To meet the assumption of normality, data for latency to visit GNE0877 patch and number of Cheerios eaten were ln transformed. All analysed data were evaluated using Levene's test for equality of variances and conformed to the assumptions for conducting parametric analyses. Body weight data were analysed using a repeated measures analysis of variance (RMANOVA) with stress condition and time as fixed effects. Data from the three motivation tests were also analysed with a RMANOVA with stress condition and time as fixed effects. Data from the three foraging tests (latency to visit a patch, switches between patches, number of Cheerios consumed) were analysed with RMANOVAs with stress condition and time as fixed effects. If time or interaction effects were detected in the RMANOVA, each time point was analysed individually using univariate ANOVA. One data point from the control group was omitted from the latency to visit a patch analysis because polyp was greater than 12 standard deviations away from the mean. Analyses were run using IBM® SPSS® Statistics (Version 21); values are reported as means ± SE. Partial eta-squared values, expressing the proportion of variance explained, are provided as effect size estimates.