Host plants and experimental warming impact fitness and infection outcomes in a migratory butterfly

Climate warming can impact host and parasite traits directly, but higher temperatures also affect parasitism indirectly through species interactions like herbivory. Given that temperature can influence plant quality, which in turn shapes nutrition and immune defense of insect herbivores, understanding plant-mediated effects of warming on insect-parasite interactions could inform insect conservation and pest management.  We examined how warming and host plant species influenced parasitism of monarch butterfly (Danaus plexippus) larvae by the protozoan parasite Ophryocystis elektroscirrha (OE) in a field cage experiment. Cardenolides, defensive chemicals produced by milkweed host plants (Asclepias spp.), increase monarch tolerance and resistance to OE. We reared parasite-exposed and uninfected control monarch larvae on tropical (high-cardenolide) or swamp (low-cardenolide) milkweed in either ambient or elevated temperatures, measuring monarch development, survival, and size, and OE infection success and intensity. To assess whether temperature effects on infection or monarch traits are influenced by altered milkweed nutrition or toxicity, we measured foliar carbon, nitrogen, and cardenolides. We found that host plant species was an important predictor of monarch development and survival, and that elevated temperatures reduced monarch tolerance to infection even when reared on high-cardenolide tropical milkweed. We also found that milkweed leaves had more nitrogen relative to carbon in temperature-elevated plots, indicating that despite improved host plant nutrition, monarchs experienced lower fitness and higher costs of infection. These findings will help predict consequences of warming for monarch health across the insect’s range, especially for southerly latitudes where tropical milkweed supports monarch residency and high OE infection risk.

Methods
We reared parasite-exposed and uninfected control monarch larvae on tropical (high-cardenolide) or swamp (low-cardenolide) milkweed in either ambient or elevated temperatures in field plots. We measured monarch development, survival, and size, and OE infection success and intensity. To assess whether temperature effects on infection or monarch traits are influenced by altered milkweed nutrition or toxicity, we measured foliar carbon, nitrogen, and cardenolides. 

We tested for the effects of plot temperature (ambient/elevated), milkweed species (swamp/tropical), and infection (inoculated/uninoculated) treatments on monarch fitness and parasite infection outcomes using linear and generalized linear models (GLMs). For each response variable, we assessed a small set of candidate models: an uninformative null, a model containing main effects, and models containing interaction terms of interest.