My research has focused on the identification, assemblage dynamics, and horizontal distribution patterns of larval fishes around rocky reefs in the Gulf of California, Mexico. In Chapter One, I analyze a series of light-trap collections taken over 35 nights at San Carlos, Sonora. Nearly 14,000 larvae from 19 families were collected. The five most abundant families contributed 90% of the larvae and the top ten families contributed 99%. Larvae of non-pelagic spawners (gobiids, labrisomids, tripterygiids, chaenopsids, pomacentrids, and bythitids) dominated the collections. Larval catches ranged from 7.5 to 2330.3 larvae/45 min, but changes in catch rate were not related to changes in ambient moonlight. In contrast, the volume of zooplankton collected was correlated with moonlight intensity. Dynamics of the ten dominant fish families were highly variable, but in most cases a large proportion of the larvae were caught on just a few nights. Taxonomic and size selectivities were apparently less severe in my study than in previous Australian studies, and the use of light-traps should be explored further. In Chapter Two, I outline the prediction that larvae of small, non-pelagic spawning fishes should more frequently be retained over reefs during development than fishes with other combinations of body size and egg type, and I describe my research testing this prediction. I made about 160 collections of fish larvae with a light-trap and plankton net at 1, 20, and 100 m from rocky shorelines. These collections yielded 27,265 larvae from about 50 families. Based on larval size frequencies, near-reef concentration gradients, and abundances offshore, I identified four families that can complete development over the reef but also have larvae dispersed offshore (Clupeidae, Engraulididae, Gerreidae, and Haemulidae). In addition, I identified seven families that primarily develop over reefs and have few or no larvae dispersed offshore (Tripterygiidae, Chaenopsidae, some Labrisomidae, Dactyloscopidae, some Gobiidae, Gobiesocidae, and Bythitidae). Adults of these seven families are mostly small, non-pelagic spawners. Larvae from four taxa of larger non-pelagic spawners (Ophioblennius, Labrisomus, Balistes, and Pomacentridae) did not appear to develop over reefs. These findings are in accord with the prediction I made. Chapter Three is a preliminary guide to identification of Gulf of California blennioid larvae. Although blennioid larvae are poorly known and few species have been described, they are well represented (ca. 20,000 larvae from five families) in my collections taken near reefs in the Gulf. Illustrations of 20 species, and brief descriptions of key characters for these and several additional species, are provided. More detailed taxonomic studies on Gulf blenniid, dactyloscopid, tripterygiid, labrisomid, and chaenopsid larvae are in progress.

No description available.

Convergence in body form and discrete morphological characters among fishes of similar habits is widespread in both freshwater and marine habitats. The identification of adaptive types of fishes alludes to predictive relationships between morphology and lifestyle. Now, a challenge confronting

Resident rocky-shore fishes were quantitatively sampled and visually censused on islands and mainlands in the Gulf of California. Patterns of biogeographic distribution were investigated in relation to the MacArthur-Wilson model of island biogeography, the lottery hypothesis of reef fish community structure. Analyses of the collections show regular patterns of distribution and relative abundance of rocky-shore fishes. Most of the variation in species number and species diversity is explained by latitude and the associated gradient of physical and oceanographic parameters. The correlation of species number and diversity is higher for island than for mainland samples. The best graphical (numerical clustering) evidence of distinct biogeographic regions in the Gulf is obtained using a similarity measure which includes proportional abundance of species (Horn's measure). Groups corresponded to upper, central, and lower Gulf areas. Greater species number, biomass, and numerical (H'(n)) and biomass (H'(b)) species diversity on islands is paralleled by greater water clarity and volume of ocean near collecting sites. This is true for the entire Gulf, the central Gulf, and for Isla San Pedro Nolasco versus the adjacent mainland rocky shoreline in the central Gulf. Community differences between island and mainland areas were greatest in the upper Gulf. 'Gamma' diversity, a proposed measure of community variance (site to site species turnover within habitats) is highest for mainlands and lowest for islands indicating more predictable community species composition on islands. Visual censuses show a good species-area curve for patch reefs and very small nearshore islands and lower species turnover in more exposed near shore habitats. There is a trend toward smaller body size of fishes on islands and an increase in body size with latitude. It is suggested that both phenomena are the result of thermally regulated growth rather than genetic population differences. High endemism in the resident rocky-shore community (the small sedentary, blennioid and gobioid fishes) is attributed to their relatively poorer means of dispersal (demersal eggs and short-lived pelagic larvae) as compared to the more mobile larger resident reef fishes with greater means of dispersal (pelagic eggs and long-lived pelagic larvae). Insular biogeographic processes in marine reef habitats are mediated by physical and biological oceanographic conditions and processes. Distance appears to be no significant barrier to dispersal for Gulf rocky-shore fishes. Immigration rates may be high in rocky habitats in general, but in inshore mainland areas physical and biological disturbance suggest higher extinction rates and consequent lower diversities. The results of these studies do not agree well with either the MacArthur-Wilson model of island biogeography or the lottery hypothesis of reef fish community structure, however, disturbance does seem to play some role in regulating diversity. Considering dispersal and colonization, oceanographic conditions (e.g., water quality, current patterns) may be important ecological factors that influence the evolution of this community.