In theory, the manner in which pathogens are transmitted will have very different effects on host populations. At one extreme, a host species will not go extinct if the pathogen is transmitted in a density dependent fashion, as is the case for influenza in humans. But if transmission is independent of density, like many sexually transmitted diseases, or the pathogen has a reservoir, as for avian malaria in Hawaii, extinction may result in the absence of intervention strategies to alter transmission, virulence, or access to all species' members. There are few empirical tests of these generalizations despite the importance of understanding emerging infectious diseases and how pathogens affect species richness and diversity.

An Ambystoma tigrinum–Ambystoma tigrinum virus (ATV) model system was used in the field and laboratory to test basic assumptions of host-pathogen theory related to disease transmission dynamics. The results indicate that the current diagnostic test for ATV infections using non-lethal sampling techniques underestimates the disease's true prevalence in wild populations. Field habitats showed more fragmentation when vegetation was sparse than dense, and ATV transmission was higher in sparsely vegetated ponds because individual contact rates increased. ATV transmission in larvae also saturated at high densities, resulting in a non-linear transmission rate. In some years epidemics occurred in habitats widely separated across the landscape. Environmental stochasticity, or a Moran effect, in the form of winter precipitation and not dispersal of infected individuals synchronized these epidemics.

Empirical tests of disease models at the population level can identify populations with transmission dynamics putting them at risk for pathogen-induced extinction. For example, disease is one likely cause of late twentieth century global amphibian declines. Theory predicts that populations with density dependent disease transmission are not at risk of pathogen induced extinction. Consistent with this expectation, our results demonstrated no evidence of extinction. In contrast, the most likely explanation for the enigmatic decline of some amphibian populations infected with the chytrid fungus, Batrachochytrium dendrobatidis, may be disease transmission that is not density dependent. These results highlight the importance of including host-pathogen interactions as an element of basic theoretical and empirical research in ecology and conservation.