Mutual Avoidance in the Spectacled Salamander and Centipede: A Discrepancy between Exploratory Field and Laboratory Data

SIMPLE SUMMARY: Despite adult amphibians often being predators of arthropods, there are cases where the roles are reversed. This study focuses on the potential predator–prey relationship between the spectacled salamander (Salamandrina perspicillata) and the centipede Scolopendra cingulata in Central Italy. In a natural site, we observed a strong avoidance pattern (negative co-occurrence) of the two study species under shelters (stones). Given the importance of olfactory cues in the salamander’s behavioural choices, we hypothesized that this species would actively avoid shelters used by the centipede so as to evade potential attacks and predation. However, when we forced the study species to choose between sharing or not sharing a given shelter in a laboratory experiment, the avoidance pattern was not confirmed. Our exploratory results show that what often appears to be a strong observation-based pattern in natural settings needs to be carefully evaluated, and experiments in controlled environments could help exclude potentially misleading processes. ABSTRACT: Interactions between amphibians and arthropods encompass a wide range of ecological relationships, predominantly characterized by predator–prey dynamics, with adult amphibians as the predators. In some instances, the roles are reversed. This study focuses on the potential predator-prey relationship between the spectacled salamander (Salamandrina perspicillata) and the centipede Scolopendra cingulata in Central Italy. Building upon previous research on chemical cue perception in amphibians, we investigated potential olfactory cue-mediated avoidance behaviours exhibited by S. perspicillata towards the potential predator S. cingulata through field observations and manipulative experiments. In a natural site, we estimated the degree of negative co-occurrence between the study species under shelters and found an avoidance pattern between S. perspicillata and S. cingulata in refuges. However, when the study species were forced to choose between sharing or not sharing a given shelter, through a manipulative experiment, the avoidance pattern was not confirmed. Potential determinants contributing to the avoidance pattern observed in nature are discussed. Our exploratory results represent a good example of how what often appears to be a strong observation-based pattern in natural settings needs to be carefully scrutinized. Hypotheses testing through experiments in controlled environments remains a valuable approach to exclude potentially misleading processes.