Quantifying within- and between-season variability in Adélie penguin fledgling weights: statistical and practical implications for detecting change

We develop statistical models of Adélie penguin fledgling weight data collected at Béchervaise Island and use them in a power analysis as a continuation of the CEMP review. The statistical models incorporate both within- and between- season variability of fledgling weights from first principles using raw data as recommended in Southwell et al. (2004). These models should be viewed as initial attempts at incorporating multiple sources of variability rather than final products because a number of issues need further consideration. Issues to be resolved include the form and direction of change of fledgling weights with a decline in resource availability, the consideration of total chick failure during severe food shortages and a possible change in variance associated with a change in the mean value. With these issues kept in mind, the major findings from these statistical models are the potential for reducing to 30 the number of birds weighed in a single 5-day period each year. If practical, this outcome could have substantial benefits by simplifying data collection. We also discuss some of the practical issues of continuing to measure fledgling weights at Béchervaise Island either by the current or the modified methodology.