Evaluation of in situ enclosures during a study of the importance of starvation to the vulnerability of herring larvae to a piscine predator

Two competing hypotheses about the effect of starvation on the vulnerability of fish larvae to predators were tested in large enclosures (mesocosms). One hypothesis predicts greater vulnerability as starvation weakens larvae. The alternative recognizes that the behaviour of starving larvae may reduce encounters with predators or fail to elicit predatory attacks, thereby decreasing vulnerability. Results of four large-scale experiments, in which progressively starved Atlantic herring Clupea harengus L. larvae were exposed to larger herring as predators, were unable to support either hypothesis because of a large amount of variability. Much of the variability was thought to be due to erratic feeding behaviour of the predators. Additional experimental observations suggest that sunshine and predator satiation can affect these experiments.