The Effect of Altitudinal Gradient on the Carbon and Nitrogen Dynamics in Coastal Atlantic Forest of Southeast Brazil

The Brazilian Atlantic forest is a vast heterogeneous region with 1.5 million km2, encompassing a large variety of forest physiognomies and compositions, containing large number of species. These forests are distributed in different topographic and climatic conditions, with high levels of precipitation. The rate of deforestation is high, approaching 350 km2 per year, showing be highly fragmented with a large number of species in extinction. The aim of this study was to understanding of the basic biogeochemistry functioning of the coastal Atlantic Forest. The study was carried out in São Paulo State, Brazil (23° 24' S and 45° 11' W). The studied areas were: Restinga Forest at sea level; Lowland Ombrophylus Dense Forest at 100m of altitude asl; Submontana Ombrophylus Dense Forest at 400m of altitude asl and; Montane Ombrophylus Dense Forest at 1000m of altitude asl. A sampling area of 1 ha in each phytophysiognomies was subdivided in contiguous sub-parcels (10 x 10m). The forest floor litter accumulated (0.06m2) was collected monthly (n=15), during 12 months, in each phytophysiognomies. Soils samples (0-0.05m depth) were collected (n=32) from square regular grids, 30m away from each other. Techniques of multivariate like principal components analysis (PCA) were used to determine correlations between the variable. The ordination graphs make possible to observe frequent of standards, representing a significant ratio of the variability of the data. The two first PCA axes cumulatively explained 60% of the total variance of the litter variables. Litter C and δ13C values were strongly influenced by altitude at 1000m. The N and δ15N of litter were influenced by altitude at 100 and 400m. The C/N relation was influenced by altitude at 0m. The lignin was elevated (p<0.01) at sea level in comparison with the other phytophysiognomies. The cellulose values did not vary significantly along the altitudinal gradient. Soil C and N concentrations progressively increased along the altitudinal gradient (p<0.01). Soil C:N ratio was significantly higher at 0m. The two first PCA axes cumulatively explained 76% of the total variance of the soil variables. Soil C and N contents were strongly influenced by altitude at 1000m and the C/N relation and sand were influenced by altitude 0m. It is also possible to observe that C, N and clay contents are inversely proportional to the soil sand fraction. This study showed that analysis of the principal components was useful in the grouping areas. The factors that explain differences along this gradient are not yet well known, but topography and the microclimate are appears to be the two main candidate factors regulating this nutritional variation of litter in the gradient.