Incidence of Hemorrhagic Disease in White-Tailed Deer Is Associated with Winter and Summer Climatic Conditions

Transcription

1 EcoHealth 6, 11 15, 2009 DOI: /s Ó 2009 International Association for Ecology and Health Short Communication Incidence of Hemorrhagic Disease in White-Tailed Deer Is Associated with Winter and Summer Climatic Conditions Jonathan M. Sleeman, 1 Jay E. Howell, 1 W. Matthew Knox, 1 and Philip J. Stenger 2 1 Wildlife Division, Virginia Department of Game and Inland Fisheries, 4010 West Broad Street, Richmond, VA Department of Environmental Sciences, University of Virginia Climatology Office, Charlottesville, VA Abstract: Hemorrhagic disease (HD) is an important vector-borne disease of white-tailed deer (Odocoileus virginianus). The objective of this study was to determine whether temperature and precipitation were associated with a measure of annual incidence of HD in white-tailed deer from Virginia. The annual percentages of deer with hoof wall growth interruptions (a clinical sign of HD) from four climate divisions in the HD endemic area of Virginia recorded during were used as indicators of annual HD incidence. Pearson s correlation coefficients between these indicators of incidence and average temperature ( F) or total precipitation (in.) for each month, as well as for winter (January February), early summer (June July), and late summer/fall (August September October) seasons were calculated. Strong direct correlations between the measure of annual HD incidence and average temperature for winter (r = 0.39, P = 0.003, n = 57), early summer (r = 0.51, P < , n = 57), and late summer/fall (r = 0.42, P = 0.001, n = 57) were evident. There also was a strong inverse correlation between the measured annual HD incidence and June precipitation (r = -0.44, P = , n = 57). Poisson regression models of seasonal temperatures and June precipitation to annual percentage of deer with hoof wall growth interruptions were developed. Based on Akaike s Information Criterion with small sample size correction (AICc), the global model was selected as the top model. Higher winter and summer temperatures may increase vector capacity and competence, and lower precipitation in June may create favorable breeding sites for midges. Keywords: climate, hemorrhagic disease, incidence, precipitation, temperature, white-tailed deer Hemorrhagic disease (HD) is one of the most important diseases of white-tailed deer (Odocoileus virginianus) in the southeastern United States. It is caused by epizootic hemorrhagic disease (EHD) or bluetongue (BLU) viruses, which are in the genus Orbivirus in the family Reoviridae. There are numerous serotypes, of which EHD serotype 2 predominates in this region (Stallknecht et al., 1995). These double-stranded RNA viruses are vector transmitted, and Published online: May 9, 2009 Correspondence to: Jonathan M. Sleeman, jsleeman@usgs.gov Culicoides midges of the species Culicoides sonorensis (formerly C. variipennis) are thought to be the most important (Howerth et al., 2001). These viruses have been associated with substantial mortality of deer populations, and clinical signs can range from acute death to chronic disease. Both BLU and EHD viruses replicate in endothelium in whitetailed deer, causing cell damage that activates the coagulation system, which results in generalized edema, hemorrhagic diathesis, and ulcerations. Infected deer will usually develop a fever that parallels the viremia. Chronic lesions

