Long-term nitrogen fertilization does not necessarily suppress soil nitrogen mineralization via enzymatic pathways

Nitrogen (N) fertilization supports high yields of cereal crops, but the extent to which such exogenous N inputs may impact soil N provisioning through enzyme-catalyzed depolymerization and mineralization is contested. We tested the hypothesis that exogenous N inputs suppress soil N mineralization by leveraging a long-term field experiment in an Argiudoll representative of the U.S. Midwest. We quantified: (1) total organic carbon (TOC) and N, (2) potentially mineralizable N (PMN) as a proxy for soil N provisioning capacity, (3) microbial biomass C (MBC) and N (MBN), and (4) activities of 12 N-hydrolytic enzymes as proxies for N depolymerization-mineralization potential rates across a 42 y urea-based fertilization gradient of 0, 67, 135, 202, and 269 kg N ha −1 . Soil PMN, MBN, TN and N-hydrolytic enzyme activities, except for two aminopeptidase and urease activities, as well as MBC, TOC and C: N were unaffected by N fertilization (i.e., 0 vs 69–269 kg ha −1 ). Under N fertilization, we found weak evidence for suppression of N mineralization: relative to the lowest N fertilization rate (67 kg ha −1 ), PMN was 23% lower ( p = 0.104) at 269 kg ha −1 and four of eight aminopeptidase activities, either on soil mass or TOC basis, were 24–284% lower ( p = 0.031–0.091) at 202 kg ha −1 than the lowest N fertilization rate (67 kg ha −1 ), suggesting partial suppression of an intermediate step of multistep N depolymerization. Despite fertilization with urea, urease activity was 78–84% lower ( p = 0.019) at 202 kg N ha −1 than 0 and 67 kg N ha −1 . Overall, we do not find evidence that N fertilization across a wide range of rates suppresses soil N depolymerization and mineralization enzyme activities, nor potentially mineralizable N, in the long-term in maize-based agroecosystems situated on Mollisols that dominate the US Midwest. Given the reliance of the over 40 million ha of maize-based production on exogenous N inputs in the U.S. Corn Belt, our results suggest limited negative impacts on soil N mineralization in the long-term. • Soil N cycling was evaluated in a 42 y N fertilization gradient (0–269 kg ha −1 ). • Fertilization (0 vs 67–269 kg ha −1 ) did not impact N pool sizes or process rates. • Potentially mineralizable N was unaffected by N fertilization ( p = 0.12). • Certain soil aminopeptidase activities were lower at 202 than at 67 kg N ha −1 . • Soil protease, deaminase and chitinase activities unaffected by N fertilization.