Abstract
An analytical method is presented for detecting nonrandom niche displacement that might result from competition by examining the degree to which niche overlap is minimized along a single niche dimension. Species'' utilization spectra for different resources are viewed as intervals, or "sticks," arranged over a finite linear resource continuum. The method tests the actual extent to which the observed configuration of sticks overlaps less than would be expected if the sticks were cast down at random. Data previously analyzed to refute the importance of niche displacement (Poole and Rathke 1980) have been found to support it when reanalyzed with a computer simulation of the shuffled-sticks model (Cole 1981).