# Divergent patterns and drivers of water use efficiency across leaf and ecosystem scales in the Horqin Sandy Land

## TL;DR

The study examines how water use efficiency varies in two ecosystems in the Horqin Sandy Land and how environmental factors influence these patterns.

## Contribution

The study reveals divergent drivers of water use efficiency at leaf and ecosystem scales, emphasizing the importance of abiotic factors and ecosystem structure.

## Key findings

- WUEleaf in SMAH closely tracks soil moisture, while WUEeco lags due to low transpiration to evapotranspiration ratios.
- WUEeco in MPA aligns with groundwater level and temperature, contrasting with leaf-scale responses.
- WUEeco is influenced by vegetation–environment interactions rather than vegetation type alone.

## Abstract

Water use efficiency (WUE) is crucial in the fragile Horqin Sandy Land, yet how global change and ecosystem traits influence its patterns remains unclear.

This study focused on two ecosystems in the Horqin Sandy Land—semi-mobile dunes with Artemisia halodendron (SMAH) and a meadow wetland dominated by Phragmites australis (MPA). Using plant carbon isotope ratios and eddy covariance data from the 2022–2023 growing seasons, we revealed the distribution patterns of leaf-scale (WUEleaf) and ecosystem-scale (WUEeco) WUE and identified their environmental regulatory mechanisms.

The results showed that the variation trend of WUEleaf in SMAH closely tracked soil moisture dynamics, whereas WUEeco lagged behind WUEleaf, demonstrating that a low T/ET (transpiration to evapotranspiration ratio) plays a key “attenuator” role. In MPA, WUEeco variation aligned with groundwater level and temperature. Overall, WUEleaf and WUEeco responded oppositely to key drivers, but were both abiotic-dominated. WUEleaf was influenced mainly by soil temperature (with an optimum of 22.8°C) and soil organic carbon, while WUEeco was driven primarily by net radiation and vegetation cover.

WUEeco variation arose from vegetation–environment interactions rather than vegetation type alone. Thus, WUEeco cannot be directly extrapolated from WUEleaf. These results underscore aligning WUE metric selection with research objectives in water–carbon coupling studies, highlighting the role of ecosystem structure and functional traits in regulating water–carbon dynamics.