Assuming the paradigm that catchment vegetation is the main source of particulate organic matter (POM) to rivers, the main objective of this study was to determine what the proportion of original C3 ...carbon from the forest had already been replaced by C4 carbon from sugar cane and pasture in the rivers of the Piracicaba Basin. In order to achieve this objective, we first produced a detailed landcover map using Landsat5‐TM images, and then we measured the carbon stable isotopic composition of the particulate riverine organic matter (δ
13
C‐POM) in seven sites along the major rivers and in two sites along a small creek. Sugar cane and pasture (C4 plants) covered almost 60% of the basin area, while silviculture, mostly of other crops, citrus, and forest that are C3 plants, covered 35%. Isotopic studies conducted in large pristine tropical rivers of South America and of Africa have shown that catchment vegetation is the main source of carbon in suspended POM. Our study demonstrates that relatively recent changes (70–80 yr ago) in landcover in the Piracicaba River Basin have already affected the composition of the riverine POM. Therefore, as in natural ecosystems, the vegetation (allochthonous source) plays an important role in the composition of the riverine POM in agricultural systems such as the Piracicaba River Basin. This control can be supported by the good correlation between cumulative area of the basin covered with C4 plants and the δ
13
C of the riverine POM. However, our study, differently from others, also shows that, during the low water period, in situ processes, such as primary production, may be an important source of carbon to the riverine POM.
Here we present the within-site, seasonal, and interannual variations of the carbon (δ13C) and nitrogen (δ15N) isotope ratios of leaves, wood, bark and litter from four sites in the Amazon region, ...Brazil. Samples were collected in Manaus (3° 06′07″ S; 60°01′30″ W), Ji-Paraná (10°53′07″ S; 61°57′06″ W), and Santarém (2°26′35″ S; 54°42′30″ W) with mean annual precipitation of 2207, 2040 and 1909 mm respectively. The overall average for all leaf samples was−32.3 ± 2.5‰ for δ13C and +5.8 ± 1.6‰ for δ15N (n = 756). The leaf δ values at these sites were often but not always statistically distinct from each other. The d13C values varied from−37.8‰ to−25.9‰ Pronounced differences in δ13C values occurred with height associated with differences in forest structure. The δ13C of leaf dry matter showed seasonal variations associated with the length of the dry season, despite the fact that total annual precipitation was similar among the studied sites. Leaf δ15N values ranged from+0.9‰ to a maximum value of+10.9‰, and the Santarém sites showed more enriched values than Manaus and Ji-Paraná sites. No seasonal variation was detected in the δ15N of leaves, but significant differences were observed among sites and with changes in canopy height. The isotope ratio data are consistent with our current understanding of the roles of light, water availability, and recycling of soil-respired CO2 influences on δ13C and consistent with our understanding that an open nitrogen cycle can lead to high δ15N values despite a significant number of legumes in the vegetation.