The observation of high losses of bioavailable nitrogen (N) and N richness in tropical forests is paradoxical with an apparent lack of N input. Hence, the current concept asserts that biological nitrogen fixation (BNF) must be a major N input for tropical forests. However, well‐characterized N cycles are rare and geographically biased; organic N compounds are often neglected and soil gross N cycling is not well quantified. We conducted comprehensive N input and output measurements in four tropical forest types of the Congo Basin with contrasting biotic (mycorrhizal association) and abiotic (lowland–highland) environments. In 12 standardized setups, we monitored N deposition, throughfall, litterfall, leaching, and export during one hydrological year and completed this empirical N budget with nitrous oxide (N2O) flux measurement campaigns in both wet and dry season and in situ gross soil N transformations using 15N‐tracing and numerical modeling. We found that all forests showed a very tight soil N cycle, with gross mineralization to immobilization ratios (M/I) close to 1 and relatively low gross nitrification to mineralization ratios (N/M). This was in line with the observation of dissolved organic nitrogen (DON) dominating N losses for the most abundant, arbuscular mycorrhizal associated, lowland forest type, but in contrast with high losses of dissolved inorganic nitrogen (DIN) in all other forest types. Altogether, our observations show that different forest types in central Africa exhibit N fluxes of contrasting magnitudes and N‐species composition. In contrast to many Neotropical forests, our estimated N budgets of central African forests are imbalanced by a higher N input than output, with organic N contributing significantly to the input‐output balance. This suggests that important other losses that are unaccounted for (e.g., NOx and N2 as well as particulate N) might play a major role in the N cycle of mature African tropical forests.

No abstract

No abstract