Changes in soil phosphorus fractions associated with altered vegetation and edaphic conditions across a chronosequence of revegetated dunes in a desert area

Revegetation through the establishment of woody vegetation on bare active dunes is an effective approach to restore degraded dune systems. Revegetation can affect soil inorganic and organic phosphorus (Pi and Po) cycling and storage by modifying biotic and abiotic conditions (vegetation characteristics and soil properties, respectively). However, how shrub revegetation and revegetation age affect soil P fractions and how biotic and abiotic factors interact to mediate revegetation effects on P fractions are poorly understood. To address these questions, we conducted a field survey in an artificially established chronosequence of revegetated dunes in a desert area of China, with five differently aged revegetated sites (19, 32, 46, 55, and 63 years) and an adjacent non-revegetated control site. We investigated changes in soil Pi and Po fractions of decreasing lability (labile P: resin Pi and NaHCO3-extractable Pi/Po; moderately labile P: NaOH-extractable Pi/Po; recalcitrant P: HCl-extractable Pi/Po and residual P) and vegetation and edaphic properties as predictors. Boosted regression trees analysis was used to evaluate the contributions of individual predictors to variability in P fractions, and structural equation modeling was used to determine the interactive effects of multiple predictors. Concentrations of soil P fractions were 65–114% greater in revegetated sites compared with the non-revegetated control, but the magnitude of this effect varied with revegetation age. Furthermore, soil total P and its fraction concentrations increased with revegetation age, but the accumulation rates of total P and most P fractions were greatest during the first 19 years after revegetation. Revegetation-induced soil total and available P increases were caused by multiple factors, including enhanced P inputs from dust deposition and plants due to increased vegetation cover and surface litter and root biomass, and enhanced Po and Pi mobilization due to increased soil alkaline phosphatase activity and decreased soil pH. Edaphic properties exerted greater controls over the spatiotemporal dynamics of soil P than vegetation changes, but the key influential factors differed among P fractions. We conclude that revegetation significantly increased the labile, moderately labile, and non-labile P pools and that the effects of revegetation were fraction- and time-dependent, with important implications for managing soil P cycling in revegetated desert ecosystems.