Natural mortality and spatial patchiness: reply to Gulland

The daily natural mortality of marine pelagic organisms ranging in size from bacteria to whales scales with weight to the power -0.301.This exponent is significantly lower than those for fish, -0.397, and for pelagic crustaceans, -0.537.The difference may reflect the action of secondary ecological factors within taxa.Spatial patchiness is one secondary factor that explains a significant amount of the variation in natural mortahty of pelagic fish eggs and earlystage larvae.The relationship remains significant if the weight exponent used to standardize mortality for size is changed from an ecosystem value of -0.25 to a 'fish' value of -0.397.To explain this observation, it is proposed that fish eggs and early-stage larvae are more vulnerable to predators when the embryos are aggregated in patches than when they are dispersed because these young fish lack the close-range antipredator behaviour that accompanies schooling in older fish.The effect on survival of the presence or absence of this component of schooling behaviour is not included in an earlier model of the relationship between predation and schooling.Introduction.In a recent paper McGurk (1986b) proposed that natural mortality, M, of pelagic fish eggs and early-stage larvae was controlled primarily by their body size, W, and secondarily by their spatial patchness, p, as measured by Lloyd's (1967) index.It was suggested that fish eggs and early-stage larvae are more vulnerable to predators than would be expected on the basis of their body size because they have not developed the motor skills and behaviour necessary to evade predators.Thus, patches of fish eggs and earlystage larvae produced by mass spawning and hatching may offer exceptional feeding opportunities to any predator that encounters them.The young fish should continue to suffer from heavy predation until the patches are dispersed by turbulent diffusion and until the larvae develop effective anti-predator behaviour.Gulland (1987) criticizes the mortality-patchiness hypothesis on 2 grounds.First, he argues that the original observation from whlch the hypothesis was developed, the deviation between the observed M of fish eggs and early-stage larvae and the M predicted from Peterson & Wroblewski's (1984) allometric model of M for fish, was not due to the secondary action of