Canine distemper virus phylogenetic structure and ecological correlates of infection in mesocarnivores across anthropogenic land use gradients

Abstract

Anthropogenic land use impacts infectious diseases at the wildlife-domestic-human interface by changing host spatial distribution, behavior, density, and population dynamics. Canine distemper virus (CDV) is a significant cause of morbidity and mortality in many wild and domestic animals. Given the propensity of CDV to infect synanthropic mesocarnivores, it is important to investigate host and environmental factors affecting mesocarnivore CDV infection. Here, we investigated patterns of CDV infection and developed a statistical model to identify environmental variables related to CDV risk in commonly affected mesocarnivores. We sampled carcasses (N = 270) submitted to the Southeastern Cooperative Wildlife Disease Study from January 2019 to December 2022 and sequenced the CDV H-gene of 32 CDV-positive animals. Overall, 158 out of 270 mesocarnivores (58.5%) and four species (raccoon, red fox, gray fox, and striped skunk) were diagnosed with CDV across 13 states. Ripley's K analysis showed positive cases were more spatially clustered at larger distances than expected due to chance. A generalized linear model for CDV-infected animals showed surface imperviousness, precipitation, and subadult/adult age classes were significant positive explanatory variables, but elevation had a significant negative association with CDV infection likelihood. H-gene sequence diversity among wild mesocarnivores in the southeastern United States was geographically separated into groups east and west of the Mississippi River, with only two eastern samples clustering with western groups. By identifying areas of intense human development at the highest risk for CDV, it may be possible to focus surveillance efforts in these areas, allowing for earlier outbreak identification, potentially preventing cross-species CDV transmission.

Importance: Anthropogenic land use change can impact infectious disease spread by altering animal distribution and behavior. Canine distemper virus (CDV) is a significant cause of morbidity and mortality in wild and domestic carnivores. This study investigated how land use influences CDV infection in wild carnivores by examining tissues collected between 2019 and 2022 from wild carnivores found dead in the southeastern United States. CDV strains were geographically distinct, with differences between populations east and west of the Mississippi river. Statistical models showed areas with increased human development and higher precipitation had higher CDV risk; however, there was lower risk associated with higher elevations and younger animals. The importance of this study is that it identifies geographic structure of CDV in the southern United States, and identifies land-use associations with potential high-risk areas for CDV transmission-information that is useful for wildlife disease surveillance and control strategies.

Keywords: Morbillivirus; anthropogenic land use; canine distemper virus; disease ecology; gray fox; mesocarnivores; phylogenetic structure; raccoon; striped skunk.

Fig 1

Map of states from which necropsy cases of mesocarnivore species were analyzed for canine distemper virus infection from January 2019 to December 2022 (n = 270). The location of Mississippi River is blue continuous line and shown by the arrow.

Fig 2

Spatial clustering analysis based on Ripley’s K-function of cases of select mesocarnivore species submitted to Southeastern Cooperative Wildlife Disease Study from January 2019 to December 2022. Plot A: Entire data set. Plot B: CDV-positive cases. Plot C: CDV-negative cases. r (the x-axis) represents distance from a point and K(r) represents the K-function. 𝐾̂ 𝑜𝑏𝑠(𝑟) (solid line) is the Ripley’s K-statistic for the observed cases. 𝐾𝑡ℎ𝑒𝑜(𝑟) (dashed line) is the K-statistic for a completely random (Poisson) point process (no spatial clustering). 𝐾̂ ℎ𝑖 (𝑟) and 𝐾̂ 𝑙𝑜(𝑟) (shading around dashed line) are the upper and lower confidence interval envelopes for the Poisson simulation. When the solid line of observed data points is greater than the dashed line and error bands (gray area), then there is spatial clustering of cases (case clustering is greater than a random distribution). When the solid line of observed data points is within the random K-statistic, then spatial clustering is similar to random.

Fig 3

Marginal effects plots of predicted CDV infection probabilities for significant predictors in the best-fit model. (A) Predicted effects of impervious surface on CDV infection probabilities with data points from this study. (B) Predicted effects of the interaction between impervious surface and binned elevation values on CDV infection probability. (C) Predicted effects of precipitation on CDV infection probability with data points from this study. (D) Predicted effects of the interaction between raccoons, temperature, and CDV infection probability with data points from this study. (E) Predicted effects of the interaction between striped skunks and CDV infection probability with data points from this study.

Fig 4

Phylogenetic tree for H-gene of canine distemper virus isolates from necropsy cases of mesocarnivore species submitted to the Southeastern Cooperative Wildlife Disease Study from January 2019 to December 2022. This study produced isolates in colored text. Fuchsia, blue, and red color isolates correspond to isolates collected from gray foxes, skunks, and raccoons, respectively. Those in black text were accessed from Genbank. The orange-highlighted clade is the proposed eastern group of viruses isolated in the study from CDV cases in states east of the Mississippi River (NC, GA, and FL). The gray shading corresponds to the western group with viruses from AR, MO, and CO.