In tropical forests growing on highly weathered soils, a significant fraction of ecosystem stocks of essential plant nutrients exists in the vegetation. Deforestation removes these nutrient stocks, perhaps leaving a nutrient-poor soil incapable of supporting sustainable agricultural practices or vigorous forest regrowth. On the other hand, many properties of tropical soils and vegetation, including pH-dependent variable charge, large effective rooting depths, and mycorrhlzal associations that can extract soil nutrients at low concentrations, may render these ecosystems resilient to disturbance. The objectives of the proposed research are to study these biogeochemical fluxes in a mosaic of primary and secondary forests and productive and degraded cattle pastures that is typical of the eastern Amazon Basin.
The project group have already measured nutrients in soils, vegetation, rain water, soil solutions, and stream water, and emissions of gases from the soils of these land uses. While addressing many of the changes in biogeochemical fluxes, this research has also raised new questions: Where does the nitrogen come from to support secondary succession when degraded pastures, which have depleted soil pools of plant-available N, are abandoned? Many of these degraded pastures arc being "reformed" by fertilizing with P and replanting in new varieties of pasture grasses, but where do the grasses obtain N to support their vigorous productivity? What are the rates of symbiotic and associative N fixation in these ecosystems? Is most of the fertilizer P taken up by the grasses, adsorbed onto iron and aluminum oxides in the soil, or lost to stream water? Can native pioneer successional species extract soil pools of P other than those identified as "plant-available" by agronomic soil test indices? Pasture and secondary forest soils that receive ash inputs following forest clearing and burning arc eventually reacidi~ed, but by what mechanisms and over what time scales?
These collaborative investigations will link mechanistic studies of nutrient loss and accumulation with field measurements of biogeochemical fluxes of plant macronutrients. Hydrologic investigations, including well transects, piezometer nests, soil moisture measurements, and energy balance calculations in each land use, will quantify flow paths of water through soil profiles and to the stream. Measures of stream water discharge and estimates of groundwater flow and unsaturated flow will be combined with measures of solution concentrations to calculate nutrient fiuxes within soil profiles and between the terrestrial and aquatic systems. Rates of biological nitrogen fixation will be estimated in field plots planted in grasses and in secondary forest pioneer species using an innovative combination of ~sN pool dilution and xsN mass balance. Changes in the isotope ratios of ~4N:~sN will provide a sensitive measure of atmospheric inputs to these systems. A mass balance of a greenhouse bioassay of P availability will identify the extractable P pools of soils sampled from various depths that are taken up by growing seedlings. In addition, the abundance of these physio-chernical pools of soil P will be compared to plant demands by measuring biomass P pools in regrowing vegetation in the field.
These investigations of key mechanisms of nutrient accumulation, loss, and retention will help explain field observations of nutrient distributions and transformations. Coupling these ecological studies to investigations of hydrologic flow paths and flow rates will permit quantification and improved understanding of the biogeochemical fluxes in this complicated landscape of changing land use.
