance, S. L. (2016). Effects of metal and predator stressors in larval southern toads (Anaxyrus terrestris). Ecotoxicology, 25(6), 1278-1286.
[Voice Of America News]. 2012, April 13. Reptiles, Amphibians in US Succumbing to Deadly Ranavirus. [Video file]. Retrieved from https://youtu.be/pVYQK5WxGr8
Credits:
All photos by Sheldon Davis. Video by Voice of America news broadcast.