Effects of Photovoltaic-Associated Microsite Heterogeneity on Soil Bacterial Community Structure in a Grassland

As solar photovoltaic (PV) infrastructure expands across open landscapes, understanding its belowground effects in grassland ecosystems is important for sustainable solar development. PV panels can create spatially heterogeneous microsites by altering light availability, precipitation redistribution, and soil environmental conditions, but their effects on soil bacterial community structure remain insufficiently understood. Here, we investigated soil physicochemical properties and bacterial communities across four PV-associated microsites: undisturbed grassland outside the PV array (Control), the front edge of the panel (FE), beneath the center of the panel (BP), and the uncovered interspace between adjacent panel rows (IS). Soil bacterial communities were characterized using 16S rRNA gene sequencing. Most measured soil physicochemical properties did not differ significantly among microsites. Available potassium was the only variable showing an overall microsite effect and was significantly lower in FE and BP than in the Control. Shannon index and Chao1 richness did not differ significantly among microsites. In contrast, Bray–Curtis-based principal coordinate analysis (PCoA) showed that FE samples occupied a relatively distinct ordination position, whereas Control, BP, and IS samples partially overlapped; permutational multivariate analysis of variance (PERMANOVA) indicated a significant overall difference in bacterial community composition among microsites. At the phylum level, no dominant bacterial phylum differed significantly among microsites after false discovery rate (FDR) correction. Redundancy analysis (RDA) indicated that alkali-hydrolyzable nitrogen and available potassium were the measured soil variables most closely associated with bacterial community composition at the operational taxonomic unit level in the final model. Together, these results indicate that differences among PV-associated microsites were detected in overall bacterial community composition but not in Shannon index or Chao1 richness. Alpha-diversity and community-composition analyses therefore provide complementary information for evaluating belowground bacterial responses to PV development in grasslands.