The porcelain crabs Petrol~sthes cinctipes (Randall) and P. eriomerus Stimpson are sympatric on exposed rocky shores in the northeastern Pacific These species often exhibit a striking pattern of non-overlapping vert~cal distributions, wlth P. eriomerus inhabiting low Intertidal and subtidal areas and P. cinctipes occurring in the middle and high intert~dal. Transplant experiments revealed that P. enomerus was unable to tolerate displacement as l~ttle as 0 3 m above ~t s normal limit, while in contrast its congener showed good survival and growth when caged at lower levels. The 2 species were equally susceptible to desiccat~on; however, P. enomerus was extremely sens~tive to thermal stress during periods of aerial exposure, and thermal tolerance varied inversely with size. Abiot~c factors also control the d~stnbution of P. cinctipes; t h~s species was competitively dominant in laboratory experiments, suggesting the lower limit is not due to ~nteractions w~t h its congener. Rather, this d~s t n b u t~o n is l~n k e d to sudden changes in substrate composition, involving an actlve avo~dance of rocks resting on sand or s~l t as demonstrated through f~e l d manipulation of substrate beneath rock patches. Patterns of resource (space) p a r t~t i o n~n g beneath rocks are largely a passive consequence of phys~olog~cal constraints and substrate preferences by adults, maintamed In part through gregarious settlement by megalopae of both species

We quantified recruitment of young-of-the-year (YOY) burrowing thalassinidean shrimp Neotrypaea californiensis to 2 habitats of differing structural complexity-epibenthic bivalve shell and bare mudflat-and examined how differential settlement and post-settlement predation influence patterns of YOY abundance. Local densities of shrimp were quantified prlor to construction of shell habitat in Grays Harbor estuary, Washington (USA). Subsequent recruitment of YOY shrimp to epibenthic shell and bare mudflat was measured during a peak settlement pulse and 10 mo post-settlement. In addition, patches of sediment overlying shell within the shell plot ('subsurface shell') were sampled 10 mo post-settlement. Differential settlement in shell and mud habitats was quantified in field and laboratory experiments. We also examined predator-prey interactions between YOY Dungeness crabs Cancer magister and newly settled shrimp in shell habitat in a laboratory experiment in which prey consumption crab-' was quantified as a function of shrimp density. Results of our studies indicate that dense coverage of epibenthic shell applied to the intertidal site reduced recruitment of ghost shrimp. Epibenthic shell habitat had significantly fewer YOY shrimp than bare mudflat at peak settlement and 10 mo post-settlement, and significantly fewer shrimp than 'subsurface shell' at 10 mo post-settlement. Successful colonization of 'subsurface shell' suggests that shnmp postlarvae settled preferentially in areas of the shell plot covered with mud or possibly were exposed to lower levels of predation than in contiguous epibenthic shell areas. Results of the field experiment revealed that 2 to 5 times fewer shrimp postlarvae settled in shell than mud treatments; a similar but non-significant trend of lower settlement in shell than mud substrate was observed in the laboratory habitat-choice experiment. YOY Dungeness crabs preyed on shrlmp in the laboratory experiment; prey density had a significant effect on consumption I-ates but not on proportional shrimp mortalities. In sum, although other processes most likely contributed to patterns of YOY shrimp distribution, postlarval habitat selection for mud substrate was a key determinant of recruitment success. Recruitment patterns may be further modified by postsettlement mortality of YOY shrimp in shell due to YOY Dungeness crab predation.