Stochastic spatial analysis of marine resources with application to minke whales (Balaenoptera acutorostrata) foraging: A synoptic case study from the southern Barents Sea

Sammendrag

A statistical method is outlined to construct multivariate confidence bounds for expected proportions of prey biomass in the sea and in the minke whale diets, as well as for the difference between these parameters. Parallel to the commercial whaling, from which 25 stomach content samples were collected, a comprehensive acoustic survey was conducted in order to identify and estimate the density (ton/nm(2)) of potential prey species. The intrinsic hypothesis was applied to the resource situation, assuming that the resource samples were realisations of a second order stochastic homogenous process. The precision of prey proportions in the sea was based on random realisations of the process, with estimated properties given by theoretical variograms and co-variograms fitted to the empirical ones. For the diet data, the precision was estimated by bootstrapping the 25 whale samples. Krill (Thyssanoessa sp. and Meganyctiphanes norvegica) and capelin (Mallotus villosus) dominated the diet. The biomass proportions of these species in the diet deviated significantly (10% simultaneous significance level) from the corresponding density proportions in the sea. A preference for foraging in the upper water layers may be a plausible interpretation of this result. When capelin and krill was examined separately from the other prey, no significant difference between diet and resource compositions was found at a 5% significance level.