In a Dec 2006 paper in Science Anticipatory reproduction and population growth in seed predators by Stan Boutin et al (314(5807):1928-3) describe how squirrel reproduction increases prior to mast events (the synchronous production of large numbers of seeds), allowing population to increase with the pulse of resources (rather than lagging behind). This paper is important because it shows that the models animals have of their environment are an important aspect of their population dyanmics.

On Faculty of 1000 Biology my colleague Andrew Gonzalez writes about the paper:

The intriguing evidence presented in this paper suggests that classical theories of consumer-resource interactions are in need of modification. The vast majority of consumer-resource theory assumes that the reproductive success and population growth of a consumer depends upon the prior abundance of its prey. That is, that consumer-resource models are based on feedback control. Boutin and co-authors present convincing evidence that the population growth of squirrels (both North American and European) may in fact involve feedforward control. This means that the squirrels seem able to anticipate future “bumper” resource (seeds) production and adjust their reproductive output to coincide with this production. Although the cues used by squirrels to predict the future are not well understood, this finding should stimulate theoretical ecologists to alter their equations.

In my 2003 Ecology paper, Model uncertainity and the management of multi-state ecosystems: a rational route to collapse (84(6) 1403-1411), my co-authors, Steve Carpenter and Buz Brock, showed how people’s expectations of the future can alter system dynamics. In that paper developed a simple integrated social-ecological model that integrates a model of the future that people use to make decisions. When agents update their models of the future based on the behaviour of the world, the actions of the agents (and consquently the behaviour of the world) can change in surprising ways which do not occur when behaviours are fixed. In economics expectations about the future are often expected to stablize system behaviour, but when people’s models do not match the complexity of the world, expectations about the future can destablize the system producing complicated cycles of behaviour.