GLMs relating penguin demographics and foraging behaviour to forage fish abundances
Marine and Coastal Management (MCM) established the Island Closure Task Team in 2010. This was prompted by concern about the conservation status of the African penguin, as data suggested that the population of the species had been decreasing rapidly over the previous few years. The role of the task team was to provide the Pelagic and EAF Scientific Working Groups (SWG-PEL and SWG-EAF) with advice relating to the merits of closing areas around certain penguin breeding colonies to purse-seine fishing (DAFF 2011). An island closure feasibility study was approved. The aim of this study was to evaluate the power of a long-term experiment to detect the effects of fishery closures around penguin colonies on penguin life-history parameters and foraging behaviour. Data have been collected both in the Western Cape (e.g. Steinfurth and Robinson 2012) and the Eastern Cape (Pichegru et al. 2012).
In December 2010, various aspects related to the impact of pelagic fishing near penguin breeding colonies were considered at the International Fisheries Stock Assessment Workshop held at the University of Cape Town. The international review panel made several recommendations in this regard (Parma et al. 2010). It was noted that while changing which islands were open or closed to fishing would not affect the estimation of the extent of residual variance, an alternation scheme involving opening closed islands and vice versa would allow data collected during the feasibility study to be used in a long term experiment (if conducted). A three-year period before switching the status of an island between open and closed was recommended.
Analysis of general linear models (GLMs) is a simple method of identifying and exploring relationships between data sets. The goal of this work is the estimation of the possible impact of pelagic fishing in the vicinities of penguin colonies on the population dynamics and foraging behaviour of penguins breeding at those colonies. A wide set of data series have been used.