Effects of Nitrogen Form and Rate on Soil Physicochemical Properties and the nirS-Type Denitrifying Bacterial Community Under Straw Addition

Denitrification is a major pathway of reactive nitrogen loss from cold-temperate farmland, but how mineral N form and rate interact with temperature to shape the nirS-type denitrifying bacterial community under straw addition is poorly resolved. Maize straw was sealed in mesh bags and buried at a fixed depth in a composite farmland soil from Jiamusi, Heilongjiang Province, and incubated for 240 d through constant-temperature phases of 5, 20, 15 and −15 °C (60 d each). Five treatments were compared: straw without mineral N (CK), straw plus full- or half-rate ammonium sulfate (A1, A2), and straw plus full- or half-rate potassium nitrate (B1, B2). The nirS-type community, soil chemical properties, and potential enzyme activities were measured at the end of each phase. Community structure differed primarily among the sequential incubation phases. Phase labels the constant-temperature sequence and is not separated from incubation time, straw decomposition or preceding thermal history. Richness of the DNA-detected library was lowest after the 5 °C phase and higher later in the sequence, as was the number of ASVs shared by all treatments. Shannon and Simpson diversities differed among treatments only after the 5 °C phase, when A1 was lower than CK, A2, B1 and B2. Community composition differed among treatments after the 5 °C phase (p = 0.042), the 15 °C phase (p = 0.001) and the −15 °C phase (p = 0.001), but not after the 20 °C phase (p = 0.060). Rhodanobacter accounted for most nirS reads in most samples, but its relative abundance was lower, and that of Herbaspirillum and other Burkholderiales higher, under A2 after the 5 °C phase. These are proportional changes in the sequenced library, not measured changes in gene copies or activity. Soil properties showed significant phase × treatment interactions. Redundancy analysis described covariation between community structure and invertase and N-related variables after the 20, 15 and −15 °C phases, but not after the 5 °C phase. Mineral N was associated with changes in nirS-type community structure after the 5, 15 and −15 °C phases, but not after the 20 °C phase. Full-rate ammonium sulfate was associated with lower nirS evenness only after the 5 °C phase; reduced ammonium and nitrate additions were associated with phase-dependent shifts in relative abundance. Denitrification potential and N2O were not measured, so these compositional shifts cannot be read as a change in denitrification rate or N loss. Whether matching N form and rate to the early and sub-zero phases would alter field N loss is a hypothesis for trials that measure application timing, crop N uptake and gaseous losses.