Niche Apportionment or Random Assortment: Species Abundance Patterns Revisited

Abstract

(1) This work examines the theoretical framework of species abundance patterns and the practical problems of applying models to real data, with particular reference to communities with a relatively small number of closely related species. (2) An attempt is made to give logical coherence to a range of niche apportionment models, including some newly developed ones. Niche apportionment can be categorized through the sequential breakage process of total niche, including the MacArthur Broken-Stick model which has traditionally been envisaged as a simultaneous breakage model. (3) Five niche apportionment models, i.e. Geometric Series, Dominance Preemption, Random Fraction, MacArthur Fraction and Dominance Decay are constrasted to a model where no conventional niche apportionment is assumed, i.e, Random Assortment model which may relate to a highly dynamic community under a variable environment. In addition, Composite model is proposed which combines niche apportionment and random assortment. (4) Species-abundance data from a community of epiphytic chironomids were used to illustrate the analyses involving the seven models. Randon Fraction and Random Assortment both successfully fitted the data on the number of individuals, whereas only Random Assortment successfully fitted the biomass data. The epiphytic chironomid community is considered to represent a highly dynamic system which is not structured through the process of niche apportionment envisaged here. (5) Subtleties of analysis involving a model-fitting exercise are discussed.