BULLETIN OF MARINE SCIENCE, 41(2): 432-440,1987 INSHORE ICHTHYOPLANKTON: A DISTINCTIVE ASSEMBLAGE? C. Lavett Smith, James C. Tyler and Loretta Stillman ABSTRACT Our collections oflarval fishes made at underwater lights on the outer shelves of coral reefs in the Caribbean yield large numbers of specimens mostly representing only a few of the families of fishes that inhabit coral reefs as adults. Although it is difficult to sample close to the reefs and methods of taking quantitative samples there have yet to be devised, preliminary comparisons of our near-reef data with those from ichthyoplankton collections made by others during extensive offshore surveys suggest that the inshore larval fishes constitute a distinctive assemblage. Not only are different taxa dominant, but few inshore larval fishes show the morphological specializations that characterize many of the offshore larvae. We postulate that tropical marine fish larvae tend to be specialized either for long distance transport or for avoiding being swept downstream by offshore currents. This seems to indicate that there are two groups of larval fishes: a far-field assemblage of larvae that are morpho- logically modified or behaviorally specialized for long distance transport by ocean currents and a near-field assemblage of un specialized larvae that avoid currents and spend their entire lives in the vicinity of the reefs. It is widely accepted that nearly all fish species that live on coral reefs have mobile life history stages, spending either the egg or larval stages or both as members of the plankton. During this pelagic phase the fish undergo tremendous morphological changes and face environmental challenges that are drastically different from those they will face on the reef during their benthic existence. Since successful completion of the planktonic phase is a prerequisite to their settlement and subsequent life history, it is hardly surprising that a great deal of research effort is being devoted to studies of ichthyoplankton and to the processes by which fish larvae settle from the open water community and become part of the benthic ichthyofauna (McFarland and Ogden, 1985). Still, this research is in its infancy and we know very little of the events that take place during the early life history stages. Although large scale offshore surveys have given a general picture of the distribution of fish larvae in the Caribbean (Richards, 1984), there is still little information on the distribution of larvae in the immediate vicinity of coral reefs. Traditionally, the planktonic larval stages of marine fishes have been regarded as rather helpless organisms that drift passively at the mercy of oceanic currents. It has been suggested (Johannes, 1978) that many coral reef fishes select spawning sites where eddies or gyres tend to keep the eggs in the vicinity of the reefs where they are spawned. Lobel and Robinson (1986) reviewed the current patterns in the vicinity of the Hawaiian Islands and concluded that mesoscale current and eddy systems can entrap larval fishes and retain them offshore for long enough periods for them to complete the planktonic phase of their life history and that these eddy currents can account for some passive return to the vicinity where the eggs originated. Leis (1986) provided evidence that more larvae are retained on the windward side of Lizard Island, Great Barrier Reef, than on the leeward side. However, Leis (I983) reported that larvae belonging to the genus Thalassoma have been taken at sea more than 1,000 km from the nearest reefs which are the habitat of the adults. Apparently, there is sufficient variation so that substantial numbers of eggs and larvae "escape" from their home area and are carried long distances. Thus, the deployment of planktonic life stages in eddies and gyres serves 432 SMITH ET AL.: INSHORE ICHTHYOPLANKTON 433 two opposing demands. It enhances the probability that a large number of eggs and larvae will remain close to the habitat necessary to complete their life cycle but at the same time allows enough larvae to escape to assure that some will be carried to distant patches of habitat for colonization and the maintenance of genetic diversity. In this paper we compare the larvae that occur close to the reef (within 10 m of the substratum) with those found in open water. Although this comparison is necessarily preliminary and in many ways crude, two observations stand out. First, the open ocean collections contain different taxonomic groups of fishes than the inshore collections; and second, many larval fishes that live in the open sea have striking anatomical specializations whereas larvae collected close to coral reefs rarely show such modifications. METHODS AND MATERIALS Over the years sophisticated methods of collecting plankton