Niche Parameters and Species Richness

This study examines how spatial niche parameters change as functions of species richness. The basis of the study was the partitioning of space by a community of small mammals in eastern Australian coastal heathland. Live—trapping was carried out in 13 habitat patches with two to seven coexisting species of small mammals on a 6.4—ha study area. Data from the 1st 2 mo of this 5—yr study are analyzed in detail. Spatial niche parameters were calculated using trapping stations as resource states. Mean niche breadth and mean niche overlap were negative functions of species richness while mean niche separation was a positive function of species richness. Spatial niche separation was estimated using nearest—neighbor analysis. Total spatial niche overlap increased to a peak and then decreased as the number of species increased. The field results were tested against two null hypotheses of noninteraction between species, one with and one without habitat selection, while a third considered the species—abundance effect using equally abundant species. The simulation values showed similar trends for the first three parameters and a continued increase for total overlap. However, field values had significantly steeper slopes, and parameter values at the extremities were usually significantly different form simulated values. Simulated niche parameters became independent of species richness when species were equally abundant, but showed increasing dependence as habitat partitioning and then microhabitat partitioning were introduced into the models. The results support Pianka's niche overlap hypthesis; mean overlaps were significantly higher than expected with a few species and significantly lower with many species, suggesting diffuse competition. Unsaturated patches showed an opposite effect, when space did not appear to be limiting, strongly implying that interspecific interactions play an important role in limiting the closeness of species packing in this community. Additional species are accommodated by significantly increasing mean spatial niche separation, thus decreasing mean spatial overlap. The rate of change in mean spatial niche breadth is not significantly different from the simulated rates.