Substituting chemical fertilizer with organic fertilizer is an important agricultural practice that improves crop yields but also affects soil biogeochemical cycling. To explore the effects of organic fertilizer on soil properties and bacterial community structure and diversity, a long-term field experiment (2011-2017) was conducted in a wheat-rice rotation system with four treatments: control (CK, no fertilizer), chemical fertilizer only (NPK), substitution of 50% of chemical nitrogen (N) fertilizer with organic fertilizer (NPKM), and substitution of 100% of chemical N fertilizer with organic fertilizer (OM). Results showed that, crop yields (wheat and rice) from 2011-2017 were highest in NPKM. Combined with alpha-diversity analysis, long-term application of NPK led to a decrease in soil bacterial diversity. However, soil bacterial community richness (Ace and Chao1 indices) increased in NPKM in the wheat season, as well as in OM in the rice season. Compared with NPK, organic fertilizer inputs (NPKM and OM) increased microbial biomass carbon and abundance of dominant bacteria (Acidobacteria, Anaerolineaceae, and Nitrospira) in both crop seasons. In NPK, relative abundance of Rhodanobacter increased significantly in the wheat season because of low soil pH. Bacterial community composition and structure were similar between NPKM and OM in the wheat season. Redundancy analysis further found that soil pH (R2 = 0.735, P = 0.013) co-varied with the bacterial community structure in the wheat season. In the rice season, relative abundance of Nitrospira increased significantly in NPKM, which could generate large amounts of nitrate nitrogen (NO3- -N), and then increase risks of N2O emissions and NO3- -N leaching in paddies. Overall, partial substitution of chemical fertilizer with organic fertilizer was optimal to increase crop yields and recover soil bacterial diversity in the wheat-rice rotation, and greater substitution with organic fertilizer in combination with chemical fertilizer is recommended only for rice seasons.
