MEPS 537:49-58 (2015)  -  DOI: https://doi.org/10.3354/meps11419

ABSTRACT: Predators non-consumptively induce prey habitat shifts, driving trait-mediated indirect interactions (TMIIs) with basal resources. Whether prey seek refuge within or avoid predator-containing patches determines the spatial re-distribution of prey and influences the nature of resulting TMIIs. In a southeastern US salt marsh, we tested how 2 species of sit-and-wait benthic predatory crab non-consumptively affect the habitat choices of grazing snails and how the resultant behavior affects cordgrass health. We first observed that snails climb higher on cordgrass around naturally occurring crab burrows and that this habitat shift corresponds with increased cordgrass leaf damage, suggesting a localized TMII. Then, by adding caged crabs to artificial burrows in the field, we found causative evidence that both crab species could drive snails upwards to the cordgrass canopy, thereby increasing leaf damage within a ~12 cm radius, but found no evidence that experimentally added crabs induce horizontal dispersal of snails. Next, in a marsh undergoing die-off, we added caged crabs to remnant cordgrass patches being rapidly colonized by snails. Crabs did not affect the rate of snail colonization of remnant patches, but did drive snails upwards once they had colonized a patch, suggesting snails do not preferentially avoid—but do alter how they locally utilize—patches of cordgrass in marsh die-off contexts. Our documentation of the spatial-scale and dimensionality of TMIIs in the field paves the way for spatially explicit models of this interaction. More generally, our results suggest that TMIIs may be consistent, predictable and tractable, lending themselves to incorporation into food-web models.

KEY WORDS: Littoraria · Marsh periwinkle · Non-consumptive effect · Salt marsh · Spartina alterniflora · Trait-mediated indirect effect · Trophic cascade