MARINE ECOLOGY PROGRESS SERIES Vol. 157: 307-310,1997 Published October 16 Mar Ecol Prog Ser NOTE Quantifying hermit crab recruitment rates and megalopal shell selection on wave-swept shores S. E. Worcester', S. D. Gaines Marine Science Institute and Depl of Ecology, Evolution and Marine Biology, Universily of California. Santa Barbara, California 93106. USA ABSTRACT. The availability of shells has been shown to limit The few studies of hermit crab recruitment to date populations of adult hermit crabs and alter the outcome of either have simply surveyed the abundance of small interspecific interactions. However, few studies have juveniles (e.g. Lowery & Nelson 1988), have placed addressed such issues in newly settled crabs. A major imped- irnent has been the absence of techniques to quantify the use small shells out in the field at relatively calm sites of shells by settling crabs. Here we present a simple tech- (either buried in the sand, e.g. Asakura 1991, or in arti- nique for assessing hermit crab recruitment in rocky habitats. ficial tide pools, e.g. Nyblade 1974, J, Alio & A, Kuris The method appears to measure differences in settlement unpubl, data), or have quantified settlement in con- rates among sites without substantial bias. By offering several laboratory cOnditlons (Gherardi lgg6).Each of different s.~eciesof mlcroaastro~odd.shells, we also found that settling pLgurus hlrs~tl~sc~lushave dist~nctshell prefer- these techniques has important llmltatlons fol quantl- ences. Such shell preferences may lead to resource l~mitat~on fying hermit crab settlement, particularly on more at the juvenile stage which may have cascading effects on the energetic wave-exposed shores. Surveys of juvenile distribution, abundance, or shell use patterns of adult hermit abundance may not accurately reflect larval availabil- crabs. ity and settlement rates at each site if there are differ- KEY WORDS: Recruitment. Hermit crab . Pagurus hirsutlus- ences in the availability or species composition of small culus Shell limitation shells or if there is substantial mortality between set- tlement and the age at which recruits are easily cen- sused. Similarly, unconstrained introductions of shells Hermit crabs typically are selective in their use of cannot maintain a fixed conlposition and abundance of shells (e.g. Vance 1972, Bertness 1980, Hazlett 1981, shells except under extremely calm conditions. Here Mesce 1993, Gherardi et al. 1994, Elwood et a1 1995). we present a technique for monitoring hermit crab In part because of preferences for different shell spe- recruitment and for assessing the pattern of shell use cies, limitations in the abundance of those shells can by recently settled hermit crabs in even the most wave- greatly affect hermit crab populations (e.g. Bollay swept environments. 1964, Kellogg 1976, Scully 1979). Although there have Study sites and methods. We conducted these been many studies on hermit crab shell preference and studies at 10 sites along the California, USA, coastline shell limitation, these studies have focused almost in San Luis Obispo and Santa Barbara Counties exclusively on adult hermit crabs. Hermit crab mega- (34"23' N, 119" 42' W to 35O 13' N, 120°51' W) from lopae settling from the plankton also must find shells June 1995 to October 1996. The common hermit crab Yet few studies have quantified the patterns of hermit species which occur along this stretch of coastline crab settlement in the field. If small shells are limiting include Pagurus hii-sutiusculus (2 subspecies; Cl-ain & to settling hermit crabs, then patterns of distribution McLaughlin 1993), P. samuelis, and P. granosirnanus and abundance of hermit crabs may be established (Morris et al. 1980). The focus of these experiments long before the adult stage (e.g. see Asakura 1991). was on the recruitment of P. hirsutiusculus, because this species was by far the most common and had the greatest number of new recruits in the field (authors' 'Present address and address for correspondence: Institute for ~~~thSystems science and policy, California state uni. PerS. 0bs.I. The chosen sites varied greatly in the versitv Monterev. Bav,. Seaside, California 93955, USA. abundance of P, hirsutiusculus (Table 1). Adult P. hir- E-mail. suzanne-worcester@ionterey edu sutiusculus are most commonly found in shells from O lnter-Research 1997 Resale of full article not perm~tted 308 Mar Ecol Prog Ser 157: 307-310, 1997 Table 1 Pagurus hirsutiusculus. Mean number of hermit hermit crab recruits in 2 tidepools at each site every 3 crabs per square meter (* 95'1;, confidence limits) at each of to 4 mo. We collected a minimum of 75 to 100 crabs at the 10 field sites. We estimated hermit crab densities by counting the number of crdbs in 0.25 m' quadrats placed each site (except at