THE RELEVANCE OF PASSIVE DISPERSAL FOR THE BIOGEOGRAPHY OF CARIBBEAN MOLLUSKS

Occurrence of teleplanic veliger larvae of sublittoral species in majo r currents of the tropical Atlantic Ocean supports the hypothesis that passive dispersal of larvae contributes importantly to the biogeograp hy of shoal-water molluscan species. Rafting may play a secondary role . Criticisms of passive larval dispersal as a factor in molluscan biog eography are: (1) that larvae have a fixed life span too short to acco unt for long-distance dispersal: both laboratory and field data show t his assertation to be mistaken; (2) that larvae after a time lose thei r competence to metamorphose: whenever tested, teleplanic larvae have been shown to retain their ability to metamorphose; (3) that dispersal is random and cannot explain congruent nonrandom associations among w idely differing taxa: dispersal is largely accounted for by advection along major ocean currents which provide quasi-permanent or seasonally reoccurring corridors for the transport of planktonic larvae; therefo re various taxa necessarily are dispersed over similar routes, leading to congruent geographic distributions; and (4) that some species lack ing long planktonic larval stages nevertheless have wide geographic ra nges and consequently there is little relationship between mode of rep roduction and geographic range of species: those species lacking exten ded planktonic larval stages yet having wide geographic ranges are mos tly sessile epibenthic forms that can attach to hard substrata, and co nsequently are preadapted for passive dispersal by rafting. Dispersal by human agencies also has contributed within historic time to the geo graphic distribution of marine mollusks. Such alternative modes of dis persal do not negate the importance of larval transport. Despite demon strable evidence for widespread passive larval dispersal and in some i nstances also for the transport by rafting of molluscan species, ecolo gical constraints place restrictions upon where and when new colonists can survive and reproduce. Geotectonics and sea-floor spreading have been significant in controlling the pattern and extent of passive disp ersal over geologic time. The closing of the Tethys Seaway during the Oligocene and Early Miocene resulted in the isolation of the southwest ern tropical Pacific Ocean from the Mediterranean Sea and tropical eas tern Atlantic Ocean. Similarly, the closing of the corridor between No rth and South America divided the tropical eastern Pacific Ocean from the Caribbean Sea. Continental drift and sea-floor spreading resulted in the initial formation and subsequent enlargement of the Atlantic Ba sin and has led subsequently to the ''mid-Atlantic barrier'' which tod ay acts as a filter between the tropical eastern Atlantic Ocean and Ca ribbean Sea. Both tectonic events and sea-floor spreading have placed important constraints on the dispersal between tropical faunas. Faunal studies of the amphi-Atlantic distributions of marine benthic mollusk s further support the hypothesis of passive dispersal. Low endemism on oceanic islands suggests that initial colonization must be largely ac complished by teleplanic larvae. Available evidence shows that althoug h no single process can completely explain the present composition of the Caribbean molluscan fauna, passive dispersal of planktonic veliger larvae must have played and continues to play an important role both in colonization and in maintaining genetic continuity between widely d isjunct regions of the tropical Atlantic Ocean. Recent new techniques now available which reveal enzyme variation and mitochondrial DNA poly morphism can revolutionize the study of bio-geography. They can make p ossible, for example, the identification of closely similar larvae in instances where morphological characteristics are inadequate and can a llow measurements of genetic exchange or gene flow and the genetic rel ationships between widely separated populations. Cladistic analysis, a lthough revealing little about the processes leading to present geogra phic distributions, can help reconstruct large-scale geographic relati onships as related to the evolution of taxa.