Estimates of population growth rates of humpback whales (Megaptera novaeangliae) in the wintering grounds off the coast of Brazil (Breeding Stock A)

Abstract

Humpback whales wintering off the eastern coast of Brazil were heavily exploited by commercial whaling in the Southern Hemisphere. During recent years, clear signs of recovery have been observed, but few estimates of population growth rate exist. In this study, quantitative estimates of rates of population increase are obtained from sighting per unit of effort data (1995–98) using generalised linear models and maximum likelihood estimation. The error distributions considered for the models were Poisson and negative binomial. Predictors of the number of sightings included the year, month and 2-week periods during which the sightings were made. Predictors were treated as factors or numeric variables. For the numeric variables, quadratic dependence was also considered for each predictor to allow for possible non-linear relationships. Using Akaike Information Criterion (AICc) as a model selection criterion, the best model included year and month as continuous predictors. The data indicated strong support for the negative binomial over the Poisson models, but did not support models based on a finer temporal scale than month. Assuming year to be a linear predictor, the best estimate of the growth rate for the population wintering off Brazil was 7.4% per year (95% CI = 0.6–14.5%) during the period 1995–98. This estimate provides additional quantitative evidence that this population has been increasing and is consistent with the observed growth rates of other humpback whale stocks.