<p>Understanding forest management impacts on understory vegetation and belowground processes is essential for sustainable forestry. This study employed a controlled and replicated field experiment to investigate how four forest management practices—three active practices (TTFM: removal of unwanted competitors of selected target trees; SBFM: optimization of forest spatial structure through selective thinning; SFCS: reduction of stand density via timber harvesting) and an unmanaged control (NT)—affect forest structure, light regimes, understory vegetation, soil nutrients, enzymes and examines their interrelationships. Current study was conducted in a natural oak forest within the Xiaolongshan Nature Reserve, Gansu Province, China, where the different management practices have been implemented since 2013. The design consisted of four randomly assigned replicate plots per treatment (16 plots in total). Within each plot, composite soil samples were collected from three depths (0–10, 10–20, and 20–30 cm) using a randomized five-point sampling method. Redundancy Analysis was employed to elucidate the key factors governing soil nutrient and enzyme activities. SBFM significantly enhanced soil acid phosphatase activity at 20–30 cm. Furthermore, both SBFM and TTFM reduced the available phosphorus content at 10–20 cm compared to the NT. Tree mingling, dominance, shrub Shannon diversity, and shrub and herb nitrogen and phosphorus content emerged as primary drivers of soil nutrient and enzyme dynamics. These findings demonstrate that SBFM enhances soil nutrient and enzyme processes. Forest structure-light coupling and understory vegetation jointly regulates soil nutrients and enzymes. This study provides guidance for improving forest management toward ecosystem sustainability.</p>