The method of binomial corner-association analysis proposed in this paper is applicable to data from grids of contiguous cells, all or most of which are sampled. After sampling, the frequency distribution of cells at successive densities is grouped into a series of successive density categories. A separate binomial analysis is then conducted for each density category, in which the unit of observation is the number, from 0 to 4, of cells of the same density category which occur around each point of intersection on the grid. Applied to a distribution of the redear sunfish (Lepomis microlophus (Günther)) in a small lake in Oklahoma, the method revealed that unoccupied cells, and cells occupied at medium and at high densities, were contagiously distributed; low density cells, however, were acceptably random. The rationale of this analytical method is compared with that of inherently single-dimensioned analyses such as the Poisson and negative binomial. Corner-association analysis is shown to be inherently two-dimensional, and to have the additional advantage of focussing attention upon the interaction of the species and the environment, such interactions producing contagion at some density levels and randomness at others.
展开▼