The larvae of most reef fishes disperse from the natal reef soon after hatching and return to the reef environment after a pelagic stage of several days to weeks. Ocean currents were assumed to disperse larvae widely, resulting in broad genetic connectivity between reef populations. In some recently studied cases, connectivity appears to be more limited and self-recruitment (settlement on natal reefs) more common than previously assumed. This may be the result of favorable hydrographic conditions, perhaps aided by a behavioral component of larvae using swimming and sensory capabilities to stay in proximity to a reef by using appropriate currents. Preliminary data on the genetic structure of the cardinalfish Apogon doederleini show significant differences between reef populations at the geographical scale of 7 to 18 km. Genetic differences between reefs using a hydrographic model would have predicted panmixis. This suggests that larval behavior plays an active role in retention and self-recruitment. This project will determine the genetic population structure of three reef fish species with different larval biology and study mechanisms of dispersal and recruitment behavior. The objectives are: (1) Determine in two consecutive years the degree of genetic substructure and its temporal stability of fish within and among reef clusters at spatial scales of approx. 10 to 20 km based on three species with different dispersal biology. (2a) Determine whether pre-settlement juveniles of the three species differentiate between odors of hydrographically more or less connected reefs and (b) Determine the temporal stability of their reef odor preference. (3) Extend the hydrographic dispersal model for different source populations among the proposed reef clusters and compare the results with connectivity results from genetic analysis and odor preference tests. For genetic analysis, geographic distances and dispersal links and barriers will be studied using DNA microsatellite markers, centered around One Tree Island in the Capricorn/Bunker group of reefs in the Great Barrier Reef, Australia. The hydrographic model of the OTI region will be employed to predict dispersal by advection. Genetic substructure and reef odor preference in settling larvae/juveniles of three species will be compared: Acanthochromis polyacanthus (no pelagic stage), (b) Apogon doederleini (pelagic, weak swimmers) and (c) Pomacentrus coelestis (pelagic, strong swimmers). A mini-flume for odor choice tests will be used on-board ships in which single settlement stage fish have shown significant odor preference for specific reefs. Intellectual merits are demonstrations of 1) the scale of genetic substructure in coral reef fishes impacting concepts about the scale of reef connectivity, 2) differences between hydrographic dispersal predictions and genetic population structure, 3) olfactory capabilities of reef fish larvae to discriminate between odors of closely linked reefs, 4) possible connections between odor preference and genetic population structure suggestive of early imprinting and retention near natal reefs. Broader impacts are in 1) coral reef conservation and management (connectivity, water quality), 2) training of the next generation of scientists in biological oceanography, behavioral ecology, evolution and (marine) conservation, 3) international collaboration.
