Invasive Burmese pythons alter host use and virus infection in the vector of a zoonotic virus

Zoonotic vector-borne diseases are significantly influenced by the composition of wildlife communities. More diverse communities often exhibit lower infection prevalence in vectors due to a dilution effect. The invasive Burmese python poses a substantial threat to this balance in southern Florida. Its predation has decimated populations of large and medium-sized mammals, fundamentally altering the local ecosystem and potentially increasing the risk of zoonotic disease transmission. This study focuses on the Everglades virus (EEVV), a zoonotic pathogen endemic to Florida, and its transmission cycle involving mosquitoes and rodents. The research question revolves around how the Burmese python's impact on mammal communities indirectly affects EEVV prevalence in its mosquito vector, ultimately influencing human disease risk. The 'perturbation hypothesis' suggests that human-induced ecosystem disruptions, like invasive species, shift interspecies transmission dynamics and pathogen prevalence, impacting human health. Previous research in the northern Everglades demonstrated shifts in vector host use following python invasion, with increased feeding on rodents and a subsequent increase in *Ehrlichia canis* infection. This study aimed to investigate whether a similar pattern exists for EEVV and to quantify the indirect effects of the python on EEVV transmission.

Extensive research highlights the impacts of invasive species on ecosystems and their cascading effects. Studies have shown that invasive species can disrupt trophic cascades, leading to significant ecosystem damage and community impacts. Of particular concern are indirect impacts on host species and the subsequent alteration of pathogen transmission dynamics. The dilution effect, where the presence of non-competent hosts reduces pathogen prevalence in vectors, is a crucial concept in disease ecology. Various factors, including land-use change and climate change, can disrupt these dynamics, impacting vector-borne disease risk. The literature emphasizes the complex interplay between biodiversity, vector-host interactions, and the risk of human infection. Prior research in the Everglades has linked the Burmese python's predation to declines in native mammal populations, altering the food web and ecosystem function. This has been demonstrated with the vector of *Ehrlichia canis*, where a dramatic shift in vector host use towards rodents has occurred. This study builds upon this foundation, investigating the potential indirect impacts of the python on EEVV prevalence in its vector mosquito.

The study compared ecologically similar areas in southern Florida with varying Burmese python densities. Researchers quantified mammal diversity and activity, vector host use, and virus incidence in areas with high and low python presence. Twelve sites were established across three preserves, with mosquitoes sampled using carbon dioxide-baited light traps and battery-powered aspirators. Mammal activity was assessed using camera trapping and Sherman live traps. Blood meals from engorged female mosquitoes were analyzed using PCR-based blood meal analysis, followed by Sanger sequencing to identify the host species. Data analysis involved binomial generalized linear mixed effects models (GLMMs) to investigate the relationship between vector infection rates, relative cotton rat host use (ratio of cotton rat blood meals to total blood meals), and metrics describing mammal community composition (cotton rat activity, non-rodent activity, and non-rodent diversity). The most parsimonious models were selected based on AICc weights. The relative probability of python presence was modeled using georeferenced python locations from the Early Detection and Distribution Mapping System database. To identify potential dilution hosts, researchers analyzed host use patterns of the vector across sites with varying python presence. Foraging rates were calculated as the ratio of species utilization to relative abundance.

Blood meals from cotton rats, the confirmed EEVV host, varied significantly among sites (0-63.4%). GLMM analysis revealed that increasing non-rodent diversity (dilution hosts) was associated with decreasing cotton rat host use, while increasing cotton rat host use was linked to higher EEVV infection rates in mosquitoes. The most parsimonious model included cotton rat activity and non-rodent diversity as predictors of cotton rat host use. Relative cotton rat host use was a strong predictor of EEVV infection rates. Non-rodent activity and species richness decreased dramatically with increasing python presence. The reduction in non-rodent activity coincided with increased cotton rat host use by the vector. Marsh rabbits and white-tailed deer emerged as potential dilution hosts, exhibiting high feeding rates by the vector at sites with low to moderate python presence, but absent at sites with high python presence. The highest EEVV infection rate was observed at a site with low mammal diversity and exclusively rodent blood meals. Overall, the study demonstrates a strong negative effect of non-rodent diversity on vector host use and an indirect, but significant, effect on EEVV infection rates in the vector.

The findings strongly support the hypothesis that invasive predators can indirectly alter vector-borne pathogen transmission. The Burmese python's impact on mammal communities significantly influences vector host use and EEVV infection rates. Changes in host community structure do not directly affect infection rates but rather indirectly impact rates through altered host-use patterns by the vector. The inverse relationship between non-rodent diversity and cotton rat host use underscores the importance of dilution hosts in regulating pathogen prevalence. Similar complex relationships have been observed in other vector-borne pathogen systems, emphasizing the importance of host community composition. The identification of marsh rabbit and white-tailed deer as dilution hosts highlights the conservation value of maintaining biodiversity to mitigate disease risk. While the study suggests increased vector infection, this does not necessarily translate to increased human disease incidence, requiring further research into human-vector contact.

This study demonstrates that the Burmese python's invasion indirectly increases the risk of EEVV transmission by altering the structure of mammal communities in southern Florida. The loss of dilution hosts leads to increased vector feeding on competent hosts (cotton rats), resulting in higher EEVV infection prevalence in mosquitoes. Conservation efforts focused on maintaining biodiversity are crucial for mitigating the risk of zoonotic diseases exacerbated by invasive species. Future research should focus on more precise quantification of human risk by incorporating additional factors that influence animal communities and enhancing sampling methodology.

The study's inferences on python presence relied on citizen science data, which might be subject to biases. Sample sizes of mosquitoes varied across sites, potentially affecting the precision of infection rate estimates. The observed increase in vector infection does not directly equate to higher human disease incidence, requiring further investigation of human-vector contact and other factors influencing human infection rates. The study area may not be fully representative of all habitats affected by the Burmese python.