in the open sea have been developed and used extensively. With large vessels and opening and closing nets it is possible to make reasonably accurate estimates of abundances of larvae at different depths and in various geographic regions. Richards (1984) has recently summarized the distribution and abundance of ichthyoplankton in the Caribbean Sea. Unfortunately, the quantitative methodology of the blue-water planktonologist cannot be directly applied to fine-scale near-shore studies. Large vessels cannot work close to reefs and the geometric complexity of the reefs makes it impossible to maneuver even small nets close to the reef from the surface. Even surface tows from small boats are subject to biases because of the difficulty of maintaining accurate speed and positions in surf, wind and currents. Consequently, in order to study the near-reef plankton we have had to resort to small nets towed by divers and to sampling the larval fishes that are attracted to lights placed on the sea floor. The latter has proved to be effective for some species, especially when the lights are employed on the outer slope of the reef; i.e., seaward of the reef crest. One of our light collections on the outer slope at Carrie Bow Cay, Belize, contained approximately 24 times as many larvae as the average of 5 collections from behind the reef crest. Our recent Carrie Bow Cay research is beginning to provide western Caribbean plankton data for comparison to our more extensive eastern Caribbean data base described below. Study Area. - The Salt River Canyon is a notch in the fringing reef of the north coast of S1. Croix, U.S. Virgin Islands. Its long axis runs approximately north and south. Its west wall is an almost vertical, thriving reef; its east wall slopes more gently and has a dense growth of gorgon ian corals. The floor of the canyon is sand, with patches of Halophila. 13 m deep at the inner end, sloping gradually to a depth of 26 m, then more steeply to about 80 m where it joins the precipitous outer wall of the fringing reef. Currents in the canyon are mostly tidal; only during periods of exceptional rainfall is there a substantial continuous seaward flow driven by freshwater runoff. Bottom tow collections were made in various parts of the canyon at depths between 16 and 40 m by divers towing a small plankton net that could be maneuvered close to the bottom and around coral colonies. The 0.505-mm mesh net was 128 em long with a 30-cm diameter mouth opening. It was mounted on an aluminum frame with two handles and fitted with a collecting bucket with side ports covered with the same mesh. The same net was used at night to sweep the beam of a flood light mounted on the NOAA habitat HYDROLAB which was located at the inner end of the canyon in water 16 m deep. Surface collections in the canyon were made with a 50-em plankton net towed from a sma1\ outboard motorboat at and just below the surface. Samples from two offshore stations were made with multiple opening and closing nets (MOCNESS). These latter collections were made by the National Marine Fisheries Service Southeast Fisheries Center, Beaufort Laboratory, as part of the OTEC project, one at Lat. 17°52.5'N, Long. 64°29.5'W; the other at Lat. 17°50.0'N, Long. 64°48.0'W. Seven nets with 1 x IA-m openings were deployed between the surface and 100 m for 10- and 20-m (vertical distance) oblique tows. Our bottom tows and light sweeps were made 26 April through 2 May 1984 and 16-21 May 1985, and our surface tows 23-28 September 1985. The MOCNESS samples of the NMFS were co1\ected 9 May 1984. RESULTS Taxonomic Comparisons. -Our preliminary survey of the literature on the fish fauna of the Caribbean area indicates that the total fish fauna is about 1400 species 434 BULLETIN OF MARINE SCIENCE, VOL. 41, NO.2, 1987 representing approximately 185 families (Smith, ms.). Obviously, such an estimate cannot be precise because undoubtedly there are some species that occur in the region which simply have not yet been reported. Nevertheless, we believe that this estimate is within 10% of the actual number. For the following discussion we have eliminated 23 families of elasmobranch fishes since they do not have planktonic larvae. We estimate that 85 families (52.5%) of bony fishes contain species that live on reefs or in other inshore en vironmen ts and 77 families (47.5%) are strictly offshore fishes. Thus, slightly more than half the families known from the Carib- bean region are inshore fishes. In his report on the larval fishes of the Caribbean, Richards (1984) reported 86 families (53.1 % of the total) of which 50 (58.1 %) are inshore families and 36 (41.9%) are oceanic. In other words, the offshore plankton samples contain about the same proportion of inshore and offshore fishes as the total fish fauna of the region.