Nitrogen cycling in wet tropical forests

Background/Question/Methods

Biogeochemists commonly make the generalization that nitrogen (N) cycles in excess of biological demand in many lowland tropical forests. Evidence supporting this generalization includes: high foliar N concentrations, elevated foliar and soil δ¹⁵N values, large soil NO₃^- pools, soil NO₃^-: NH₄⁺ ratios that exceed 1, and high rates of N loss in both aquatic and gaseous forms. However, most of these data come from the driest third of the tropical forest biome (those forests receiving < 2500 mm annual precipitation). Analyses of data from wetter lowland sites, including our sites in SW Costa Rica, imply that the paradigm of “N excess” does not hold for the wetter fraction of lowland tropical forests. Instead, data show comparatively low foliar and soil δ¹⁵N values, low soil NO₃^- pools, and low values for N losses as N₂O or stream NO₃^-. To better understand the mechanisms that may produce such patterns, we conducted an ¹⁵N pool dilution experiment in a forest on the Osa Peninsula of Costa Rica which receives ~5000 mm of annual precipitation.

Results/Conclusions

Results show high rates of gross N mineralization (9.5 ± 1.8 μg N g soil^-1 d^-1, mean ± 1 SE), suggesting that overall N supply is high and in accordance with previous data showing high rates of litterfall and decomposition in this site. However, while drier lowland forests often display high rates of nitrification relative to mineralization, in this wet site, microbial NH₄⁺ assimilation (5.1 ± 1.0 μg N g soil^-1 d^-1) exceeded rates of gross nitrification (2.6 ± 0.4 μg N g soil^-1 d^-1). In addition, we observed relatively high rates of dissimilatory nitrate reduction to ammonium (DNRA, 0.5 ± 0.1 μg N g soil^-1 d^-1), a mechanism that conserves inorganic N within the ecosystem. We also report strong seasonal differences in rates of the production and likely fate of NO₃^-, where windows of especially wet conditions may drive significant gaseous N losses. Collectively, these data support a more nuanced conceptual model of how N cycles in lowland tropical forests; we suggest that precipitation > 2500 mm y^-1 leads to increasing rates of biological N demand and greater losses of organic N. These factors contribute to low nitrification rates, and lower potential losses to atmospheric and aquatic realms.