Bayesian estimate of Australian humpback whale calving interval under sparse resighting rates: 1987–2009

Abstract

This study estimates a calving interval for humpback whales from a longterm photo-ID catalogue of 2,973 individuals resighted in Hervey Bay, East Australia. The study proposes a modification of two existing methods to handle partial identification of sex and age-classes of whales from visual surveys. One method truncates the data to just breeding females and discards all resighting events prior to the first observed breeding event. The second method utilises the multi-stage mark recapture (MSMR) framework and multi-event extension to include all resighted individuals and their entire encounter history. The performance of each method is assessed and the conditioning required to handle ambiguity of sex and age-classes is detailed, which is subtly different from most other mark-recapture methods. Both truncation and the multi-event methods led to similar estimates of calving intervals: 2.98 years (95% CI: 2.27–3.51) and 2.78 years (95% CI: 2.23–3.68) respectively. More importantly, estimates were more sensitive to the exact specification of resighting probabilities among age and sex classes than to the type of conditioning. However, the multi-event framework resulted in more precise estimates of other important life-history parameters such as apparent survival, and included a wider constituency of age and sex classes.