Nitrogen oxide fluxes and nitrogen cycling during postagricultural succession and forest fertilization in the humid tropics

The effects of changes in tropical land use on soil emissions of nitrous oxide (N(2)O) and nitric oxide (NO) are not well understood. We examined emissions of N(2)O and NO and their relationships to land use and forest composition, litterfall, soil nitrogen (N) pools and turnover, soil moisture, and patterns of carbon (C) cycling in a lower montane, subtropical wet region of Puerto Rico. Fluxes of N(2)O and NO were measured monthly for over 1 year in old (more than 60 years old) pastures, early- and mid-successional forests previously in pasture, and late-successional forests not known to have been in pasture within the tabonuco (Dacryodes excelsa) forest zone. Additional, though less frequent, measures were also made in an experimentally fertilized tabonuco forest. N(2)O fluxes exceeded NO fluxes at all sites, reflecting the consistently wet environment. The fertilized forest had the highest N oxide emissions (22.0 kg . N . ha(-1) . y(-1)). Among the unfertilized sites, the expected pattern of increasing emissions with stand age did not occur in all cases. The mid-successional forest most dominated by leguminous trees had the highest emissions (9.0 kg N . ha(-1) . y(-1)), whereas the mid-successional forest lacking legumes had the lowest emissions (0.09 kg N . ha(-1) . y(-1)). N oxide fluxes from late-successional forests were higher than fluxes from pastures. Annual N oxide fluxes correlated positively to leaf litter N, net nitrification potential nitrification, soil nitrate, and net N mineralization and negatively to leaf litter C:N ratio. Soil ammonium was not related to N oxide emissions. Forests with lower fluxes of N oxides had higher rates of C mineralization than sites with higher N oxide emissions. We conclude that (a) N oxide fluxes were substantial where the availability of inorganic N exceeded the requirements of competing biota; (b) species composition resulting from historical land use or varying successional dynamics played an important role in determining N availability; and (c) the established ecosystem models that predict N oxide loss from positive relationships with soil ammonium may need to be modified.