Comparison of the GLMM year effect variance from models of west coast response variables in the island closure experiment with the variance of anchovy recruitment

The variances of the random year effects estimated in the GLMM models for penguin response variables at Robben and Dassen islands are compared to the variance of annual anchovy recruitment over a comparable period. Values for the former are found to be considerably less than the latter. A possible hypothesis to explain this, together with its possible implications for the estimates of the effects of fishing from the experiment, are discussed.