Determining the independent impact of soil water on forest transpiration: A case study of a black locust plantation in the Loess Plateau, China

Transpiration (Tr) is influenced by environmental factors, vegetation properties, and anthropogenic management. The effect of environmental factors on Tr are taking place from two aspects: evaporative demand (i.e., potential evapotranspiration, PET) and water supply (i.e., soil water). Soil water is one of the most important factors that limit plant transpiration in terrestrial ecosystems, especially in semi-arid and arid regions. Investigating the relationship between Tr and soil water is crucial for an improved understanding of plant survival strategies and predicting hydrological cycles and water resources under climate change. Although the relationships of soil water and Tr have been widely studied, the independent effects of soil water on Tr are difficult to separate because soil water and PET occur concurrently under natural conditions. This study carried out field observations of sapflow density, meteorological factors and soil water in a black locust (Roblin:a pseudoacacia) plantation in the semi-arid Loess Plateau of China. This information was used to develop a model integrating the daily PET and relative extractable soil water (REW, an index that represents the available soil water): Tr = (0.27 x PET - 0.02 x PET2 - 0.32) x (1 - e(-5.70 x) (REW)). The model fitted the measured data well (R-2 = 0.65 and RMSE = 0.06 mm day(-1)). We found that the daily Tr increased as the REW increased under varying PET levels. Additionally, the independent effects of soil water on Tr were analysed using the factorial experiment analysis method. The REW was manipulated and PET varied naturally during the measurements to separate the independent effects of the REW on Tr. The results showed that Tr increased with the REW during the study period at a rate of 0.53 mm day(-1) per 0.1 REW when the REW < 0.4 and 0.09 mm day(-1) per 0.1 REW beyond the threshold (REW = 0.4). Tr increased by 27.3 mm (43.7%) compared to the controlled Tr (observed PET and lowest REW) due to the effects of increasing REW (ranging from 0.14 similar to 0.81) during the study period (growing seasons of 2015 and 2016). Furthermore, canopy-level stomatal conductance was calculated using a simplified inverted Penman-Monteith equation. The reference canopy-level stomatal conductance (Gsref), which represents the Gs capacity and its sensitivity to soil water, showed an increasing response with increasing REW, which is the possible reason for the Tr dynamics with the REW. Our study provided an improved understanding of the soil water-Tr relationship and predicted effects of climate change on tree water use. Although the method required abundant field observation data, we have provided a feasible method of accurately quantifying the independent contribution of soil water on forest Tr.