Background and aims Mountain ecosystems in drylands are generally constrained by phosphorus (P) availability, depending on the abundance and composition of soil inorganic and organic P (Pi and Po) fractions. However, shifts in P fractions and their edaphic drivers across vegetation types along elevational gradients in dryland mountain ecosystems remain poorly understood. Methods Using topsoil (0–20 cm) samples collected from four primary vegetation types at different elevations in the Qilian Mountains of China’s drylands: grassland (2700–2800 m), spruce forest (2900–3000 m), subalpine shrubland (3400–3500 m), and alpine shrub-meadow (3600–3700 m), we investigated changes in different bioavailable P fractions, including labile Pi (resin-Pi + NaHCO3-Pi) and Po (NaHCO3-Po), moderately labile Pi (NaOH-Pi) and Po (NaOH-Po), and relatively stable Pi (HCl-Pi) and Po (HCl-Po), and their relationships with soil biogeochemical properties (clay content, pH, alkaline phosphatase activity, exchangeable cations, and reactive Fe/Al oxides). We evaluated the individual and interactive effects of edaphic variables on P fractions using boosted regression trees model and path analysis. Results Soil Po fractions showed consistent changes across the sites, with their concentrations peaking in the alpine shrub-meadow and being lowest in the spruce forest, whereas Pi fractions exhibited distinct shifting patterns with their maximum and minimum concentrations occurring in variable vegetation types. Edaphic variables explained most of the variation in Pi (76–87%) and Po (65–81%) fractions, but they exhibited complex mechanisms governing the dynamics of P fractions. Notably, alkaline phosphatase activity negatively affected Po concentrations but positively influenced Pi concentrations, whereas amorphous Fe/Al oxides negatively impacted labile Pi but enhanced Fe/Al-associated P storage. Conclusions Our results reveal contrasting distribution patterns of Pi and Po fractions across vegetation types and highlight the crucial roles of soil biogeochemical drivers in regulating multi-pool Pi and Po dynamics in mountain ecosystems.

Soil biogeochemical controls on the distribution of phosphorus fractions in the Qilian Mountains of China’s drylands

期刊论文
发表日期：

2026年1月17日

查看全文

Zeng-Ru Wang,
Ji-Liang Liu,
Feng-Rui Li

摘要

Background and aims

Mountain ecosystems in drylands are generally constrained by phosphorus (P) availability, depending on the abundance and composition of soil inorganic and organic P (P_i and P_o) fractions. However, shifts in P fractions and their edaphic drivers across vegetation types along elevational gradients in dryland mountain ecosystems remain poorly understood.

Methods

Using topsoil (0–20 cm) samples collected from four primary vegetation types at different elevations in the Qilian Mountains of China’s drylands: grassland (2700–2800 m), spruce forest (2900–3000 m), subalpine shrubland (3400–3500 m), and alpine shrub-meadow (3600–3700 m), we investigated changes in different bioavailable P fractions, including labile P_i (resin-P_i + NaHCO₃-P_i) and P_o (NaHCO₃-P_o), moderately labile P_i (NaOH-P_i) and P_o (NaOH-P_o), and relatively stable P_i (HCl-P_i) and P_o (HCl-P_o), and their relationships with soil biogeochemical properties (clay content, pH, alkaline phosphatase activity, exchangeable cations, and reactive Fe/Al oxides). We evaluated the individual and interactive effects of edaphic variables on P fractions using boosted regression trees model and path analysis.

Results

Soil P_o fractions showed consistent changes across the sites, with their concentrations peaking in the alpine shrub-meadow and being lowest in the spruce forest, whereas P_i fractions exhibited distinct shifting patterns with their maximum and minimum concentrations occurring in variable vegetation types. Edaphic variables explained most of the variation in P_i (76–87%) and P_o (65–81%) fractions, but they exhibited complex mechanisms governing the dynamics of P fractions. Notably, alkaline phosphatase activity negatively affected P_o concentrations but positively influenced P_i concentrations, whereas amorphous Fe/Al oxides negatively impacted labile P_i but enhanced Fe/Al-associated P storage.

Conclusions

Our results reveal contrasting distribution patterns of P_i and P_o fractions across vegetation types and highlight the crucial roles of soil biogeochemical drivers in regulating multi-pool P_i and P_o dynamics in mountain ecosystems.