Response of the daily transpiration of a larch plantation to variation in potential evaporation, leaf area index and soil moisture

Tree transpiration (T) is a major water budget component and varies widely due to the integrated effects of many environmental and vegetation factors. This study aimed to separate, quantify, and then integrate the effects of the main individual factors, to improve water use estimation and manage the hydrological impacts of forests. A field study was conducted at 3 plots of larch (Larix principis-rupprechtii) plantation in the semi-humid area of the Liupan Mountains, northwest China. The main influencing factors were the atmospheric evaporative demand expressed by potential evapotranspiration (PET), the soil water availability expressed by volumetric soil moisture (VSM) within the 0–100 cm layer, and the canopy transpiration capacity expressed by forest canopy leaf area index (LAI). The daily stand T was estimated through the up-scaling of sap-flow data from sampled trees. It displayed a high degree of scattering in response to PET, VSM and LAI, with an average of 0.76 mm·day(−1) and range of 0.01–1.71 mm·day(−1) in the growing season of 2014. Using upper boundary lines of measured data, the response tendency of T to each factor and corresponding function type were determined. The T increases firstly rapidly with rising PET, VSM and LAI, then gradually and tends to be stable when the threshold of PET (3.80 mm·day(−1)), VSM (0.28 m(3)·m(−3)) and LAI (3.7) is reached. The T response follows a quadratic equation for PET and saturated exponential function for VSM and LAI. These individual factor functions were coupled to form a general daily T model which was then fitted using measured data as: T = (0.793PET − 0.078PET(2))·(1 − exp(−0.272LAI))·(1 − exp(−9.965VSM)). It can well explain the daily T variation of all 3 plots (R(2) = 0.86–0.91), and thus can be used to predict the response of daily T of larch stands to changes in both environmental and canopy conditions.