The importance of the relationship between population density and habitat area

Abstract

The relationship between population density and habitat size, the density-area relationship, has received considerable, but often sporadic and unfocused attention (Root 1973, Risch 1981, Kareiva 1983, Bach 1988, Matter 1997). Recent reviews of the topic have revealed that there is a great deal of variability in the relationship with species showing patterns of increasing, decreasing, and constant density with habitat size (Bowers and Matter 1997, Bender et al. 1998, Connor et al. 2000). Despite this considerable variability, the overall trend across species appears to be increasing density with habitat size, especially for insects and birds (Bender et al. 1998, Connor et al. 2000; but see Gaston et al. 1999). Mammals show fairly constant densities with habitat area with about as many positive as negative density-area relationships (Bowers and Matter 1997, Connor et al. 2000). A variety of different mechanisms have been offered to account for density-area relationships. Root (1973) explained the positive density-area relationship seen for herbivorous insects in the 'resource concentration hypothesis'. He proposed that density would increase with area via dispersal if emigration rates were greater for smaller, and immigration rates were greater for larger patches (Root 1973). Subsequently, predatory rates (Risch 1981), habitat quality (Hanski 1994, Matter 1997), and social interactions (Bowers and Matter 1997, Bowers and Dooley 1999) that vary with patch size have been proposed to account for both positive and negative density-area relationships. Bowers and Matter (1997) proposed that habitat selection at small spatial scales may produce negative density-area relationships while positive relationships may arise at larger scales through colonization-extinction dynamics. Density-area relationships may also arise from methodological problems such as the mis-estimation of patch size for edge and interior species (Bender et al. 1998) or increasing amounts of non-habitat with census area (Smallwood and Schonewald 1996, Gaston et al. 1999). Gaston et al. (1999) detail many of the problems that can occur when estimating density-area relationships; however, many of the statistical problems discussed can be avoided through the use of alternative methods. Gaston et al. (1999) point out that for log density-log area plots the minimum allowable value is 1/area, which may produce a non-zero expected slope as well as heteroscedasticity. These statistical difficulties can be overcome by using an area-sensitive Poisson model (Feller 1968, Connor et al. 1997, Matter 1997). Here the expected abundance of organisms at a site is simply the mean density across all sites multiplied by site area raised to a power. This model can be fit with and without the exponent using generalized linear models to determine if density varies with area. The model has the advantage of having the appropriate variance structure and allows for the inclusion of inhabitable sites with zero density (Matter 1997). Despite research describing patterns of density with habitat area and attempting to attribute mechanisms responsible for it, researchers have yet to appreciate the importance of density-area relationships in a wider context. Similar to patterns of aggregation, density-area patterns influence a variety of ecological processes. Here, I illustrate how the density-area relationship of individual species is of fundamental importance for spatial population and community ecology and is pertinent to conservation issues.