Epidemiologists increasingly realize that species interactions (e.g. selective predation) can determine when epidemics start and end. We hypothesize here that resource quality can also strongly influence disease dynamics: epidemics can be inhibited when resource quality for hosts is too poor and too good. In three lakes, resource quality for the zooplankton host (Daphnia dentifera) was poor when fungal epidemics (Metschnikowia bicuspidata) commenced and increased as epidemics waned. Experiments using variation in algal food showed that resource quality had conflicting effects on underlying epidemiology: high-quality food induced large production of infective propagules (spores) and high birth rate but also reduced transmission. A model then illustrated how these underlying correlations can inhibit the start of epidemics (when spore production/birth rate are too low) but also catalyse their end (when transmission becomes too low). This resource quality mechanism is likely to interface with other ones controlling disease dynamics and warrants closer evaluation.

Some aspects of habitat seem to enhance the spread of disease whereas others inhibit it. Here, we illustrate and identify mechanisms that connect habitat to epidemiology using a case study of disease in plankton. We see a pronounced relationship between the basin shapes of lakes and fungal (Metschnikowia bicuspidata) disease in the zooplankton grazer Daphnia dentifera. As we work through seven mechanisms that could explain why Daphnia in some lakes are sicker, we can eliminate some hypotheses (i.e., those relating an index of lake productivity to disease through host density, links between resource quality and transmission rate, and variation in host susceptibility) and find support for others involving food-web actors (e.g., selective predation on infected hosts by fishes, "sloppy predation" by an invertebrate, a possible dilution effect in V-shaped lakes). Furthermore, we identify physical mechanisms (gravity currents, turbulence) that could lead to greater transport of fungal spores to habitat occupied by Daphnia hosts in U-shaped lakes. These results highlight how habitat structure, through its effects on food-web structure and physical processes, can shape wildlife disease.

The "healthy herds" hypothesis suggests that selective predators, by acting as parasite sinks, may inhibit the start of epidemics and reduce prevalence of infection. Here, we describe a counter-example using field patterns, experiments, and a model. The predator Chaoborus releases infective spores of a fungal parasite and, in doing so, may facilitate epidemics in Daphnia populations. In the field, epidemics occur in lakes with higher Chaoborus densities. Experiments revealed that nonselective Chaoborus release many of the spores contained in their prey. Since these released spores remain infective, this predator can catalyze epidemics when a lake's physical environment might otherwise impede them. Without Chaoborus, Daphnia dying of infection may sink to the lake bottom before releasing spores. A model tracking hosts and spores in the water column (where hosts contact spores) and in bottom sediments (where they cannot) illustrates this mechanism. Thus, by dispersing spores while feeding, this predator spreads disease. Many invertebrates are parasitized by obligately killing parasites, offering a variety of systems for additional tests of this "predator-spreader" hypothesis. In the meantime, this planktonic disease system prompts a very important, general warning: before we use predators to keep the herds healthy, we need to carefully think about the interface between predator feeding biology and the underlying epidemiology of wildlife disease.

No abstract