Human activities, particularly agricultural intensification, are causing fundamental ecological changes across various spatial scales. While research has focused on the impact of land-use change on local species diversity, the broader issue of biotic homogenization at larger scales is equally important. Biotic homogenization, the increase in taxonomic or functional similarities among spatially distributed ecological communities, can be driven by various factors, including the introduction of exotic species and the loss of native species. Previous studies have primarily focused on aboveground habitats, leaving a critical knowledge gap regarding the impacts on belowground communities, which are crucial for ecosystem services and conservation. Land-use change and agricultural conversion alter community assembly processes, composition, and species diversity through species extinction, colonization, and shifts in relative abundance. Intense agriculture can lead to soil degradation (compaction, salinization, acidification, etc.), resulting in structural shifts in microbial taxonomic and functional composition. Existing research is predominantly site-specific and lacks a generalizable understanding of how soil microbial profiles respond to agricultural conversion across diverse soil and climate types. This study addresses whether agricultural effects lead to taxonomic and functional biotic homogenization of soil microbiomes at large spatial scales and how land-use changes alter soil microbial community composition and functions across a wide range of soil and climate types.

A substantial body of research indicates that intensive agriculture reduces biodiversity across various taxa and ecosystems. Studies have highlighted the loss of local species diversity due to agricultural expansion, but the phenomenon of biotic homogenization at larger spatial scales has received less attention. Biotic homogenization, characterized by increased similarity in community composition across geographic locations, is a significant concern for ecosystem services and conservation. While research has documented the homogenizing effects of agriculture on aboveground communities, studies focusing on belowground microbial communities remain limited. Understanding the impact of agricultural conversion on soil microbial communities is essential for developing effective conservation strategies. The consequences of agricultural land use changes on soil properties (compaction, nutrient imbalance, etc.) and their implications for microbial community structure and function have been investigated in specific locations, but a global perspective is needed to generalize the findings.

This study used a two-pronged approach to investigate the global impacts of agricultural land conversion on soil microbial communities: a continental soil survey and a global meta-analysis. The continental survey involved the collection of 1185 soil samples from 44 regions across China, encompassing croplands and adjacent natural ecosystems (forests, grasslands, wetlands). Samples were collected during the crop growing season, and soil physicochemical properties were measured. DNA was extracted from the samples, and 16S rRNA gene amplicon sequencing was conducted to characterize bacterial community composition. A subset of samples underwent shotgun metagenomic sequencing to analyze functional gene content. The global meta-analysis involved a comprehensive literature search to identify studies with paired sequencing data from agricultural and natural ecosystems across multiple continents. Raw sequencing data were processed and analyzed using bioinformatics tools. Statistical analysis included linear mixed-effects models (LMMs) to assess agricultural impacts on microbial communities, principal coordinate analysis (PCoA) to visualize community dissimilarities, and PERMANOVA to test for significant differences among ecosystems. Mantel correlations examined relationships between soil properties and fungal communities. Structural equation modeling (SEM) was used to investigate the causal relationships among agricultural impacts, soil and fungal variables, and soil functions. To estimate the relative importance of five community assembly processes (homogeneous selection, variable selection, dispersal limitation, homogenizing dispersal, and undominated assembly), the β-nearest taxon index (βNTI) and Raup-Crick dissimilarity were calculated.

The continental survey and global meta-analysis consistently demonstrated that land conversion to agriculture leads to taxonomic homogenization of soil bacterial communities. Beta-diversity, a measure of community turnover across sites, was significantly lower in croplands than in natural ecosystems (forests, grasslands, wetlands). Croplands exhibited increased similarity in microbial composition compared to paired natural ecosystems. This homogenization was driven primarily by the increased geographic range of taxa prevalent in croplands, indicating the expansion of generalist species and the decline of habitat-specific taxa. Although functional composition showed less pronounced homogenization, specific functional genes involved in nutrient cycling (nitrogen fixation, phosphorus mineralization, and transportation) were depleted in croplands compared to natural ecosystems. Analysis of metagenomic data showed that while there were no significant differences in overall carbon metabolism, there were shifts in specific carbon-degrading genes. Several phyla (Chloroflexi, Gemmatimonadota, Planctomycetota, Myxcoccota, and Latescibacterota) were significantly enriched in croplands. The relative importance of homogeneous selection in shaping microbial community assembly was higher in croplands than in natural ecosystems. Soil pH was identified as a major driver of bacterial community composition, and interactions with fungal communities (saprotrophs and pathogens) also played a significant role. Links were observed between bacterial taxonomic composition and soil functions, but these links were not as strong for functional composition, suggesting functional redundancy in soil microbiomes. Specifically, the abundance of Firmicutes and Actinobacteria, known for spore-forming ability, was lower in croplands, suggesting a shift in microbial dormancy strategies.

The findings strongly support the hypothesis that agricultural conversion induces taxonomic and functional homogenization of soil microbiomes across large spatial scales. The observed homogenization, characterized by a reduction in beta-diversity and the dominance of cosmopolitan taxa, has significant implications for ecosystem services. The depletion of functional genes involved in nutrient cycling suggests a potential decline in soil fertility and ecosystem resilience. While the functional redundancy of soil microbiomes offers some buffering against changes, the observed loss of nitrogen fixation and phosphorus mineralization potential highlights the risks of agricultural intensification. The strong influence of soil pH and fungal interactions on bacterial community composition underscores the need to consider these factors in managing agricultural land. These findings reinforce the urgent need to consider biotic homogenization when assessing the sustainability and health of agricultural practices.

This study provides compelling evidence that agricultural land conversion leads to significant taxonomic and functional homogenization of soil microbial communities on continental and global scales. This homogenization poses a threat to soil health and ecosystem services, highlighting the need for sustainable agricultural practices and ecological restoration to conserve biodiversity. Future research could focus on more detailed investigations of functional gene expression to further understand the functional consequences of biotic homogenization and explore the potential for manipulating microbial communities to enhance soil functions in agricultural systems.

While this study provides a comprehensive overview, several limitations should be considered. The global meta-analysis relied on publicly available data, which may not be uniformly distributed across different regions and ecosystems, potentially introducing biases in the results. The sampling strategy, while extensive, might not fully capture the heterogeneity of soil microbial communities within and across ecosystems. Furthermore, reliance on 16S rRNA gene amplicon sequencing for characterizing bacterial communities may have missed the diversity of unculturable or rarely sequenced organisms. Future studies might benefit from incorporating more sophisticated techniques to assess microbial diversity.