Inferring surface time of Minke whales from inter-surfacing interval data using a hidden Markov model

• Main Authors: Christiansen, F., Rasmussen, M., Lusseau, D.
• Format: Conference Object
• Language: English
• Published: 2011
• Subjects: Faxaflói; Balaenoptera acutorostrata; Iceland; minke whale; North Atlantic
• Online Access: https://researchrepository.murdoch.edu.au/id/eprint/28967/

Surfacing rate data of Minke whales is an important factor used in the abundance estimates of Minke whale (Balaenoptera acutorostrata) stocks, both in aerial and vessel based surveys. Today, most abundance estimates of Minke whales rely on VHF-transmitters data rather than visual data. Visual data collected from land has the advantage of being relatively cheap to collect, which allows data to be collected from a larger number of individuals while causing no effect on the surfacing rates of the animals being studied, hence limiting biases. In this study, individual follows of Minke whales were conducted from a land-based station in Faxaflói bay, Iceland, and data on inter-breath intervals (IBI) were collected. Two distinct dive types were present within the surfacing data, which we defined as regular dives and deep dives. Those emerged from two different biological processes: whales spending time at the surface and whales engaging in foraging activities. A hidden Markov model was used to identify and define the density distribution of IBI as the observation state of these two hidden diving processes. Regular dives had a mean surfacing interval of 43 seconds (SD=44.8) and deep dives had a mean surfacing interval of 155 seconds (SD=115.1). The transition probabilities between the two dive types were estimated, from which the relative proportion spent in each dive type could be inferred. Minke whales perform regular dives during 62% and deep dives during 38% of their time. The relative proportions spent in each dive type can be used as estimates of how much time a whale will be typically at the surface available to be detected during cue counting surveys and to estimate the odds that a whale is in a long dive and therefore unlikely to be detected. Data was also collected from commercial whalewatching boats in the same bay, and were analysed together with the land based data to measure the effect of whalewatching boat interaction on Minke whale surface intervals. The proportion of time spent in deep dives decreased from 38% to 14% during interactions with whalewatching boats, while regular dives increased from 62% to 86%. The inter-surfacing interval used in abundance estimates of Minke whales in the North Atlantic today is derived from VHF-transmitter data and is about 77 seconds. Our mean values of surface intervals lies below and above this mean, which raises the question if a single mean value of surfacing interval can be used to make reliable abundance estimates of Minke whales, as both the dive type and the presence of vessels is likely to affect this value.