Developmental changes in opacity and visibility of larval Atlantic herring, Clupea harengus L., and their vulnerability to visual predation by whiting, Merlangius merlangus L.
Features associated with the visibility of larval herring, Clupea harengus, to predators were investigated in the laboratory using larvae reared from artificially fertilised eggs. Herring larvae were reared up to the point at which they metamorphosed into juveniles. Photography and image analysis were used to measure the relative sizes of opaque structures. Lateral body area increased with body length anisometrically (exponent 2.62) and eye area increased at a similar rate (exponent 2.57), but the opaque pigmented lateral area increased as length4.41. The lateral area of the full gut increased with length3.04, the proportional increase in opacity resulting from feeding decreasing with body size. Under anaesthetic, larger larvae increased the dispersion of their melanin whereas smaller larvae contracted their melanophores. Herring larvae expanded their melanophores when confined near a black background; in white surroundings they contracted them by up to 90&37. At ecologically realistic intensities, changes in downwelling light did not affect melanophore state but the introduction of sidewelling light had a significant influence when it reached a threshold intensity. Vulnerability to predation by whiting, Merlangius merlangus, was tested. Larvae with full guts were more vulnerable than unfed larvae, the difference increasing with decreasing light intensity. The probability of capture increased with light intensity and body size. There was no evidence of an increasing ability to evade capture as the larvae grew. Against a dark background, larvae with contracted melanophores were more vulnerable than more pigmented larvae; the situation was reversed with a lighter background. The activity of larvae were observed using infra red illumination. Neither the presence of food in the gut nor changing light intensity affected activity levels. The light intensity threshold for feeding was found to differ according to prey type. During encounters with a potential predator the frequency of larvae responding was found to be low. Finally, it was demonstrated that juvenile herring are unaffected by a potential 'alarm substance'.