AREAS OF ENDEMISM OF SOUTHERN SOUTH AMERICA: AN EXAMPLE USING A LARGE DATA BASE OF INSECT DISTRIBUTIONS

De Mendoza CONICET

The studied area is located in Southern South America and correspond to the Patagonian and Subantartic subregions (Morrone, 2001). The definition of this region has been based mainly on its floristic composition and only a few studies were based on mammal or insect distributions. The present study is to our knowledge, the largest of this kind for the area. Originally the matrix was conformed by 19845 data entries of 1041 species. This matrix was filtered by eliminating the species with only one registry or species with more than one registry but from the same locality. The final matrix was conformed by 999 species of insects based on more than 12000 georeferenced data entries of insects of the orders Coleoptera and Hymenoptera. This information was gathered from systematic revisions and from the collections of the main museums and institutions of Argentina, Chile and Brazil. The program NDM version 2.2 was used to analyze the matrix by using different grid sizes and fill scores. Five analyses were made: 1) 1o X 1o grid without fill, 2) 1o X 1o grid with fill of 30, 3) 0.5o X 0.5o grid, 4) 1o X 0.5o (West – East oriented grid) and 5) 0.5o X 1o (North south oriented grid). Nested areas, conflictive areas and disjunctive areas were identified. Results of the search 1o X 1o grid with fill of 30 with posterior strict consensus of 40% was selected for a more detailed analysis. This search was selected because it recognizes the major number of not superimposed areas, 17 in total. The same search but without fill recognize six more areas not found in the first analysis. The other searches did not recognize new areas. Finally, four of the seven areas recognized by Dominguez et al., (2006) for the Patagonian steppe were recovered here. Furthermore, six areas matching with only one ecoregion (Valdivian forest, Chilean matorral, Magallanic sub-antartic forest, Patagonian steppe), several areas with associations of two or more ecoregions that were also proposed in previous studies (eg: Patagonian steppe - Low Monte), and two major areas that cover most of Chile and part of western Argentina were recovered. The use of a large data base allowed us the identification of a larger number of endemic areas with higher endemicity scores (defined by a higher number of species). The use of different grid sizes allowed the different areas raising the sensitivity of the analysis (e.g. the use of 1x0.5 grid allowed the differentiation of areas within the Maule region of Chile). Nevertheless, there are still many downfalls in the determination of areas of endemism: the lack of information for large areas such as the eastern region of the Argentine Patagonia. This problem does not arise in Chile, that is mostly well know.

Herramientas personales