Processes, that is, mechanisms, of habitat choice have to be distinguished from patterns of space use, that is, distribution of individuals in the environment resulting from individual decisions. Patterns of individuals' spatial distribution or variation in fitness are often used to infer habitat choice processes by individuals, because determining whether habitat selection occurs can be difficult. Empirical studies often analyze habitat choice processes by comparing site characteristics and patterns of site use in different types of habitats. Occupied sites are expected to be of higher quality than unoccupied or randomly picked up sites if habitat choice occurs. However, the same patterns can result from different processes, and from processes other than active habitat selection by individuals.
A widely used concept is the ideal free distribution (IFD), defined as the distribution of individuals among habitat patches expected under the assumptions that individuals (1) distribute themselves so as to optimize their fitness, (2) are free to move among habitat patches, that is, without any cost or constraint, and (3) have a perfect (ideal) and instantaneous knowledge of the relative quality of habitat patches and local density dependence function. At equilibrium, (1) mean individual fitness is equal on all patches because individuals are distributed among patches proportionally to the relative quality (availability of resources) of each patch and (2) individuals cannot increase their fitness by changing patch (Figure 12).
More realistic refinements of the IFD model including (1) different forms of density dependence (e.g., nonmonotonic density-dependent functions such as Allee effects) and (2) interindividual differences in competitive ability (allowing some individuals to monopolize resources and thus achieve higher fitness than others: ideal despotic or dominant distribution (IDD)) have been proposed, but the IFD represents a null model describing the spatial distribution of individuals in the environment that maximizes fitness at the population scale given the distribution of patches, and to which distributions generated via different habitat choice processes incorporating constraints on information accessible to individuals (the ideal assumption) and their movements (the free assumption) can interestingly be compared (Figure 13).
Inferring habitat choice solely from patterns of habitat use by individuals can be misleading because high densities may be observed on low-quality patches, for instance, because of constraints for individuals in obtaining reliable information about potential habitat patches quality. A direct investigation of habitat choice, aiming at identifying information cues and decision rules used by individuals and determining the extent to which habitat choice strategies affect fitness, is often more appropriate than indirect inferences. Such an approach links proximate factors (elements of the environment used by individuals for choosing in a mechanistic way) and ultimate factors (evolutionary causes of individual choices, that is, linked to the relative fitness of individuals adopting different habitat choice strategies) involved in habitat choice.
Was this article helpful?