Giving-up density (GUD) experiments are a common technique used to evaluate perceived predation risk by prey species. These experiments use food consumption as a proxy for predation risk, assuming greater amounts of food will be left unconsumed as risk perception increases. Implicit to this experimental design is the assumption that the perceived value of a food item does not change across levels of predation risk, so that differences in consumption are a direct function of perceived risk. However, recent studies have shown that in some species, such as grasshoppers, predation risk elevates metabolic rate and causes prey to shift dietary preferences toward carbohydrate-rich (C-rich) foods and away from protein-rich (P-rich) foods. This suggests that the interpretation of GUD experiments may be confounded by perceived food values that change across levels of predation risk. We conducted a series of lab experiments to test the hypothesis that carbohydrates and proteins change in relative importance to prey when facing predation risk, and consequently the GUD framework may over- or under-estimate predation risk depending on the nutrient value of the food used. We presented grasshoppers with two artificial diets (C-rich or P-rich) in a choice or no-choice factorial design crossed with predation risk from a spider predator. Next we conducted a GUD experiment with standardized levels of risk by using enclosures stocked with a known number of spiders (0-5) to test our hypothesis that food nutrient preference can covary with predation risk.
As predicted, the presence of spiders in our choice and no choice experiment reduced consumption of the P-rich food and increased consumption of the C-rich food. In our giving-up density experiment, the GUD of P-rich food increased with spider density, but was invariant when using C-rich food. These results suggest that GUD experiments using a C-rich food may underestimate predation risk, whereas using a P-rich food may overestimate predation risk.