Conclusion

This meta-analysis indicated that although overall BNEN tended to be lower than BNEA on median basis, there were no significant differences between BNEA and BNEN. Yet, of the ten natural ecosystems studied, BNEN was higher in riparian zones than in boreal forests and tundra. What is more, when comparing the six agricultural land systems assessed, vegetable crop fields and intentional fallow areas showed higher BNEA than cropland, pasture, and rice paddy. The observed increases in BNE in certain ecosystems and agricultural lands can likely be the combined result of increased exogenous N inputs, absence of vegetation for N uptake, accelerated soil N dynamics, or warmer-wetter climatic conditions. Moreover, soil properties such as soil organic C, N, pH, and/or BD were recurrently associated with the outcome of BNE across the assessed terrestrial ecosystems. Interacting effects amongst these controlling factors on BNE showed to be complex and multidirectional, and required additional investigation. Likewise, there is a need to further assess and discern the key soil processes and specific N sources contributing to production or consumption of BNE in a broad variety of natural and agricultural systems. We also noted negative BNE within the compiled global dataset. These reported negative BNE or N₂O uptakes were infrequent and of low magnitude, and hence, their contribution to global N₂O budget can be currently minor. Due to potential bias derived from selection of study sites and use of diverse methodologies for quantifying BNE, it is uncertain whether datasets compiled and analyzed through meta-analyses adequately represent the distribution of BNE from terrestrial ecosystems at global scale. There is a need for additional studies to clarify these unknowns.