Testing for MacArthur-Wilson Equilibrium with the Arthropods of the Miniature Spartina Archipelago at Oyster Bay, Florida

SYNOPSIS. The unique features of the MacArthur-Wilson modelof equilibrium biogeography are the immigration and extinctioncurves. In the model, the immigration rate for each island isnegatively correlated with species number, and island extinctionrate is positively correlated with species number. Direct tests of the sign and statistical significance of immigrationand extinction curves were done, with data on arthropods oftiny Spartina alterniflora islets, in north Florida. Rey (1981)defaunated six islets in the spring, then censused the recolonizationweekly, for one year. Two control islets and a mainland plotwere also censused. Distinct tendencies in favor of MacArthur-Wilsoncorrelations were present for both immigration and extinction.But sufficient extraneous variation occurs in immigration andextinction rates to make correspondence to MacArthur-Wilsoncurves statistically non-significant, unless probabilities arecombined among islands. Thus, mathematical deductions from theMacArthur-Wilson model, which treat the curves as deterministiclines with no variance, are less appropriate than approachesthat incorporate stochastic variation. The study of community patterns, by models such as the MacArthur-Wilsonmodel, are not a substitute for careful autecological studies.Mechanisms of colonization and co-existence are only dimly hintedat by gross community patterns.