Deciphering spatial variability and dominant controls of soil quality under four types of grassland along a southeast-northwest transect in Tibet, southwestern China

Abstract

Grasslands on the Qinghai-Tibetan Plateau are under threat from ongoing degradation due to natural and anthropogenic perturbations, resulting in widespread decline in the soil quality of grassland. However, the variations, distributions and dominant drivers of soil quality under different grassland types on the plateau have not been studied due to great difficulty in large-scale sampling at high altitudes. We quantitatively assessed soil quality under alpine meadow (AM), alpine steppe (AS), alpine desert steppe (ADS), and alpine desert (AD) along a 900-km southeast-northwest transect in Tibet, southwestern Qinghai-Tibetan Plateau using an improved soil quality index (SQI) approach. Results showed that there was a high spatial variability of SQI across the transect, with an overall increase of SQI from northwest to southeast (P < 0.01). The average SQI of AM (0.51 ± 0.12) and AS (0.47 ± 0.13) was significantly higher than that of ADS (0.33 ± 0.10) and AD (0.30 ± 0.10) (P < 0.05). Random forest analysis demonstrated that soil property, vegetation factors, environmental conditions and human activity exhibited a decreasing order of the relative importance on SQI. Interactions of soil property and vegetation factors significantly influenced SQI as indicated by the structural equation modeling (P < 0.05). Pedotransfer function (PTF) was applied to adequately predict the soil quality of grassland in Tibet (RMSE = 0.087, R2 = 0.647, P < 0.01), but an extrapolation of the established PTF for estimating grassland soil quality in another alpine region in northeastern Qinghai-Tibetan Plateau requires the nonlinear least squares correction to improve the prediction accuracy. This study unravelled the spatial variability and dominant controls of the soil quality of grassland in Tibet and established a high-precision forecasting model of SQI by considering the grassland type, providing a basis for accurately predicting and sustainably managing grassland soil quality in the alpine region. Reducing grazing intensity and vegetation restoration are needed to improve the soil quality of grassland on the Qinghai-Tibetan Plateau.