Lowland tropical rain forests are considered to be a major source of nitrous oxide (N2O) to the atmosphere. Though abiotic factors strongly regulate denitrification rates in these and other ecosystems, the importance of plant functional and chemical diversity in driving local variation is not well understood. This question is especially relevant in lowland tropical forests, like those of the Osa Peninsula of Costa Rica, which are characterized by substantial spatial heterogeneity in canopy foliar nitrogen (N) concentrations. Across a common soil type and using hyperspectral imager based maps of foliar N, we established 0.25 ha plots where foliar N was significantly above or below the regional mean. Compared with low foliar N plots, high foliar N plots displayed significantly elevated soil inorganic N concentrations and cycling rates, and greater inputs of relatively N-rich litter. Thus, we hypothesized that these greater N cycling rates and inputs would support higher denitrification rates in high foliar N plots. We made monthly static chamber measurements measure N2O fluxes in the field, and used reciprocal litter transplants, denitrification enzyme activity, and microbial functional gene abundance analyses to elucidate mechanisms driving possible differences in N2O fluxes across the landscape.
In the field, N2O fluxes were 2-3 times greater in the high foliar N plots than in the low foliar N plots, consistent with our hypothesis. Soil moisture and respiration rates were comparable between plot types, suggesting that differences in N2O fluxes were driven by microbial community composition or substrate inputs rather than abiotic conditions. To test substrate effects, we conducted lab denitrification enzyme activity assays using leachate from high or low N plant litter or high C and N media. Leachate additions generally stimulated higher rates of denitrification in soils from high foliar N plots. However, when supplied with ample C and N substrate, denitrification potentials were highest in soils from low foliar N plots, suggesting an interaction between enzyme abundance and substrate limitation in microbial communities from the two cover types. We conclude that along with well-recognized drivers such as seasonal rainfall variation, canopy foliar N heterogeneity influences in denitrification rates in this tropical forest. Furthermore, because canopy N can be remotely sensed over large areas, the relationship between foliar N and N2O fluxes could be useful in scaling plot-level N2O emissions to the landscape scale.