Study design

 

Land cover gradient: Scale, Representativity, extent and grain: 

The problem of relating phenomena across scale is a central problem in biology and in all of Science (Simon 1992). Addressing the problem of scale also has a fundamental applied importance. In consequence there are some problems related to how interpolate obtained results on bigger areas. We addressed the challenges by first exploring the forest loss and isolation pattern (as suggested by the RAPELD methodology) as a preliminary procedure before selecting the areas to be sampled. To define the landscapes where field samples will be collected, we divided the entire region in hexagons of 5000 hectares, hereafter landscapes. Using the information about altitude, we defined as the entire study area as those landscapes located above 150 meters of altitude (360 landscapes). By using the information from SOS Pantanal, the latest land use cover map generated from this region, we differentiated three land cover types, deciduous forest, riparian vegetation, and open areas, in all landscapes in the region . By using the combined information from the natural vegetation types, we performed an analysis of the total remaining vegetation and some configuration measurements of the remaining vegetation within the 360 landscape analyzed. To perform this analysis we transformed the shape files into raster files and processed the information using Fragstagts (McGarigal 2002). Our results showed that most landscapes in the region have less than 500 ha of total native remaining vegetation. As evidenced in other fragmentation studies we found a clear correlation between total native vegetation and largest patch index . The fragmentation process showed that as total native vegetation decrease there is an increase of total number of fragments, whereas when most native vegetation is loss there is also a decrease in the number of fragments . 

 

Although recent studies carry out at a landscape scale have used various landscape sizes (from 250 ha to 10000ha) our landscape size (5000ha) use an intermediate size, in which even the larger mammals recorded in this area may use as a home range area (Tapirus terrestris, Tayassu pecari) . Nonetheless it is important to mentioned that is still a huge area given the kind of experiments taken place, however we expect that location areas where  experiments are taken place will represent what is happening in the area.