Abstract: We investigated pure Eucalyptus urophylla × E. grandis plantations (PP) and their mixtures with Erythrophleum fordii (EE), Dalbergia odorifera (ED), Castanopsis hystrix (EC), and Parashorea chinensis (EP), and studied soil phosphorus (P) fractions, soil physicochemical properties, fine-root traits of the dominant E. urophylla × E. grandis, microbial biomass, and extracellular enzyme activities in the 0-20 cm layer, as well as their interrelationships. We analyzed the mechanisms underlying the effects of mixed plantations of E. urophylla × E. grandis on soil P fractions. The results showed that, compared with PP, EE and EC significantly increased labile P by 77.7% and 28.2%, moderately labile P by 25.4% and 13.4%, respectively. ED significantly increased moderately labile P by 24.2%, whereas EP significantly increased labile P by 44.3%. Furthermore, compared with PP, EE and EC had significantly higher soil organic carbon, microbial biomass carbon, and activities of C- and N-acquiring enzymes. ED significantly increased soil ammonium and nitrate, accompanied by a significant decline in C- and N-acquiring enzyme activities. EP exhibited significant reductions in both soil total nitrogen and the activities of C- and N-acquiring enzymes. Mixed plantations increased acid phosphatase activity by 24.7%-43.5%. Random forest analysis indicated that soil microbial biomass carbon, fine-root P content and root surface area of the dominant E. urophylla × E. grandis, and N-acetyl-β-glucosaminidase were the primary drivers of labile and moderately labile P. Structural equation modeling indicated that mixed plantations influenced the content and availability of soil P fractions by modulating fine-root traits and soil microbial metabolic traits. Overall, our results indicated that establishing EE and EC mixtures represented particularly effective strategies for improving soil P availability in Eucalyptus plantations.

Key words: mixed plantation, soil phosphorus fraction, root trait, microbial biomass, enzyme activity