INDIVIDUAL-CENTERED ANALYSIS OF MAPPED POINT PATTERNS REPRESENTING MULTISPECIES ASSEMBLAGES

On the basis of Ripley's combined count-distance method, Juhasz-Nagy's information theoretical functions and the proposition of Williams et al. for the study of small-scale community pattern, a new procedure is suggested for elucidating multi-species point patterns based on digit ized field data. The method utilizes nested circular plots with increa sing radii drawn around each individual and determines changes in flor istic composition along this space series. The information provided by detecting the species composition around the sample plant is calculat ed, and its observed mean for all individuals is compared, for each ra dius, to the expectation under the null model, i.e. for complete spati al randomness of all points. The departure from randomness is illustra ted by conventional profile diagrams and is tested for significance ba sed on confidence envelopes simulated by Monte Carlo methods. One adva ntage of the individual-centered sampling strategy is that the role of each species in influencing its own neighbourhood can be analyzed sep arately, providing information for the assessment of guild structure a nd assembly rules in communities. The performance of the method is eva luated using artificial and simulated point patterns.