| Summary: | `titrebThe forest patches biodiversity, “island anthroposystems” in openfield landscapes : the Western Gâtinais (France) case.`/titrebTo the south of Paris, between the Fontainebleau and Orléans forests, an extensive plateau forms the Western Gâtinais. The openfield landscape is made of many forest patches. A sample of 71 woodlots was studied in order to characterize their flora and to test the predictions of Island Biogeography and Landscape Ecology as regards to the richness of their species assemblages.We found 671 species overall. From one patch to the next, the species numbers vary greatly (minimum : 18 maximum : 270). The composition of assemblages is diversified, as 57 % of the species are present in less than 6 woodlots.The classical species-area relationships is not falsified, despite of some contrary examples. On the other hand, results are unclear as regards the isolation effect. Our results suggest that two landscape characteristics have a positive influence on the species richness : 1 - when forest patches form “archipelagos”, the species flows are probably facilitated, resulting in a higher species number in each patch (the “archipelago effect”) 2 - the species richness of a patch is related to the level of ecological heterogeneity of the surrounding landscape matrix. The comparison of old woodlots (existing before 1880) and “neo-patches” (created since 1880) does not reveal any correlation between species number and age : neo-patches can be as rich as old patches this observation suggests rapid dissemination processes.Human causes are to be taken into account to explain the differences of species richness between patches. We have established a global index of human influence : in large woodlots (5 ha and more), the species richness is significantly correlated to this index. The diversity of practices depends on the size of the woodlots the areas, shapes and relative locations of woodlots result from the evolution of land use : woodlots are “anthroposystems”. The trajectory of each one and ...
|
|---|