Mortality
We found considerable diet-dependent mortality during the experiment. Specifically, there was a significant interaction between the average amount of food eaten and diet type in their effect on mortality during the 3-week feeding period (F1,488 = 4.8509; p = 0.0281). Mortality decreased with average mass of food eaten in both treatments, but mortality was higher in males held on the high protein diet, at all levels of food consumed, except at lower consumption (Fig. 2A). Additionally, both diet (F1,387 = 13.514; p< 0.001) and infection cue (F4,387 = 14.427;p < 0.001) significantly affected survival through the 2-day calling trials; in this case, the interaction term was not significant (F4,382 = 1.032; p = 0.3905) and was not included in the final model. Males held on the high protein diet were the least likely to survive through the end of the 2-day calling trial. Controlling for diet, males receiving the highest dose infection cue also were the least likely to survive the 2-day calling trial (Fig. 2B). Males that died prior to the conclusion of the calling trials were eliminated from further analyses of reproductive effort, immune function, or survival.