Experimental Community Studies: Time‐Series Tests of Competition between African and Neotropical Bees
1983
Roubik, David W.
Experimental studies of food competition were performed in lowland forest of French Guiana by controlled introduction and removal of a small number of African honey bee colonies, totaling 100 000—150 000 bees. Trends in brood production and food storage by observations colonies of native social bees, Melipona favosa and M. fulva, were described with linear stochastic statistical models. Data were taken after the introduction of honey bee colonies, and the models were used to construct falsifiable hypotheses that would suggest whether the honey bees disrupted colony growth and resource harvest of native species. During a 1—mo introduction of honey bees in the wet season, when floral resources were at low abundance, time—series intervention analyses showed that African Apis had no significant effect on native bee colonies. Repeated—measures ANOVA also showed no treatment effect on bee colonies, and removal of 5—10 honey bee colonies had no effect on brood dynamics and food storage of four conspecific colonies. All colonies foraged during experiments, yet no competition was recorded. Since the three species are diet generalists and share many of the same resources, competition at flowers was not disproved, although any cumulative effects were negligible. Linear, stochastic integrated autoregression/moving average (ARIMA) models were useful in removing statistical dependence in data series and allowing hypothesis tests, but intracolony variables showed seasonal decline that prevented introduction of deterministic components that would have improved some models. Perturbation of social tropical bee communities by colonizing African Apis is unlikely to become evident within a short time or at relatively low honey bee colony density.
显示更多 [+] 显示较少 [-]