PS 72-153 - Using openModeller to analyze the geographical distribution of the Centridini bees (Apidae, Hymenoptera)

Friday, August 10, 2007
Exhibit Halls 1 and 2, San Jose McEnery Convention Center
Isabel Alves-dos-Santos1, Tereza Cristina Giannini1, Sandra R. C. Naxara1 and Antonio Mauro Saraiva2, (1)Ecologia, Universidade de São Paulo, São Paulo, Brazil, (2)Escola Politécnica, Universidade de São Paulo, São Paulo, Brazil
Bees of the tribe Centridini are native from the Western Hemisphere and especially diverse in Neotropical region. The tribe is composed of two genera Centris and Epicharis, which together account for about 170 species. The centridine represent sometimes more then 20% of the richness and about 30% of individuals of the local beefauna, a considerable number for a solitary bee. Females of these bees collect floral oil to surface the brood cells and to mix with pollen to feed the larvae. That is why they are called oil-collecting bees. Centridini are narrow associated to Malpighiaceae flowers, which produce the floral oil they need. Some species of Centris are of economic interest because they pollinate crops like wildcherry and cashew. In the present study we used the OpenModeller, which is an open source spatial distribution modeling tool, to analyze the geographical distribution of 17 species of Centris and 10 Epicharis. Our databank is based on the results of 46 bee surveys performed in Brazil. Our databank accounts for 4576 Centridini individuals. The maps resulted from modeling were compared to the results with the distribution presented by others biodiversity information systems like Discover Life. In many cases the occurrence area predicted by the model were confirmed with the maps produced by the others biodiversity centre, like Centris analis, C. bicolor, C. flavifrons, C. fuscata, C. lutea, C. similes, C. trigonoides, C. varia and Epicharis flava. On the opposite side, the modeling of some species agrees with our data, like Epicharis dejeanii and E. rustica. In both cases we have to consider the species misidentification or the fact of having not enough samples to conclude. In Brazil information on the predictive modeling of species distribution can be very useful to determinate conservation areas and places where we should intensify faunistic inventories.
Copyright © . All rights reserved.
Banner photo by Flickr user greg westfall.