Spatial scales of variability in cod recruitment in the North Atlantic

Abstract

We reexamine the hypothesis that there are large-scale (thousands of kilometres) patterns of recruitment of Atlantic cod (Gadus morhua) in the North Atlantic. Previous analyses have used sequential population analysis to reconstruct population histories; however, these methods are subject to a variety of biases and rely on the accuracy of commercial catch-at-age data. Several different studies have arrived at conflicting interpretations using virtually the same data. Here we analyze alternative data from research surveys using statistical methods that explicitly account for estimation error and we also employ detrended sequential population analysis. We conclude that the spatial scale of recruitment correlations generally does not exceed 500 km.