Reconstructing Humpback Whale Populations in the North Atlantic

No abstracts are to be cited without prior reference to the author.

Our understanding of changes in populations of cetaceans is often based on fitting single species density-dependent population models to contemporary data. This understanding can be extended by also using historical data. The process of fitting population models to both contemporary and historical data can provide a focus on the history of the population over time periods of multiple life spans, tests of the consistency of the historical and contemporary data, and a test of the structure of the population model. A recent example of an attempt to fit historic and contemporary data for North Atlantic humpback whales is described. A complex spatial model with compensatory and depensatory density-dependence on calf survival was used to integrate historical removal data with contemporary information on population breeding structure, regional abundance, and rates of increase in different feeding grounds. The model was unable to mimic the observed positive rates of increase in most areas of the North Atlantic. As such, the modeling provided an improved historical focus for North Atlantic humpback whales. The apparent inconsistency between the historical and contemporary data provides a basis for improving both. Further, it raises questions about the ability to model the complex feeding and breeding population structure of this species. We place this modeling attempt in a larger context, and suggest that such modeling exercises can be one of the few ways of testing, for cetaceans, the adequacy of central single-species model assumptions and concepts, such as that the population was in equilibrium prior to harvesting, density-dependent depensation and compensation in vital rates, reversibility of the effects of harvesting on the population dynamics, and the lack of effects of harvesting on other components of the ecosystem. Although short-term management advice for cetaceans is often based on single-species population models that ignore historical removals, greater confidence in long-term predictions may be expected if the model can successfully integrate historic and contemporary data. Finally, we suggest that this practice of selecting data to fit models, rather than selecting models that fit the available data, is a misapplication of Occam’s Razor that may inhibit our understanding of the long-term effects of whaling