Size selectivity of krill by penguins is shown to be highly sensitive to the statistical assumptions made during the analysis of data. The nested ANOVA method is proposed as being the correct approach for analysis because lack of independence between krill specimens found in the stomach of a given penguin prevents the pooling of krill lengths from different samples. Samples taken in Bahía Esperanza were used to illustrate the different approaches to this analysis. A highly significant linear regression between krill size and time of sampling was found.

Full use of information obtained from analysis of penguin stomach contents requires block sampling designs. Replicating samples from the same penguin or studying portions of stomach samples at different stages of digestion are proposed as alternative methods.