Event Analysis Applied to Timing in Marine Fish Ontogeny

Abstract

We illustrate the utility of event analysis for characterizing timing in marine fish life histories, comparing event series among experimental groups, and predicting event timing in untested environments. Two examples using early life history traits are given, in example (1) we analyzed parental influence on metamorphosis timing in experimental laboratory populations of winter flounder (Pseudopleuronectes americanus). We fit a Weibull distribution to right-censored observations of age at metamorphosis and then estimated the fraction of individuals from each of two families that had become juveniles by a predetermined day. Offspring of one family were 1.25 times more likely than the second to enter the juvenile period under the applied time constraint. In example (2) we evaluated temperature effects on age at death of unfed capelin hatchlings (Mallotus villosus). Using an accelerated failure time model we predicted age at death at an untested temperature and, assuming a patchy distribution of prey, estimated the number of days available to a cohort to encounter a suitable food patch and still maintain the potential for generating an above average year-class. These examples demonstrate the utility of event analysis for more thorough use of observations on ontogenetic timing in experimental and natural populations.