Our research further confirmed that the integrated analysis of NMDS and PLS-PM methods effectively elucidates the trends in the divergence of bacterial and eukaryotic β-diversity under soil acidification.

Wang (2025) has expressed concerns regarding our publication in Global Change Biology (Duan et al. 2025). In our study, the first NMDS axis was used as a proxy to assess the shifts in bacterial and eukaryotic community composition (Duan et al. 2025) following peer-reviewed methodologies (Wang et al. 2016; Xu et al. 2021; Zhang et al. 2024). Additionally, a PLS-PM analysis was incorporated into our research study. Wang (2025) noted that our NMDS and PLS-PM analyses indicate opposing β-diversity trends between bacteria and eukaryotes, something that requires additional investigation. In response, we provide further evidence affirming the appropriateness of using the first NMDS axis as an indicator of compositional changes within microbial communities.

Use of soil pH as a key environmental variable for ecology analysis; NMDS based on Bray–Curtis dissimilarity distance, supplemented by Mantel tests, was introduced to evaluate the influence of soil pH on the microbial community composition (Rousk et al. 2010). The results demonstrated that soil pH exerts a pronounced effect on bacterial and eukaryotic community compositions across the entire pH gradient (Figure 1). This relationship is observed through NMDS ordination of pairwise Bray–Curtis dissimilarity distances, so the spatial arrangement of communities along the first axis relative to pH variations is visualized (Figure 1A,D). The Mantel test further revealed statistically significant correlations between soil pH and both bacterial (Spearman's p = 0.28, p = 0.001) and eukaryotic communities (Spearman's p = 0.25, p = 0.001). These patterns were also evident in the regression analyses between soil pH and the first NMDS axis scores for both (R2 = 0.65, p < 0.001) (Figure 1B) communities (R2 = 0.41, p < 0.001) (Figure 1E). Meanwhile, Random Forest analysis identified soil pH as the principal determinant influencing the first NMDS axis scores of both bacterial and eukaryotic communities (Duan et al. 2025). In summary, our study highlights that the first axis of NMDS analysis serves as a reliable indicator of compositional variations for bacterial and eukaryotic communities.

The PLS-PM analysis further supported that soil pH is a dominant variable for bacterial (direct effect = 0.62) and eukaryotic (direct effect = −0.34) community composition (Duan et al. 2025, figure 4B, tables S7 and S8). Moreover, NMDS ordination revealed distinct clustering patterns in bacterial communities under acidic conditions (Figure 1A), whereas eukaryotic communities exhibited greater dispersion (Figure 1D). To validate these findings, a comparative β-diversity analysis between near-neutral (pH 6.5–7.5) and acidic (pH < 6.5) samples was conducted, using the PERMDISP approach, following the methodological recommendations by Wang (2025). PERMDISP analysis demonstrated a statistically significant reduction in bacterial β-diversity in acidic environments compared to near-neutral conditions (perm = 0.037) (Figure 1C). In contrast, eukaryota displayed an opposing trend, with increased β-diversity under acidic conditions (perm = 0.504) (Figure 1F). These results support our prior conclusions that soil acidification leads to a substantial decline in bacterial β-diversity while concurrently promoting an increase in eukaryotic β-diversity. Feng and Deng (2025) have provided an independently commissioned commentary that further validates and supports our findings.

https://doi.org/10.1111/gcb.70396