Repeatability of foraging behavior following a simulated predation attempt depends on color morph, sex, and foraging metric in Red-backed salamanders (Plethodon cinereus)

Abstract

Behavioral repeatability greatly affects the capacity of an individual to respond to varying environments. When multiple behaviors within individuals are repeatable and correlated across time or across contexts, it is termed a behavioral syndrome. However, not all behaviors exhibit the same level of repeatability, and relatively few studies have examined repeatability in amphibians. We examined the repeatability of foraging behavior in the Eastern Red-backed salamander (Plethodon cinereus), a color-polymorphic terrestrial salamander, following a simulated predation attempt. We tested several hypotheses: (1) Simulated predation would negatively affect foraging, increasing latency to feed and decreasing the number of prey items eaten in a fixed time period compared to a control group; (2) Because striped color morphs of P. cinereus are more aggressive, striped individuals would exhibit “bold” behavior by resuming foraging sooner and consuming more prey; and (3) Foraging behavior would be more repeatable for males. We found that the predation treatment inhibited foraging behavior, although neither morphs nor sexes differed in either forging metric. The number of prey eaten was repeatable for all groups of salamanders. Latency to feed, however, was not repeatable for control salamanders. Simulated predation induced repeatable latencies, but when morphs and sexes were analyzed separately, only unstriped and male salamanders were repeatable, suggesting characteristics of these groups related to behavioral syndromes drive this response. We speculate that the greater repeatability of the unstriped morph’s latency to feed may result from more frequent encounters with predators in the leaf litter matrix while foraging. Striped salamanders from the source population, in turn, exhibit greater territorial success, and thus may experience more variation in encounters with predators and conspecifics over the course of their lifespans. Our results illustrate the need to carefully define the behavior and subset of the population to be tested when studying behavioral repeatability or behavioral syndromes.