Apparent Short-term Stability of Forest Soil Resistomes under Litter Manipulation is Associated with Microbial Communities and Mobile Genetic Elements

Forest ecosystems are important environmental reservoirs of antibiotic resistance genes (ARGs), yet the effects of plant litter inputs on soil resistomes remain poorly understood. Here, we combined a litter manipulation experiment (with litter, WL; no litter, NL) with metagenomic sequencing to investigate ARG dynamics across three forest stands (coniferous pure, CP; broad-leaved pure, BP; and coniferous mixed, CM) during an approximately three-month experimental period. We identified 43 antibiotic drug classes and 1,595 ARG subtypes. The litter treatment was associated with significant changes in soil physicochemical properties, particularly nutrient availability, with the strongest effects observed in the CM stand. However, within stand comparisons revealed no significant differences in microbial diversity indices between WL and NL treatments, despite a significant main effect of litter treatment on the Shannon index in the two-way ANOVA ( p < 0.05). Notably, total ARG abundance and ARG Shannon diversity did not differ detectably between WL and NL within any forest stand. A descriptive beta-distribution abundance occupancy model adapted from the Sloan framework showed similar descriptive relationships under the two treatments. Integrating co-occurrence networks, generalized additive models, and partial least squares path modeling, we found that ARG abundance was more strongly associated with microbial community attributes and MGE abundance than with soil variables. Collectively, our results indicate limited detectable short-term changes in ARG metrics despite treatment-associated changes in soil conditions. These findings indicate apparent short-term stability of resistome metrics at a single sampling endpoint and comparatively stronger associations of ARG abundance with biotic variables than with abiotic soil variables.