A model for estimating meal frequency and meal size from stomach data with an application to Atlantic cod (Gadus morhua) feeding on capelin (Mallotus villosus)

The feeding of a predator on prey that occurs in schools is modelled as a stochastic process involving two random variables: (i) the number of schools that the predator finds per unit time and (ii) the number of prey items captured in each encounter. The number of encounters is governed by a Poisson process and the number of prey items can be either a Poisson random variable or more generally a negative binomial. The principal parameters in the model can be interpreted as the average meal frequency and the average meal size. This model is applied to Atlantic cod (Gadus morhua) feeding on capelin (Mallotus villosus). Cod stomachs from six length-groups were sampled off eastern Iceland and the number of capelin in individual stomachs recorded. The parameters in the stochastic feeding model are estimated by the method of maximum likelihood using the observed frequency distribution of capelin numbers in cod stomachs. The data fit the model well where the meal size is modelled by a negative binomial distribution, but less successfully where the meal size has a Poisson distribution. Finally, using the estimates of average meal frequency and average meal size the average and the standard deviation of consumption rate are calculated.