An Initial Analysis of the Power of Monitoring certain Indices to Determine the Effect of Fishing on Penguin Reproductive Success from an Experiment where Pelagic Fishing is Prohibited in the Neighbourhood of Robben Island, but Continues around Dassen Island.

Historical data on fledging success and the breeders per adult moulter ratio for penguins at the Robben and Dassen island colonies are analysed using General Linear Modelling (GLM) to provide the basis for a simulation evaluation of the power of an experiment which closes a neighbourhood (about 20 km radius perhaps) around Robben Island to detect a statistically significant (5% level) impact of the (assumed multiplicative) effect of fishing on penguin reproductive output. Process error for each index is estimated to be about 20%. For an actual effect size of 30% (corresponding to a change in the population growth rate of about 3% p.a.), the experiment would have to continue for some 10 years for a better than 50:50 chance of such detection for the breeders per adult moulter index, and at least 20 years for fledging success. An initial GLM analysis to determine the direction and magnitude of the effect of pelagic catches on penguin reproductive output in terms of the two indices and colonies considered suggests that these catches enhanced this output in all four instances (a potential consequence of the hypothesis that fishing scatters large shoals, thus increasing the probability of natural predators finding their prey), though none of the estimates is statistically significant at the 5% level.