The shorebird models of J. D. Goss-Custard, R. A. Stillman, and co-workers are good examples of IBMs that needed to include adaptive behavior because empirical model rules would not have been sufficient. The IBMs were developed to predict the impact of land reclamation, resource harvesting, and recreation on the winter mortality of species of shorebirds and waterfowl, for example, the oystercatcher (Haematopus ostralegus) in the Exe Estuary in England. The IBMs had to predict the effect of new environmental conditions, for which no empirical rules or data were available. The models had thus to operate on basic principles, that is, physiology and fitness-seeking feeding behavior that is based on adaptive decisions.
The habitat is divided into discrete patches, which vary in their exposure and their quantity and type of food. During each time step, birds choose where and on what to feed, or whether to roost. Time steps typically represent 1-6 h. The bird's state variables include foraging efficiency, dominance, location, diet, assimilation rate, metabolic rate, and amount of body reserves. Key environmental inputs to the models are the timings of ebb and flow and temperature, which both affect feeding and the amount of food needed to survive.
A main behavioral process of the model is interference competition (e.g., food stealing), which is related to the individual's dominance status and to local bird density on the patches. The submodels describing the bird's decision where to move, what to eat, and how much time to spend feeding, are based on principles from optimal foraging theory and game theory. The individuals are assumed to always try and maximize their own chance of survival.
Model predictions were compared with many observed patterns, and after several iterations of the modeling cycle, patch selection, prey choice, and the proportion of time spent feeding were accurately predicted for many species and sites. In one case, the increase in winter mortality due to land reclamation was known from observations. The model was parametrized for the pre-impact situation, then run for the situation with reduced feeding area, and the increase in winter mortality determined. The match of observed and predicted increase in winter mortality was almost perfect. It could also be shown that the model, if it had existed at the time the land reclamation took place, could have been used to recommend a certain mitigation measure that was under discussion but not realized because it was unclear whether it could really compensate the loss of original feeding areas.
The first shorebird model of the group of Goss-Custard and Stillman needed several years for implementation, parametrization, and testing, but subsequently simpler and more flexible models were developed, that could be used for management support within 1-2 years. Currently, the models are used for a suite of species of shorebirds and waterfowl at more than ten different sites all over Europe.
Was this article helpful?