Desired domiciles: shell selection by two species of hermit crabs from a temperate rocky intertidal in Washington state, US

Abstract

While shells are key resources for hermit crabs, there is great variation among the parameters of different species of shells, and hermit crabs often show species- and even site-specific preferences for particular shells. Two species of hermit crabs, Pagurus granosimanus (Stimpson, 1859) and Pagurus beringanus (Benedict, 1892), apparently have increased their vertical distribution overlap in the intertidal regions of San Juan Island, Washington state, US compared with studies from 50 years ago. Behavioral plasticity, driven by increased direct and indirect competitive interactions, may have altered shell preference in either or both of these species. We investigated whether the degree of shell preference overlap displayed by the two species of hermit crabs had altered compared with previous findings, and also whether any preferences changed depending on the wave exposure and density of crushing predators at the site of collection, the size class of the hermit crabs, or the previous shell inhabited by the hermit crab. We also examined whether the weight or length of the shell determined choice, from among the species of shells provided as choices (our design did not experimentally isolate shell characteristics from species). In the laboratory we simultaneously provided four of the most common types of sympatric shells to individual hermit crabs. Pagurus beringanus (Benedict, 1892) tended to choose shells of Nucella lamellosa (Gmelin, 1791) or Lirabuccinum dirum (Reeve, 1846), whereas P. granosimanus typically did not have a marked preference for certain species of shells, but often avoided using shells of Calliostoma ligatum (A. Gould, 1849). Neither shell length nor previous-domicile significantly predicted shell choice for either hermit crab species. Shell weight did not determine shell choice for P. granosimanus, and where it was a significant factor for P. beringanus, that result may have been driven by preference for shell species. The recently-reported distribution of hermit crab-inhabited domiciles in the field did not match the rankings of utilized shells reported herein. Appropriately-sized, quality empty shells appear to be limiting in this system. The overlap of distribution of these two hermit crabs in the low intertidal and their similar range of shells used may cause the potential for interspecific competition, but the lack of shell specialization by P. granosimanus and preference of P. beringanus for a different species of shell (N. lamellosa) than that often used most by P. granosimanus (L. dirum) likely reduces interspecific competition. Our results largely concur with results reported as long as 50 years ago, despite the likelihood of environmental changes in the intervening time.