The role of stochastic processes in producing nested patterns of species distributions

Abstract

Nestedness has received considerable attention in community ecology and conservation biology from both theoretical and empirical perspectives. This has lead to the creation of various metrics and null models to analyze nested subsets, all of which rely on the random placement of species to assess significance. However, if immigration and extinction are the processes that underlie species distributions on island systems, then null models might be better determined on the basis of randomly placed individuals. Consequently, we examined the effects of species–abundance distributions (uniform, dominance–decay, random–assortment, and dominance–preemption), island–size distributions (uniform and linear decrease), and total abundances (128, 256, 512, 1024, 2048, 4096 and 8192) on the degree of nestedness and its significance. Generally, matrices of species presence and absence created from the random placement of individuals were nested significantly according to null models based on the random placement of species. Island size and abundance had less of an effect on nestedness in systems dominated by only a few species than in systems in which abundances were distributed more evenly. Stochastic processes, such as the random placement of individuals, predispose systems to evince patterns of nestedness at the species level, which may account, in part, for the ubiquity of nestedness in nature.

A comparison of the leaf-litter ant fauna in a secondary atlantic forest with an adjacent pine plantation in southeastern Brazil

“…Ultimately, whilst we can not confi rm what determined the species clusters in the secondary forest, we can not exclude that stochastic processes, such as the random placement of individuals, could produce the observed patterns of species distributions (Higgins et al 2006).…”

The empirical Bayes approach as a tool to identify non-random species associations

“…Total species numbers m and total sites numbers n per matrix were also sampled from uniform distributions (10 B m B 100 and 10 B n B 100). This sampling protocol produced random matrices with relatively high matrix fills that were moderately to strongly nested due to passive sampling (Higgins et al 2006). In the second type of random matrix (M E ), species occurrences were sampled from a uniform random distribution until a randomly chosen number of species (again 10 B m B 100 and 10 B n B 100) and matrix fill (0.1 \ fill \ 0.9) was reached.…”

Spatio-temporal nested patterns in macroinvertebrate assemblages across a pond network with a wide hydroperiod range

“…Dispersal also is favored by physical links between sites which can occur following heavy rainfall (Fahd et al 2007). As high connectivity is one of the main drivers of nestedness in biological communities (Boecklen 1997;Cook and Quinn 1995;Wright et al 1998;Higgins et al 2006) and dispersal has been suggested as being important in driving nested patterns (Patterson 1990;Atmar and Patterson 1993;Loo et al 2002;Hausdorf and Hennig 2003), we expected, and found, good dispersers to be more highly nested than poor and non-dispersing taxa (Cook and Quinn 1995). Isolation can affect the degree of nestedness observed across systems (Wright et al 1998;McAbendroth et al 2005;Monaghan et al 2005), something that we have observed in distant ponds located in northern and southern areas of the park, which had more idiosyncratic assemblages.…”