Effects of Ageratina adenophora Invasion on Soil Nutrients, Enzyme Activities, and Microbial Composition/Function in Karst Areas of Central Guizhou

This study elucidates the regulatory effects of Ageratina adenophora invasion on soil nutrient cycling, enzyme activities, and microbial composition/function in karst ecosystems of central Guizhou and reveals the plant–soil–microbe feedback-driven invasion loop. Four invasion gradients were established in Guanling County: CK (uninvaded), L (10–30% cover), M (30–60%), and H (60–90%). Soil nutrients, enzyme activities, metagenomic profiles, and nutrient contents in vegetative organs of A. adenophora and native Artemisia argyi were analysed. Key results: (1) Soil microbial communities included 7 domains, 198 phyla, 179 classes, 364 orders, 856 families, 3,531 genera, and 22,111 species. No significant differences were detected in soil nutrients, enzyme activities, or microbial composition/structure/diversity across invasion gradients. (2) Invaded soils presented significantly greater soil organic carbon (SOC), total nitrogen (TN), and pH than CK soils. Sucrase and urease activities were significantly elevated. Compared with A. argyi leaves, A. adenophora leaves contained significantly greater TN and total potassium (TK). (3) Microbial Shannon diversity ( H' ) and Pielou evenness ( J ) were significantly decreased. Verrucomicrobia and Candidatus Eisenbacteria were significantly enriched in invaded vs. CK soils. (4) Functional gene (K13038 and K00639) levels were significantly correlated with invasion intensity. Carbohydrate metabolism genes ( GT10 and CBM4) were upregulated by 1.5- and 1.7-fold, respectively; the stress resistance gene merG was upregulated 2.3-fold. These changes collectively demonstrate that A. adenophora invasion increases SOC/TN content and sucrase/urease activity, restructures microbial communities, and upregulates carbohydrate metabolism/stress resistance-related genes, establishing a self-reinforcing "invasive plant‒soil–microbe" feedback loop in karst ecosystems.