2 12 Jonathan M. Sleeman et al. include a heavy overgrowth of the hooves with an indentation or crack in the wall (Figure 1), corresponding to interruption of growth during acute illness (Howerth et al., 2001). In addition, ruminal scarring can occur, secondary bacterial infections, and subsequent emaciation. Deer that survive acute infection may be observed with these chronic lesions, such as hoof wall growth interruptions, which provide an easily recognizable measure of recent HD disease events. The disease is seasonally predictable corresponding to the seasonal activity of the midges but is somewhat unpredictable on an annual basis. Vector-borne pathogens are considered to be particularly sensitive to variations in climate (Purse et al., 2005), and previous studies of BLU virus exposure of domestic ruminants have indicated that climatic variables, such as temperature, rainfall, and humidity, may be important for viral transmission by increasing vector capacity and competence (Gibbs and Greiner, 1989; Gibbs, 1992; Wright et al., 1993). Consequently, the objective of this study was to determine whether selected climatic conditions were associated with annual percentage of white-tailed deer from Virginia with hoof wall growth interruptions, which was used as an indicator of annual incidence of HD. Since 1993, hunters in Virginia who participate in the Virginia Department of Game and Inland Fisheries Deer Management Assistance Program (DMAP; virginia.gov/wildlife/deer/dmap.asp) have collected data on the number of deer harvested that have hoof wall growth interruptions, which as described earlier is a clinical sign of Figure 1. Hoof of white-tailed deer showing typical hoof wall growth interruption or cracked hoof as a result of hemorrhagic disease. HD in deer that survive infection. Approximately 10% of the annual deer harvest of ,000 deer was examined by hunters for these lesions (mean ± SD of deer examined for hoof lesions during inclusive = 16,658 ± 2,881, minimum = 12,177, maximum = 21,893). Only deer with two or more hooves displaying clear evidence of cracked hooves were recorded as positive for HD. These data were collected during hunting season, i.e., mostly during November and December of each year. Consequently, the annual percentages of deer with hoof wall growth interruptions from 1993 to 2006 inclusive were calculated and used as a standardized indicator of annual HD incidence. Only deer from the counties in four climate divisions (1 4) east of the Blue Ridge Mountains were included in the study, and percentages were calculated separately for each climate division (Figure 2). This area is considered the HD endemic area in Virginia (Stallknecht et al., 1991), and data were very sparse from counties west of the Blue Ridge Mountains. In addition, data on average temperature ( F) and total precipitation (in.) by month for each of the four climate divisions for the same time period were obtained from the University of Virginia Climatology Office database. Examples of the data used in this study are listed in Table 1. Pearson s correlation coefficients between the annual percentage of deer with hoof wall growth interruptions (our indicator of annual HD incidence) and temperature or precipitation for each month, as well as for winter (January February), early summer (June July), and late summer/fall (August September October) seasons were calculated (n = 57, representing data from four separate climate divisions for 14 years as well as data from 2007 for one climate division). These seasons were selected based on their potential importance in the life cycle of the vector as well as known transmission dynamics of the virus. All statistical analyses were performed using the SAS Statistical Package (SAS Institute, Inc., Cary, NC, USA). Direct correlations between annual percentage of deer with hoof wall growth interruptions and several average monthly temperatures were found (January, February, April, June, July, August, September, and October); however, average monthly temperatures often were correlated with each other. More meaningfully, strong direct correlations between this measure of annual HD incidence and average temperature for winter (r = 0.39, P = 0.003, n = 57), early summer (r = 0.51, P < , n = 57), and late summer/fall (r = 0.42, P = 0.001, n = 57) also were evident. There also was a strong inverse correlation between

