Improving estimates of Adélie penguin breeding population size: developing factors to adjust one-off population counts for availability bias
New methods are presented for the collection and development of data to adjust counts of Adélie penguins made at any time in the breeding season to an estimate of the breeding population. The development of ‘availability adjustment factors’ involves the collection of time-series counts of population objects, such as adults, nests or chicks, throughout the breeding season and standardisation of the time series to a reference point one week after
the peak in egg laying, consistent with Standard Method A3 of the CCAMLR Ecosystem Monitoring Program (CEMP). Remotely operating cameras were used to obtain time-series counts of adults and nests, while time-series counts of chicks were obtained manually. Standardised time series were modelled using a generalised additive model framework to allow interpolation across days when no counts were available and to provide a formal estimate of uncertainty around the average function. The method is illustrated using data collected at the Béchervaise Island CEMP site in 2007/08. Estimates of the breeding population obtained by adjusting a one-off count using availability adjustment factors will not account for penguins that left the breeding site prior to the reference point because their
breeding attempt failed, breeding-age penguins that failed to return to the breeding site in that breeding season, nor pre-breeding penguins. The abundance of these components of the population can only be estimated through alternative methods such as mark-recapture of individually tagged birds. Given the paucity of adjustment data in the literature, further collection of availability adjustment data across space and time is recommended.