Arroyo Hondo where crabs were every 2 m along transects run horizontally through the mid- too rare) to quantify the size frequency distributions of intertidal zone at each site. The number of quadrats sampled crabs. As part of a larger study, we sorted the crabs into is given in the third column 6 size classes based on shell diameter (<3, 3-6, 6-10, 10-15, 15-20, 20-25, and >25 mm), and separated I site Crab density No. of quadratr I them based on hermit crab species and occupied shell species. We quantified the overall density of hermit Alegria crabs at each site by running transects through the Arroyo Hondo Coal 011Point middle intertidal zone and counting all hermit crabs in D~abloCove 0.25 rn2 quadrats. The recruit density of crabs at a site Field's Cove was measured as the proportion of recruits (crabs in Jalama shells <6 mm) from the size frequency distributions Lompoc Landing North Vandenberg times the average density of crabs from the quadrats. Shell Beach Our initial focus was on trying to measure recruit- Shoreline Park ment rates of hermit crabs at different sites, so we hap- hazardly collected microgastropod shells to place in the bags. The preliminary settlement patterns, how- Olivella biplicata and Acanthina spirata (Morris et al. ever, suggested that megalopae may be quite selective 1980). in their use of different shells. The recruitment bags We tried a variety of techniques to assay the recruit- contained a relatively haphazard mixture of shell ment of hermit crabs to microgastropod shells that types. Bags contained primarily shells of (93 to 98%) failed before using the simple approach of placing Lacuna sp. and Littorina sp. (by far the most abundant shells in a mesh bag. The holes in the mesh need to be 2-3 mm size snails in these habitats; authors' pers. small enough to retain the shells used by settling crabs, obs.), but we also included some shells of much less but large enough to permit entrance by megalopae common snail species such as Tricolia sp., ALia sp., Epi- seeking shells. For the species of crabs and shells in tonium sp, and Homolapoma sp. (all species combined this system, polyester mesh (mesh diameter = 1.7 mm; 2 to 7 %). The 'commonness' of shells was assessed by polyester netting #9622, Research Nets Inc.) proved our difficulty in finding the snails in the field. After effective. To quantify hermit crab recruitment rates, many days of searching we were unable ever to collect we placed 50 small shells (- 2-3 mm diameter) in bags large numbers of shells of Tricolia sp., Alia sp., Epito- made of 30 X 30 cm squares of the polyester mesh. nium sp. or Homolapoma sp. (Homolapoma sp. in par- Each mesh square was folded to form a pouch, then ticular was extremely difficult to collect.) On the other securely closed with a cable tie and affixed to the sub- hand, thousands of Littorina sp. and Lacuna sp. could stratum with cable ties looped through eye bolts be collected with a sweep of a hand at most sites. screwed into plastic wall anchors in the rock. We Results. Pagurus hirsutiusculus recruited to mesh placed 10 mesh bags at each of the 10 sites to assess bags in densities ranging from 0 to 29 recruits per bag whether the bags collected hermit crab recruits. (mean + SD = 0.5 + 1.8 for all sites and sample dates). We replaced the mesh bags elther bi-weekly or Settlement density varied greatly over the year with monthly from June 1995 to October 1996. The number the highest recruitment occurring between May and of hermit crab recruits and the shell type occupied by November (Fig. l).The seasonal patterns for the site each crab in each bay was determined in the lab on the with the highest rate of recruitment (Field's Cove) is day the bags were collected in the field. We identified shown in Fig. 1. Almost no recruitment occurred in the all crabs to species and all occupied shells to genus. winter. The seasonal pattern of recruitment in the bags Settling hermit crabs within 2 molts of the megalopal mirrored the observations of recruits on the shore. stage were often difficult to distinguish at the species Small juvenile crabs were typically observed on the level. Therefore, we retained the youngest crabs in shore 1 to 2 mo following peaks of settlement to shells individual running seawater containers in the lab with in the bags. an abundant supply of larger shells until they had Settlement densities also varied substantially among molted into a stage where they were identifiable to sites. The range of recruit densities among sites species (usually after 2 molts). spanned more than 2 orders of magnitude, with high- To test whether recruitment into the mesh bags was est settlement at Field's Cove (max. 129 recruits per similar to 'natural' recruitment into unrestrained shells month) and lowest settlement at Jalama (max.