The macrofauna associated with individually collected clumps of pelagic Sargassurn was influenced strongly by the location of sampling. Algae collected in the Sargasso Sea carried more species with lower dominance and higher evenness and variation in abundance than algae collected in the Gulf Stream. Gulf Stream samples were dominated by the shrimp Latreutes fucorum (67.8 %) with little compositional variation whereas the gastropod Litiopa melanostoma was the most abundant but weakly dominant species (25.1 %) in the Sargasso Sea. Generally, species endemic to pelagic Sargassum species were most abundant in the Sargasso Sea while numerous temporary associates were found in the Gulf Stream. Numbers of individuals and species increased with the weight of algal clumps but clump size had little effect on the species composition of algal associates. Variation in species composition and trophic organization of macrofaunal associates with geographic region and with algal age were related to differences in the epiphytization of the algae. Because of the dissimilarity in algal associates of the Gulf Stream and Sargasso Sea, the influx of littoral fauna from the West Indies to the Sargasso Sea is probably very slow and the present forms may have evolved in that water mass.

Laboratory experiments showed that the three epifaunal amphipods Cymadusa cornpta, Grandidierella bonnieroides, and Melita elongata, normally associated with seagrasses, are capable of detecting small differences in the density of seagrass and actively select areas of high blade density. When different seagrass species -Thalassia testudinurn, Syringodiurn filiforme, and Halodule wrightii -were offered with equal blade biomass, H, wrightii was selected due to its high ratio surface area: biomass. When offered equal surface area all three amphipod species were randomly distributed among the seagrass species, confirming the hypothesis that surface area is the habitat characteristic chosen. It was concluded that blade surface area of macrophytes provides the best estimate of habitat complexity in seagrass meadows. The surface area relationship also held for C. compta when the highly branched red alga Gracilaria foliifera was used as a substratum. Degree of habitat selectivity by the three species of amphipods was related to known differences in vulnerability to predation by Lagodon rhornboides. It is concluded that habitat preference and biological interactions must both be examined to fully understand distributional patterns in seagrass epifauna.

GrowthABSTRACT: The roles of nutrition, predation, and habitat choice were tested as mechanisms for shaping natural distribution of newly settled queen conch in seagrass Thalassia testudinum meadows in the central Bahamas. Small animals from 3 size classes were enclosed at 2 sites in 3 different habitats across a seagrass gradient (bare sand, low-density seagrass, medium-density seagrass) for 3 wk. Medium (11 mm shell length) and large (22 mm) newly settled conch grew faster in both seagrass habitats than on bare sand, but differences were not significant for small (5 mm) conch. Small and medium conch grown on sand, however, weighed more than those of the same length grown in seagrass. In laboratory habitat-choice experiments, newly settled conch preferred living seagrass and seagrass detritus over bare sand. Medium conch tethered in the 3 habitats at both sites suffered very high mortality (50 to 96% killed in 11 d), with those tethered in medium-density seagrass showing an advantage. Juvenile conch over 1 yr old are primarily associated with medium-density seagrass, but younger animals have been observed on sand. Results from this study show that nutrition is adequate on sand for newly settled conch, but an ontogenetic shift from sand into seagrass is unlikely because predation occurs before food becomes limiting or a habitat choice can be made. Nevertheless, post-settlement mortality is high regardless of where conch settle in the seagrass gradient, and predation is the most important mechanism influencing distribution.

The relatlve roles of habltat quallty and predation were examined to test the hypothes~s that surv~vorshlp In a manne gastropod, Strombus glgas, IS enhanced by livlng In a g g r e g a t~o n s when young Enclosure and tether experiments were conducted In and around a juven~le aggregatlon located In the m~d d l e of a tidal flow field in the Bahamas whlch occupied ca 21 ha Elght s t a t~o n s were selected to represent a gradient from bare sand to h~g h seagrass blomass and a transect through the aggregatlon (~n s i d e density = 0 40 to 0 56 )uveniles m-' outside denslty SO 08 )uveniles m-2) Seagrass characteristics proved to be poor pred~ctors of habltat sultab~lity in terms of S glgas growth and compared w~t h the presence of conspecifics relatively unimportant predictors In terms of shelter from predation Although growth was > 0 1 mm d.' at all but 2 stations, juveniles (60 to 80 mm shell length) suffered higher proport~onal mortal~ty when tethered outside the aggregatlon (48 to 100%) than inside (18 to 20%) Mortal~ty showed a strong negatlve correlat~on w~t h wild S glgas density, suggesting that small-scale distnbut~on is greatly influenced by denslty-dependent phenomena Presence of conspecifics proved to be a n important factor affect~ng survivorshlp In seagrass meadows of the Bahamas KEY WORDS. Strombus gigas. Habitat quality Predat~on . Seagrass . T e t h e r~n g . Bahamas

Abundance and species richness of crustaceans associated with the calcareous green alga Penicillus capitatus in Puerto Rico increased with algal size. Faunal abundance was higher in the dry season than in the wet season, but species-area regressions were identical for the 2 periods. Faunal abundance rather than algal size, however, proved to be the best predictor of crustacean species richness on P. capitatus. Although there were no negative associations among species pairs, qualitative and quantitative changes in the crustacean assemblages during recolonization of defaunated P. capitatus and increasing dominance with animal density suggest the potential significance of differential immigration rates, displacement of certain taxa, and positive associations in explaining variation in species composition and species richness. Higher abundances of crustaceans (particularly amphipods and tanaidaceans) were found on P. capitatus than in the surrounding seagrass (Halodule wrightii) habitat; this is probably a function of protection from fish predators and habitat selectivity.