3 Climate and Hemorrhagic Disease 13 Figure 2. Climate divisions of Virginia. Table 1. Selected Data Used in Our Study of the Association of Annual Incidence of Hemorrhagic Disease (HD) as Measured by Percentage of Deer Examined with Hoof Wall Growth Interruptions, and Climatic Conditions, i.e., Monthly and Seasonal Average Temperatures ( F) and Total Precipitation (in.) Data June temperature June precipitation % HD incidence Data shown are means for all four Virginia climate divisions included in the study the measure of annual HD incidence and June precipitation (r = -0.44, P = , n = 57). To further explore these relationships Poisson regression models of the seasonal temperatures and June precipitation to annual percentage of deer with hoof wall growth interruptions were developed. The model parameters are shown in Table 2. Based on Akaike s Information Criterion with small sample size correction (AICc), the global model was selected as the top model. The model was then used to predict the percentage of deer with hoof wall growth interruptions for 2007 (Table 3), and the average difference between the observed incidence, and the predicted incidence was 0.48%. Climatic conditions seem to have a strong influence on our measure of incidence of HD in Virginia, and our findings are consistent with previous studies on BLU virus transmission dynamics (Purse et al., 2005). Higher winter and summer temperatures in Virginia probably increase vector capacity and competence (Gibbs, 1992; Wright et al., 1993; Purse et al., 2005; Saegerman et al., 2008) by increasing the vector s ability to acquire, maintain, and transmit the virus. In particular, higher summer temperatures will enhance replication of the virus within the vector, resulting in higher viral load and decreased extrinsic incubation period. Culicoides activity is positively correlated with temperature, and higher summer temperatures increase vector abundance, vector survival, and biting and transmission rates (Purse et al., 2005; Saegerman et al., 2008). In addition, Culicoides sonorensis overwinters in the larval stage as third instars (Vaughan and Turner, 1987) and are mostly recovered from ponds at the ice/mud interface of water bodies. Consequently, warmer winter temperatures increase overwinter survival of the larvae greatly influencing summer seasonal emergence patterns and population density of adult flies (Purse et al., 2005). Furthermore, the inverse correlation between the measure of annual HD incidence and June precipitation may be explained by the vector life cycle. During the summer months most C. sonorensis larvae inhabit the surface mud at or near the water line (Vaughan and Turner, 1987). Therefore, the relatively lower precipitation in June may create favorable breeding sites for midges due to the creation of larger mudflats from receding waterlines. However, other studies have stated that increased summer and fall precipitation will increase seasonal incidence of BLU virus

4 14 Jonathan M. Sleeman et al. Table 2. Poisson Regression Model Parameters of the Seasonal Temperatures and June Precipitation to Annual Hemorrhagic Disease Incidence in White-Tailed Deer from Virginia, as Measured by Percentage of Deer Examined with Hoof Wall Growth Interruptions, Parameter df Estimate Standard error Lower CI Upper CI v 2 P Intercept < June precipitation < Winter temperature < Early summer temperature < Late summer/fall temperature < Table 3. Predictions of Percentage of White-Tailed Deer with Hoof Wall Growth Interruptions (as a Measure of Hemorrhagic Disease Incidence) from Four Climate Divisions in Virginia During 2007 Using Poisson Regression Model of Seasonal Temperatures and June Precipitation Climate division Observed incidence (%) Predicted incidence (%) Difference (%) transmission by increasing vector abundance, among other factors (Purse et al., 2005). These apparent differences in the influence of precipitation on EHD and BLU viral transmission warrant further study. Finally, warmer and drier summer climatic conditions may have an impact on white-tailed deer that influence viral transmission dynamics and the development of clinical disease. Because the hoof of deer grows at 0.5 cm per month (Miller et al., 1986), by next hunting season any previous growth interruptions should have been removed. Therefore, the annual percentage of harvested deer with cracked hooves should be measuring new cases each year, making the monitoring of cracked hooves potentially an excellent measure of HD incidence, and therefore risk of disease. However, there are potential issues with the numerator in that there may have been other causes of cracked hooves, and presumably not all deer that survive infection with HD will develop hoof wall growth interruptions. Furthermore, there may be issues with the denominator in that not all deer harvested were potentially susceptible to HD or at risk for exposure. However, assuming these other factors remained relatively constant then the annual percentage of cracked hooves is a potentially good indicator of annual incidence of HD, and our data show some interesting trends with an increasing incidence since 1998, and a regular 3-year cycle of rising and falling incidence (Figure 3). This 3-year cycle is consistent with previous studies (Stallknecht et al., 1991). Further studies at multiple spatial scales using additional climatic data, such as number of days of rainfall and evaporation rates, are warranted. In addition, these studies should include biotic factors that may influence incidence, including acquired and innate immunity of deer, as well as deer densities and virus serotype (Howerth et al., 2001). The development of other standardized indicators of annual incidence as well as predictive models based on other HD diagnostic criteria also would be beneficial to wildlife managers. Whereas a link to global warming is speculative, this study illustrates the potential effects that climate change may have on a wildlife vector-borne disease, and with average temperatures continuing to increase, we may see increasing frequency of more severe outbreaks of HD or its spread into new geographic areas due to the change in vector distribution (Gibbs, 1992; Sellers, 1992; Purse et al., 2005). This could have a negative impact on deer populations with resultant need to alter deer management practices. Virginia has seen a general warming during the past decade; 1998 was the warmest year on record (statewide average temperature of 57.4 F), and 9 of the 10 hottest years have occurred since that date (University of Virginia Climatology Office database). Interestingly, this change in statewide average temperature coincides with the apparent increasing annual incidence of HD as measured by the percentage of deer with cracked hooves (Figure 3). Coincidentally, 1998 was the year when various BLU viral strains started to spread across 12 European countries and 800 km further north than previously reported (Purse et al., 2005). It has been suggested that this spread into Europe was the result of recent changes in climate.

