Costly learning: preference for familiar food persists despite negative impact on survival.

Animals often rely on events in their environment that provide information (i.e. experience) to alter their future decision-making in ways that are presumed to be beneficial. Such experience-based learning, however, does not always lead to adaptive decision-making. In this study, we use the omnivorous harvestman Heteromitobates discolor to explore the role of past diet on subsequent food choice and survival. We first tested whether a short-term homogeneous diet (rotten crickets, fresh crickets or dog food) influenced subsequent food choice (rotten cricket versus fresh cricket). We next examine the impact of diet on survival. We found that following experience with a homogeneous cricket diet, adult harvestmen displayed a learned preference for familiar food, regardless of whether it was rotten or fresh crickets; individuals experiencing dog food were equally likely to choose rotten versus fresh crickets. We additionally found that individuals that ate rotten crickets suffered shorter survival than those that ate fresh crickets. Together, our results suggest that the diet an individual experiences can lead to maladaptive food preferences-preferences that ultimately result in reduced longevity.

Costs and benefits of freezing behaviour in the harvestman Eumesosoma roeweri (Arachnida, Opiliones).

Animals present an enormous variety of behavioural defensive mechanisms, which increase their survival, but often at a cost. Several animal taxa reduce their chances of being detected and/or recognized as prey items by freezing (remaining completely motionless) in the presence of a predator. We studied costs and benefits of freezing in immature Eumesosoma roeweri (Opiliones, Sclerosomatidae). Preliminary observations showed that these individuals often freeze in the presence of the syntopic predatory spider Schizocosa ocreata (Araneae, Lycosidae). We verified that harvestmen paired with predators spent more time freezing than when alone or when paired with a conspecific. Then, we determined that predator chemical cues alone did not elicit freezing behaviour. Next, we examined predator behaviour towards moving/non-moving prey and found that spiders attacked moving prey significantly more, suggesting an advantage of freezing in the presence of a predator. Finally, as measure of the foraging costs of freezing, we found that individuals paired with a predator for 2h gained significantly less weight than individuals paired with a conspecific or left alone. Taken together, our results suggest that freezing may protect E. roeweri harvestmen from predatory attacks by wolf spiders, but at the cost of reduced food and/or water intake.