Marine species possess dispersive stages that interconnect subpopulations, which may inhabit ‘source’ and ‘sink’ habitats, where reproduction and emigration either exceed or fall short of mortality and immigration, respectively. Postlarval supply, juvenile density and adult abundance of the Caribbean spiny lobster, Panulirus argus, were measured at four widely separated sites spanning >100 km in Exuma Sound, Bahamas. Adult abundance was lowest at a site with the highest postlarval supply and little nursery habitat; hence, it was tentatively classified as a sink. Circulation in Exuma Sound is dominated by large-scale gyres which apparently concentrate and advect postlarvae toward the nominal sink. The remaining three sites, including one marine reserve, had higher adult abundances despite lower postlarval supply, and are therefore tentatively classified as sources. Postlarval supply is probably decoupled from adult abundance by physical transport. Adult abundance is likely decoupled from postlarval supply by the effects of varying habitat quality upon postlarval and juvenile survival, as indicated by non-significant differences among sites in juvenile density. It appears that some sites with suitable settlement and nursery habitat are sources of spawning stock for Panulirus argus, whereas others with poor habitat are sinks despite sufficient postlarval influx.

In general, hypoxia (< 2 mg O 2 l -1) is detrimental to marine food webs because of faunal declines associated with persistent, severely low oxygen. However, transfer of benthic production to higher trophic levels could be facilitated under hypoxia if infauna migrate to shallower burial depths, increasing their availability to predators. A series of outdoor mesocosm and laboratory experiments were conducted in 3 years (1999 to 2001) to quantify the effects of hypoxia upon (1) predation by the blue crab Callinectes sapidus on the Baltic clam Macoma balthica, a biomass dominant in Chesapeake Bay, and (2) ) either without or with a male blue crab that was allowed 2 d to feed upon the clams. Predation-induced proportional mortality of clams was significantly lower under low DO than under normoxia in all 3 years. Thus, under short-term hypoxia, both crab feeding efficiency and trophic transfer from M. balthica to blue crabs were reduced. Changes in clam burial depth due to oxygen levels was determined by establishing normoxic and low DO treatments in replicate 208 l aquaria in 2 years. Burial depth after exposure to oxygen treatments for 48 h did not differ as a function of oxygen level. None of the clams died after 2 d of low DO, 27% died after ~6 d, and 90% died after 21 d. Short-lived hypoxia therefore reduces the ability of crabs to forage upon clams efficiently and increases clam survival, whereas long-term hypoxia may increase the availability of clam prey to predators through mortality and movement to the surface. Thus, shortlived hypoxia is likely to reduce the transfer of benthic prey to higher trophic levels, although longerterm exposure may increase transfer.

An aggregation of juvenile queen conch Strombus gigas vath densities exceeding 250 ind. m-' was observed in the central Bahama Islands during April 1987. The newly documented wave-like aggregation comprised over 100 000 individuals of a single year class averaging 101 mm in total shell length. The aggregation moved an average of 4.8 m d-', en masse, in the direcbon of the ebb tidal currents. Observations between May and August 1987 showed that the mass migration was long-lived. We hypothesize the mass migration to be seasonally synchronous and to serve as a density-dependent or habitat-dependent dispersal mechanism for newly emerged l + year classes from centers of larval recruitment.