Scale‐dependent spatial variance patterns and correlations of seabirds and prey in the southeastern Bering Sea as revealed by spectral analysis

Abstract

Uni‐ and bivariate spectral analyses of the spatial distribution of thick‐billed murres Uria lomvia and acoustic estimates of prey biomass in the southeastern Bering Sea were used to examine the spatial variance patterns of a predator and its prey at multiple spatial scales. Power, phase and coherency spectra from individual transects, as well as those averaged over all transects, were examined. The average spectra, representing a temporal scale of months, showed that murres and prey had similar spatial variance patterns and were positively correlated over the range of spatial scales studied. The individual spectra, representing a temporal scale of hours, showed several patterns that were not evident in the average spectra. In particular, the transect‐level analyses showed that the correlation between murres and prey was poor at spatial scales where prey variance was relatively low. This result suggests a new hypothesis to explain poor small‐scale correlations between consumers and resources: resource distribution is relatively uniform at small scales resulting in only a slight increase in foraging return for consumers showing an aggregative response at these scales. The differences among spatial scales and between the average and individual spectra illustrate how ecological patterns can vary with temporal and spatial scale.