Evaluation of rice–legume–rice cropping system on grain yield, nutrient uptake, nitrogen fixation, and chemical, physical, and biological properties of soil

Abstract

To achieve higher yields and better soil quality under rice–legume–rice (RLR) rotation in a rainfed production system, we formulated integrated nutrient management (INM) comprised of Azospirillum (Azo), Rhizobium (Rh), and phosphate-solubilizing bacteria (PSB) with phosphate rock (PR), compost, and muriate of potash (MOP). Performance of bacterial bioinoculants was evaluated by determining grain yield, nitrogenase activity, uptake and balance of N, P, and Zn, changes in water stability and distribution of soil aggregates, soil organic C and pH, fungal/bacterial biomass C ratio, casting activities of earthworms, and bacterial community composition using denaturing gradient gel electrophoresis (DGGE) fingerprinting. The performance comparison was made against the prevailing farmers’ nutrient management practices [N/P2O5/K2O at 40:20:20 kg ha−1 for rice and 20:30:20 kg ha−1 for legume as urea/single super-phosphate/MOP (urea/SSP/MOP)]. Cumulative grain yields of crops increased by 7–16% per RLR rotation and removal of N and P by six crops of 2 years rotation increased significantly (P < 0.05) in bacterial bioinoculants-based INM plots over that in compost alone or urea/SSP/MOP plots. Apparent loss of soil total N and P at 0–15 cm soil depth was minimum and apparent N gain at 15–30 cm depth was maximum in Azo/Rh plus PSB dual INM plots. Zinc uptake by rice crop and diethylenetriaminepentaacetate-extractable Zn content in soil increased significantly (P < 0.05) in bacterial bioinoculants-based INM plots compared to other nutrient management plots. Total organic C content in soil declined at 0–15 cm depth and increased at 15–30 cm depth in all nutrient management plots after a 2-year crop cycle; however, bacterial bioinoculants-based INM plots showed minimum loss and maximum gain of total organic C content in the corresponding soil depths. Water-stable aggregation and distribution of soil aggregates in 53–250- and 250–2,000 μm classes increased significantly (P < 0.05) in bacterial bioinoculants-based INM plots compared to other nutrient management plots. Fungal/bacterial biomass C ratio seems to be a more reliable indicator of C and N dynamics in acidic soils than total microbial biomass C. Compost alone or Azo/Rh plus PSB dual INM plots showed significantly (P < 0.05) higher numbers of earthworms’ casts compared to urea/SSP/MOP alone and bacterial bioinoculants with urea or SSP-applied plots. Hierarchical cluster analysis based on similarity matrix of DGGE profiles revealed changes in bacterial community composition in soils due to differences in nutrient management, and these changes were seen to occur according to the states of C and N dynamics in acidic soil under RLR rotation.