Sustainable agriculture requires high yields with minimal environmental impact. Soil microbiota are crucial for nutrient cycling, organic matter decomposition, and plant health, influencing ecosystem productivity. Fertilizer application is a key agricultural practice, but its impact on soil microbiomes remains unclear. While fertilization can increase microbial diversity via nutrient enrichment, excessive application or organic fertilizer-induced enrichment of antibiotics and resistance genes can reduce it. Contrasting findings are attributed to soil type, climate, and vegetation, which impact microbial habitats, substrate availability, and plant-microbe interactions. Legumes, with their symbiotic nitrogen fixation and release of growth-stimulating compounds, are known to promote soil ecological functions. Continuous monoculture can decrease soil quality, while crop rotations enhance substrate diversity and plant growth-promoting microorganisms. This study hypothesized that fertilization differentially affects soil microbiomes in legume versus non-legume systems due to nitrogen's role in regulating soil microbiomes. The study examined long-term fertilization effects on soil bacterial communities in three contrasting cropping systems (continuous alfalfa, continuous winter wheat, and a grain-legume rotation) in a highland region, analyzing soil nutrients, microbial activity, bacterial diversity, composition, co-occurrence networks, and associations with crop productivity.

Existing research highlights the complex relationship between fertilization and soil microbiomes. Studies show that combined chemical and organic fertilizers can increase crop yields by improving soil properties and influencing microbial communities, for example increasing the relative abundance of Bacillus and Flavobacterium in rice paddies. However, results are inconsistent, with some showing no significant changes in microbial diversity, composition, or activity after fertilization. These inconsistencies are largely attributed to differences in soil types and climates, which vary in their physicochemical properties, substrate availability, and vegetation. The inclusion of legumes in cropping systems has been shown to improve soil structure, nutrient availability, microbial diversity, and disease resistance. In contrast, continuous monoculture can lead to soil quality degradation. Prior research lacks comparative analysis of fertilization effects across various cropping systems within the same soil type and climate, hindering a clear understanding of the system-specific responses of soil microbiomes.

A 36-year long-term fertilization experiment was conducted in Changwu County, Shaanxi Province, China. Three cropping systems were studied: continuous alfalfa (AC), continuous winter wheat (WC), and a grain-legume rotation (GLR) of winter wheat-millet-pea-winter wheat. Four fertilization treatments were applied: unfertilized control (CK), phosphorus (P), P and nitrogen (NP), and NP plus manure (NPM). Soil samples were collected from the 0-20 cm layer. Soil physicochemical properties (moisture, ammonium, nitrate, soil organic carbon (SOC), total nitrogen (TN), available phosphorus (OP), pH) were measured. Soil enzyme activities (β-1,4-glucosidase, 1,4-β-D-cellobiohydrolase, β-xylosidase, β-1,4-N-acetylglucosaminidase, L-leucine aminopeptidase, alkaline phosphatase) and potential soil respiration were determined. 16S rRNA gene amplicon sequencing was used to analyze soil bacterial diversity and composition. Statistical analyses included linear mixed-effects models, Wilcoxon rank sum tests, principal coordinate analysis (PCoA), permutational multivariate analysis of variance (PERMANOVA), distance-based redundancy analysis (dbRDA), DESeq2 analysis, co-occurrence network analysis using SparCC, Mantel tests, random forest analysis, variance partitioning, and piecewise structural equation modeling (SEM).

Cropping system significantly affected soil properties and microbial activity. AC had the highest SOC, TN, C/N ratio, ammonium, nitrate, and enzyme activities but the lowest OP. Fertilization increased OP (P treatment) and soil nutrients, enzyme activities, and crop productivity (NPM treatment). The NPM treatment's effects were more pronounced in WC and GLR than in AC. Bacterial α-diversity (Observed ASVs, Shannon, Chao1) was higher in AC and WC than GLR. Fertilization increased α-diversity in AC but decreased it in GLR. NPM had the largest effect on diversity. Bacterial communities differed significantly by cropping system, fertilization treatment, and their interaction. Long-term fertilization did not significantly affect the abundance of core taxa. The NPM treatment resulted in a greater number of fertilization-responsive ASVs (166–263) compared to P and NP treatments (4–129). The number and identity of fertilization-responsive taxa varied across cropping systems. Bacterial co-occurrence network analysis revealed higher complexity and robustness in the AC system. dbRDA and random forest analyses showed that bacterial communities were primarily influenced by OP and N/P in AC, and SOC, TN, and NO3 in WC and GLR. Soil properties, core, and responsive communities significantly contributed to predicting microbial activity and crop productivity. SEMs revealed direct and indirect effects of soil nutrients and microbial communities on microbial activity and crop productivity.

The study's findings support the hypothesis that fertilization differentially impacts soil microbiomes in legume and non-legume cropping systems. The increase in bacterial diversity in the AC system with fertilization, particularly NPM, is likely due to the combined effects of fertilization and legume nitrogen fixation, providing diverse niches for microbes. The lack of effect of fertilization on bacterial diversity in WC and its decrease in GLR highlight the contrasting responses of different cropping systems. The increased number of cropping system-responsive ASVs with fertilization reflects the enhanced impact of plants on soil microbiota under favorable growth conditions. The higher network complexity and robustness in the AC system are likely related to higher nutrient pools and diversity. The influence of soil properties on bacterial communities varied by cropping system, reflecting the differing nutrient demands and stress levels. The strong association between fertilization-responsive taxa and crop productivity underscores the role of specific microbial communities in supporting agroecosystem productivity. The identification of key taxa enriched under fertilization across systems points towards microbial mechanisms contributing to enhanced nutrient cycling and plant growth.

This study demonstrates the contrasting responses of soil microbiomes to long-term fertilization across legume and non-legume cropping systems. Fertilization significantly altered bacterial community composition and diversity, with manure application exacerbating inter-system differences. Fertilization-responsive taxa were strongly linked to crop productivity. The findings highlight the need to consider cropping system-specific responses when assessing the impact of fertilization on soil health and productivity. Future research should focus on manipulative experiments to establish causality and further investigate the functional roles of key taxa in these agroecosystems.

The study's correlational nature limits the establishment of definitive causal relationships between fertilization, microbial community changes, and crop productivity. Further research using manipulative experiments is needed to disentangle direct and indirect effects. The study focused on bacterial communities; the inclusion of other microbial groups (fungi, archaea) would provide a more comprehensive understanding. The long-term nature of the experiment limits the ability to assess short-term responses to fertilization.