Behaviour that influences dispersal and connectivity in the small, young larvae of a reef fish

Determining the scale of larval dispersal and population connectivity in demersal fishes is a major challenge in marine ecology. Historically, considerations of larval dispersal have ignored the possible contributions of larval behaviour, but we show here that even young, small larvae have swimming, orientation and vertical positioning capabilities that can strongly influence dispersal outcomes. Using young (11–15 days), relatively poorly developed (8–10 mm), larvae of the pomacentrid damselfish, Amblyglyphidodon curacao (identified using mitochondrial DNA), we studied behaviour relevant to dispersal in the laboratory and sea on windward and leeward sides of Lizard Island, Great Barrier Reef. Behaviour varied little with size over the narrow size range examined. Critical speed was 27.5 ± 1.0 cm s−1 (30.9 BL s−1), and in situ speed was 13.6 ± 0.6 cm s−1. Fastest individuals were 44.6 and 25.0 cm s−1, for critical and in situ speeds, respectively. In situ speed was about 50% of critical speed and equalled mean current speed. Unfed larvae swam 172 ± 29 h at 8–10 cm s−1 (52.0 ± 8.6 km), and lost 25% wet weight over that time. Vertical distribution differed between locations: modal depth was 2.5–5.0 and 10.0–12.5 m at leeward and windward sites, respectively. Over 80% of 71 larvae observed in situ had directional swimming trajectories. Larvae avoided NW bearings, with an overall mean SE swimming direction, regardless of the direction to nearest settlement habitat. Larvae made smaller changes between sequential bearings of swimming direction when swimming SE than in other directions, making it more likely they would continue to swim SE. When swimming NW, 62% of turns were left (more than in other directions), which would quickly result in swimming direction changing away from NW. This demonstrates the larvae knew the direction in which they were swimming and provides insight into how they achieved SE swimming direction. Although the cues used for orientation are unclear, some possibilities seemingly can be eliminated. Thus, A. curacao larvae near Lizard Island, on average swam into the average current at a speed equivalent to it, could do this for many hours, and chose different depths in different locations. These behaviours will strongly influence dispersal, and are similar to behaviour of other settlement-stage pomacentrid larvae that are older and larger.