Abstract
Mortality rates of marine fish larvae are incredibly high and can determine year‐class strength. The major causes of larval mortality are predation and starvation, and the performance of larvae in survival skills that can mitigate this mortality (predator evasion, foraging) varies among individuals and cohorts, but the causes of the variation are not known. Transcriptomics can link gene expression variation to phenotypic variation at the whole‐system level to investigate the molecular basis of behavioural variation. We used tag‐based RNA‐sequencing to examine the molecular basis of variation in predator evasion and routine swimming (trait related to foraging efficiency) in the larval red drum, Sciaenops ocellatus . We looked for functional gene networks in which interindividual variation would explain variation in larval behavioural performance. We identified co‐expressed gene groups (“modules”) associated with predator evasion traits and found enrichment of motor, neural and energy metabolism pathways. These functional associations and pattern of correlations between modules and traits suggest that energy availability and allocation were responsible for the magnitude of startle responses, while differential neural and motor activation were associated with differences in response latency.