Improving methods to identify areas of endemism

GOLOBOFF, PABLO.; SZUMIK, CLAUDIA A.

Helsinki

Congreso; XXI Annual Meeting of the Willi Hennig Society; 2002

Willi Hennig Society

Szumik et al. (2002) proposed an optimality criterion to evaluateareas of endemism. That paper proposed to divide the study regioninto cells, assigning a score to a possible area (= set of cells) based onthe number of species which can be considered endemic for the area.This is an all-or-none criterion, by which a species either contributes ordoes not to the endemicity score. Improved criteria and heuristicalgorithms are proposed here. The score of endemicity for a given setof cells is the sum of the individual scores for each of the species; thescore for a species is higher when the species adjusts better to the areaand lower when it adjusts less well. In contrast to the previously usedcriterion, the improved one can produce meaningful evaluations fordisjunct sets. Heuristic procedures to find the sets of cells with highestscores are discussed. The problem of comparing incompatible sets ofcells is discussed; since two contradictory sets may have their scoresgiven by different species, simply discarding sets contradicted by otherswith higher scores may lead to missing some well-supported areas ofendemism. Thus, the computational problem is more complex thansimply finding the sets of highest score.