Bing Zhang, Mingyu Ma, Dongmei Xue, Hai Wang, Fangyu Gao, Xiaoyun Zhang, Stephen M Mureithi
The results showed that SWC in the 0-5 cm layer exhibited pronounced temporal variability and isotopic enrichment due to evaporation, whereas stability increased with depth. The 20-40 cm layer, constrained by a calcic horizon, served as a stable and isotopically depleted water source during the late growing season. The MixSIAR model exhibited superior performance in quantifying water source contributions compared with the DCM and IsoSource approaches. Both species predominantly utilized water from the 0-10 cm layer during the early and peak growing stages, shifting to water from the 20-40 cm layer when shallow moisture became limited later in the season. Under intensified competition for surface soil water, C. songorica increased its reliance on 10-20 cm water, thereby reducing hydrological niche overlap with S. klemenzii. The proportional similarity (PS) index peaked in July-August, indicating intensified hydrological niche overlap during the period of maximum water demand.
INTRODUCTION: Global climate change alters the spatiotemporal distribution of soil water, reshaping plant water use strategies in water-limited ecosystems. However, seasonal shifts in plant water-source use and uncertainties associated with different water-source partitioning approaches remain insufficiently quantified in desert steppe ecosystems.
METHODS: Using stable hydrogen and oxygen isotopes combined with soil water content (SWC), we investigated the seasonal water sources of Stipa klemenzii and Cleistogenes songorica in a desert steppe in Inner Mongolia from May to September 2025. Three approaches, the Direct Comparative Method (DCM), IsoSource, and MixSIAR models, were applied to identify and quantify plant water sources.
RESULTS: The results showed that SWC in the 0-5 cm layer exhibited pronounced temporal variability and isotopic enrichment due to evaporation, whereas stability increased with depth. The 20-40 cm layer, constrained by a calcic horizon, served as a stable and isotopically depleted water source during the late growing season. The MixSIAR model exhibited superior performance in quantifying water source contributions compared with the DCM and IsoSource approaches. Both species predominantly utilized water from the 0-10 cm layer during the early and peak growing stages, shifting to water from the 20-40 cm layer when shallow moisture became limited later in the season. Under intensified competition for surface soil water, C. songorica increased its reliance on 10-20 cm water, thereby reducing hydrological niche overlap with S. klemenzii. The proportional similarity (PS) index peaked in July-August, indicating intensified hydrological niche overlap during the period of maximum water demand.
DISCUSSION: These findings demonstrate depth- and season-dependent water partitioning among dominant herbaceous species in the desert steppe and highlight the importance of Bayesian isotope mixing models for quantifying plant water sources under changing precipitation regimes. The results provide insights into the adaptive water-use strategies of desert steppe plants in response to future climate change.