Meso-Scale Variability of Soils and Forest Canopy Properties is Connected to Geomorphologic Features in Eastern Amazonia

In this study we investigated the relationships between landscape features such as terrain elevation and slope with two variables that drives forest productivity, soil texture and leaf area index (LAI). The study was carried out at the Tapajos region in Para State, eastern Amazonia. Twenty-four 0.25 ha plots were sampled along a ~150 km north-south transect in October 2002. Soil samples were collected (0-10 cm) in three random points in each plot for texture analysis. LAI was measured at 25 points regularly distributed in each plot. The geomorphologic attributes for each plot were extracted from the Shuttle Radar Topography Mission (SRTM) data linearly resample to 10 m spatial resolution. Terrain slope was linear and negatively related to the soil clay content (r² =0.73). Soil sand content had an expected opposite pattern (r² =0.72). The soil content of clay and sand along the elevation gradient can be strongly explained by a cubic polynomial curve (r² =0.82 and 0.81, respectively). LAI showed to be a logarithmic function of slope (r²=0.61), excluding plots located in the Valley regions. Moreover, LAI showed a linear and positive relationship with soil clay content (r² =0.52). Similarly to the relationships found between terrain elevation and soil texture, the 3rd order polynomial could explain 64% of the LAI variability over the Tapajos. Therefore, we concluded that topography is a major driver of the patterns of soil texture at the landscape scale and the combined effect of topography and soil can largely explain the patterns of LAI over the Tapajos. The combination of SRTM data and field-based information has the potential to increase the accuracy of ecosystem scale estimations of forest productivity in the Amazonia.