Figure legends
Figure 1: Diet consumption by male G. sigillatusfed either of two diet types (P:C = 5:1 and 1:8). (A) The amount of food eaten (dry weight of trays containing diet both before and after offering) across tray number (i.e., feeding period time). (B) The average amount of diet eaten during the feeding trial between diet types and based on pronotum width, a proxy of structural body size. Feeding periods were 3 weeks long and trays were switched every 3 days for each male.
Figure 2: Mortality of male G. sigillatus during the 3-week feeding trial (A) and subsequent 2-day calling period (B) based on diet (P:C = 5:1 or 1:8) and infection cue treatments.
Figure 3: MCMCglmm zero-altered Poisson (ZAP) analysis of the likelihood of calling (a logistic regression for zeros) and time spent calling (zero-truncated Poisson) in male crickets (G. sigillatus ) fed either of two diet types and administered an infection cue from a gradient of increasing intensity using heat-killed bacteria. Diet effects are shown as the deviation from the P:C 1:8 diet treatment and overall infection effects are shown as differences from the naive level.
Figure 4: Model predicted effects of diet type and infection cue (from a gradient of increasing intensity using heat-killed bacteria) on the (A) likelihood of calling and (B) time spent calling (given that a male produced a call) in G. sigillatus . Points show predicted effects with 95% confidence intervals, taken from MCMCglmm ZAP analysis.