Spatiotemporal dynamics of soil bacterial and fungal communities at a small scale.

BACKGROUND: Soil is a common form of trace evidence in forensic investigations, and its indigenous microbial communities hold promise for geographic provenance determination. However, most existing studies focus on macro-scale soil microbial patterns, with limited understanding of their small-scale spatiotemporal dynamics. This study collected soil samples across multiple time points from a small-scale forest site in Guiyang, China, to characterize bacterial and fungal community distributions. We aimed to quantify the effects of temporal variation and geographic distance on soil bacterial and fungal community structure, and to evaluate their utility as biomarkers for geographic provenance inference. METHODS: Forest surface soil samples were collected from five geographic locations at five time points, with paired environmental and meteorological data recorded in situ. High-throughput sequencing of the 16S ribosomal RNA gene and the internal transcribed spacer 2 region was performed on the MGI DNBSEQ-G99 platform. We integrated alpha and beta diversity analyses, distance-decay models, neutral community models, and random forest algorithms to systematically assess spatiotemporal associations of microbial community assembly and the performance of soil microbes for small-scale geographic provenance prediction. RESULTS: Bacterial Pielou's evenness differed significantly across sampling locations, but alpha diversity of neither bacteria nor fungi varied significantly within the observation window of this study. Permutational multivariate analysis of variance revealed that geographic location was strongly correlated with community structure variation, with no significant temporal effect within the 14-day observation window. Fungal communities exhibited a stronger distance-decay relationship, lower fit to neutral community models, and lower dispersal rate than bacterial communities, indicating stronger deterministic environmental selection. Both bacterial and fungal communities demonstrated exceptional potential for closed-set recognition of predefined sites (accuracy > 92%). CONCLUSION: Soil microbial community structure remains relatively stable over the study period, with spatial heterogeneity far exceeding temporal variability. In small-scale forest habitats with relatively stable vegetation, fungal communities exhibit stronger spatial heterogeneity than bacteria, a pattern likely driven by their high specificity for the co-varying host plants and elevation gradients across the sampling sites. However, inferences across different vegetation types and longer temporal scales require further validation.