Broken pieces: Can variable ecological interactions be deduced from the remains of crab attacks on bivalve shells?

Shell fragmentation patterns that result from attacks by durophagous predators on hard-shelled marine invertebrates are a rich source of indirect evidence that have proved useful in interpreting predation pressure in the fossil record and recent ecology. The behaviour and effectiveness of predators are known to be variable with respect to prey size. It is less well understood if variable predator–prey interactions are reflected in shell fragmentation patterns. Therefore, we conducted experimental trials to test the behavioural response of a living crab, Carcinus maenas, during successful predatory attacks on the blue mussel Mytilus edulis on two prey size categories. Further, we examined resultant shell fragments to determine whether specific attack behaviours by C. maenas could be successfully deduced from remaining mussel shells. In contrast to previous studies, we observed no significant differences in attack behaviour by the predators attributable to prey size. In most experimental predation events, crabs employed an ad hoc combination of five mechanisms of predation previously described for this species. We identified seven categories of shell breakage in predated mussels, but none of these were unambiguously correlated with specific attack behaviour. Combined attack behaviours may produce shell breakage patterns that have previously been assumed to be attributable to a single behaviour. While specific patterns of shell breakage are clearly attributable to durophagy, the results of this study provide important insights into the limitations of indirect evidence to interpret ecological interactions.