Soil sample sizes for DNA extraction substantially affect the examination of microbial diversity and co-occurrence patterns but not abundance

As the cornerstone of most soil microbial diversity and abundance research, soil DNA extraction is inherently influenced by soil sample sizes, yet the effects remain far from fully understood. In this study, soils were collected from desert, meadow, forest, and cropland ecosystems to examine the effects of soil sample sizes (i.e., 0.01, 0.025, 0.25, 0.50, and 1.00 g) on prokaryotic and fungal community analysis. Soil microbial abundance and diversity were analyzed through real-time PCR and NovaSeq high-throughput sequencing, respectively. We found that soil sample sizes for DNA extraction generally did not affect microbial abundance determination. In contrast, microbial richness, community composition, and co-occurrence patterns varied with different soil sample sizes. The microbial richness and co-occurrence frequency obtained from 0.01 to 0.025 g of soils were lower than those determined based on 0.25–1.00 g of soils in most ecosystems. Microbial community composition was also affected by the soil sample sizes when they were 0.50 g or less. In particular, small soil sample sizes for DNA extraction (i.e., ≤ 0.025 g) revealed dramatic variations in microbial community profiles among soil microhabitats. Collectively, in most ecosystems, at least 0.25 g of soils should be used to determine the overall diversity of microbes. As for the soils with low microbial cellular density, DNA should be extracted from 0.50 g or even more soils. Thus, the analysis of soil microbial diversity and co-occurrence patterns rather than abundance are substantially affected by the soil sample sizes for DNA extraction, and the choice of optimum soil sample sizes depends on soil types and research purposes.