Landscape-level toxicant exposure mediates infection impacts on wildlife populations

Abstract

Anthropogenic landscape modification such as urbanization can expose wildlife to toxicants, with profound behavioural and health effects. Toxicant exposure can alter the local transmission of wildlife diseases by reducing survival or altering immune defence. However, predicting the impacts of pathogens on wildlife across their ranges is complicated by heterogeneity in toxicant exposure across the landscape, especially if toxicants alter wildlife movement from toxicant-contaminated to uncontaminated habitats. We developed a mechanistic model to explore how toxicant effects on host health and movement propensity influence range-wide pathogen transmission, and zoonotic exposure risk, as an increasing fraction of the landscape is toxicant-contaminated. When toxicant-contaminated habitat is scarce on the landscape, costs to movement and survival from toxicant exposure can trap infected animals in contaminated habitat and reduce landscape-level transmission. Increasing the proportion of contaminated habitat causes host population declines from combined effects of toxicants and infection. The onset of host declines precedes an increase in the density of infected hosts in contaminated habitat and thus may serve as an early warning of increasing potential for zoonotic spillover in urbanizing landscapes. These results highlight how sublethal effects of toxicants can determine pathogen impacts on wildlife populations that may not manifest until landscape contamination is widespread.

Keywords: ecotoxicology; host–pathogen interaction; mathematical model; pollution; urbanization.

Figure 1.

(a) Schematic of the compartmental model. Squares represent host population according to infection status (susceptible or infected) and habitat (pristine or toxicant-contaminated). The parameter f represents the fraction of toxicant-contaminated landscape. Horizontal arrows (purple) represent movement between pristine and toxicant-contaminated habitats, vertical arrows (orange) represent transitions between susceptible and infected classes, and diagonal arrows (green) represent demographic processes. Dotted arrows represent processes affected by toxicants. (b) Differential equations of the model, colour-coded to represent movement, infection and demographic processes as in (a). (c) Model parameters with definitions, units and default values used in model simulations.

Figure 2.

Equilibrium population size (a), infection prevalence (b) and spillover risk (the density of infected animals in toxicant-contaminated habitat) (c) plotted as a function of the proportion of toxicant-contaminated habitat. In all panels, the dispersal cost from toxicant-contaminated habitat is relatively low (c_σ = 0.2). Transmission is constant in pristine habitat (β_P = 0.006) and varies in toxicant-contaminated habitat (β_T = 0.0015, 0.006 and 0.0105; orange, blue and purple lines, respectively). Population size in the absence of infection is shown for comparison in (a) (black line). Other parameter values are provided in figure 1.