5 Climate and Hemorrhagic Disease 15 Figure 3. Percentage ± 95% confidence interval of white-tailed deer examined by hunters exhibiting cracked hooves from east of the Blue Ridge Mountains, Virginia, from There is widespread scientific agreement that the world s climate is changing and that the weight of evidence demonstrates that anthropogenic factors have and will continue to contribute significantly to global warming and climate change. It is anticipated that continuing changes to the climate will have serious negative impacts on public, animal, and ecosystem health due to extreme weather events, changing disease transmission dynamics, emerging and re-emerging diseases, and alterations to habitat and ecological systems that are essential to wildlife conservation. ACKNOWLEDGMENTS This project was funded by Federal Aid to Wildlife Restoration Act under Project WE-99, administered by the Virginia Department of Game and Inland Fisheries. The authors thank Kendell Ryan for her generous assistance with the figures. REFERENCES Gibbs EPJ (1992) Epidemiology of orbiviruses-bluetongue: towards 2000 and the search for patterns. In: Bluetongue, African Horse Sickness, and Related Orbiviruses, Walton TE, Osburn BI (editors), Boca Raton, Florida: CRC, pp Gibbs EPJ, Greiner EC (1989) Bluetongue and epizootic hemorrhagic disease. In: The Arboviruses: Epidemiology, Ecology, Vol 2, Monath TP (editor), Boca Raton, Florida: CRC, pp Howerth EW, Stallknecht DE, Kirkland PD (2001) Bluetongue, epizootic hemorrhagic disease and other orbivirus-related diseases. In: Infectious Diseases of Wild Mammals, Williams ES, Barker I (editors), Ames, Iowa: Iowa State University Press, pp Miller KV, Marchington RL, Nettles VC (1986) The growth rate of hooves of white-tailed deer. Journal of Wildlife Diseases 22: Purse BV, Mellor PS, Rogers DJ, Samuel AR, Mertens PPC, Baylis M (2005) Climate change and the recent emergence of bluetongue in Europe. Nature Reviews Microbiology 3: Saegerman C, Berkens D, Mellor PS (2008) Bluetongue epidemiology in the European Union. Emerging Infectious Diseases 14: Sellers RF (1992) Weather, Culicoides, and the distribution and spread of bluetongue and African horse sickness viruses. In: Bluetongue, African Horse Sickness, and Related Orbiviruses, Walton TE, Osburn BI (editors), Boca Raton, Florida: CRC, pp Stallknecht DE, Blue JL, Rollor EA III, Nettles VF, Davidson WR, Pearson JE (1991) Precipitating antibodies to epizootic hemorrhagic disease and bluetongue viruses in white-tailed deer in the southeastern United States. Journal of Wildlife Diseases 27: Stallknecht DE, Nettles VF, Rollor EA III, Howerth EW (1995) Epizootic hemorrhagic disease virus and bluetongue virus serotype distribution in white-tailed deer in Georgia. Journal of Wildlife Diseases 31: Vaughan JA, Turner EC (1987) Seasonal microdistribution of immature Culicoides variipennis (Diptera: Ceratopogonidae) at Saltville, Virginia. Journal of Medical Entomology 24: Wright JC, Getz RR, Powe TA, Nusbaum KE, Stringfellow DA, Mullen GR, et al. (1993) Model based on weather variables to predict seroconversion to bluetongue virus in Alabama cattle. Preventive Veterinary Medicine 16:

6