Habitat fragmentation as a result of biotic and abiotic factors controls pathogen transmission throughout a host population

1. The hypothesis that habitat fragmentation (biotic or abiotic) alters the transmission of disease within a population is explored using field data from a well-studied amphibian-pathogen system. 2. We used the Ambystoma tigrinum-A. tigrinum virus (ATV) model system to show how habitat fragmentation as a result of emergent vegetation and habitat management affects disease transmission dynamics in ponds across a landscape. 3. We quantified variation in ATV infection over time and across the landscape. ATV infection was significantly higher in ponds modified for livestock use (P = 0.032). Disease incidence decreased with increased amounts of emergent vegetation (P < 0.001). These factors appear to control disease transmission by altering the host contact rate and with it disease transmission. 4. A field experiment to test the effect of emergent vegetation on the distribution of larvae in ponds demonstrated a behavioural change in larvae found in sparsely vegetated ponds. Microhabitat choices resulted in larvae being concentrated at the pond edge resulting in a 'halo effect' in sparsely vegetated ponds, whereas larvae in heavily vegetated ponds were distributed more evenly throughout. Microhabitat choice affects the effective density that larvae experience. This 'halo effect' increases contact rates in the shallows of sparsely vegetated ponds and increases the transmission of a directly transmitted pathogen. 5. Despite recurrent epidemics of a lethal Ranavirus in tiger salamanders on the Kaibab Plateau, Arizona, USA, these populations persist. We discuss the implications of our results in the context of density-dependent transmission and homogeneous mixing, two increases key assumptions of epidemiological theory.