Search

The production of reactive nitrogen worldwide has more than doubled in the last century because of human activities and population growth. Advances in our understanding of the nitrogen cycle and the impacts of anthropogenic activities on regional to global scales is largely hindered by the paucity of information about nitrogen inputs from human activities in fast-developing regions of the world such as the tropics. In this paper, we estimate nitrogen inputs and outputs in Brazil, which is the world’s largest tropical country. We determined that the N cycle is increasingly controlled by human activities rather than natural processes. Nitrogen inputs to Brazil from human activities practically doubled from 1995 to 2002, mostly because of nitrogen production through biological fixation in agricultural systems. This is in contrast to industrialized countries of the temperate zone, where fertilizer application and atmospheric deposition are the main sources of anthropogenic nitrogen. In Brazil, the production of soybean crops over an area of less than 20 million ha, was responsible for about 3.2 Tg N or close to one-third of the N inputs from anthropogenic sources in 2002. Moreover, cattle pastures account for almost 70% of the estimated 280×106 ha of agricultural land in Brazil and potentially fix significant amounts of N when well managed, further increasing the importance of biological nitrogen fixation in the nitrogen budget. Much of these anthropogenic inputs occur in the Brazilian savannah region ( Cerrado), while more urbanized regions such as the state of São Paulo also have high rates of nitrogenous fertilizer inputs. In the Amazon, rates of anthropogenic nitrogen inputs are relatively low, but continuing conversion of natural forests into cattle pasture or secondary forests potentially add a significant amount of new nitrogen to Brazil given the vast area of the region. Better measurements of biological fixation rates in Brazil are necessary for improving the nitrogen budgets, especially at a more refined spatial scale. [ABSTRACT FROM AUTHOR]

Human activities that modify land cover can alter the structure and biogeochemistry of small streams but these effects are poorly known over large regions of the humid tropics where rates of forest clearing are high. We examined how conversion of Amazon lowland tropical forest to cattle pasture influenced the physical and chemical structure, organic matter stocks and N cycling of small streams. We combined a regional ground survey of small streams with an intensive study of nutrient cycling using N additions in three representative streams: a second-order forest stream, a second-order pasture stream and a third-order pasture stream. These three streams were within several km of each other and on similar soils. Replacement of forest with pasture decreased stream habitat complexity by changing streams from run and pool channels with forest leaf detritus (50% cover) to grass-filled (63% cover) channel with runs of slow-moving water. In the survey, pasture streams consistently had lower concentrations of dissolved oxygen and nitrate (NO) compared with similar-sized forest streams. Stable isotope additions revealed that second-order pasture stream had a shorter NH uptake length, higher uptake rates into organic matter components and a shorter NH residence time than the second-order forest stream or the third-order pasture stream. Nitrification was significant in the forest stream (19% of the added NH) but not in the second-order pasture (0%) or third-order (6%) pasture stream. The forest stream retained 7% of added N in organic matter compartments and exported 53% (NH = 34%; NO = 19%). In contrast, the second-order pasture stream retained 75% of added N, predominantly in grasses (69%) and exported only 4% as NH. The fate of tracer N in the third-order pasture stream more closely resembled that in the forest stream, with 5% of added N retained and 26% exported (NH = 9%; NO = 6%). These findings indicate that the widespread infilling by grass in small streams in areas deforested for pasture greatly increases the retention of inorganic N in the first- and second-order streams, which make up roughly three-fourths of total stream channel length in Amazon basin watersheds. The importance of this phenomenon and its effect on N transport to larger rivers across the larger areas of the Amazon Basin will depend on better evaluation of both the extent and the scale at which stream infilling by grass occurs, but our analysis suggests the phenomenon is widespread. [ABSTRACT FROM AUTHOR]

Anthropogenic N inputs and riverine export were determined for a meso-scale river basin in one of the most developed and economically important regions of South America. The Piracicaba River basin is located in southeastern Brazil and drains into a tributary of the Paraná River. The basin supports over 3 million people (about 2% of the population of Brazil) with intensive agricultural and industrial activities. During two years from 1995 to 1997, biweekly samples were collected at 10 stations along the Piracicaba River and its tributaries for analyses of dissolved and particulate N. The average annual flux of dissolved inorganic N and total N increased by a factor of 15 and 20 times, respectively, from the headwaters to the lower reaches of the main channel, whereas discharge increased by only 7 times. On a per area basis, the export of TN varied according to land use and was significantly correlated to the net input of anthropogenic N. Among 10 sub-catchments composing the basin, areas mostly covered by pasture and forest had the lowest export, whereas more agricultural and urban areas had higher export. The amount of N exported from each sub-catchment varied widely, but inputs were consistently higher than fluvial outputs. Losses and retention of N occurred throughout the basin but were especially high in the sub-catchment with a main-stem reservoir, suggesting that aquatic processing plays an important role in controlling riverine N export. Total net anthropogenic input to the Piracicaba River basin was 4,500 (± 900) kg N km-2 yr-1 of which about 40% was exported via fluvial outputs. [ABSTRACT FROM AUTHOR]