Environmental and canopy conditions regulate the forest floor evapotranspiration of larch plantations

Background: Integrated forest-water management focusing on forest-water coordination is an important way to alleviate water use conflicts among forests and other sectors in vast dryland regions. Forest floor evapotranspiration (FE), which is an important component of forest evapotranspiration, accounts for a large proportion of the water consumed in arid forests. Elucidating how environmental and canopy conditions impact FE has important significance for guiding integrated forest-water management in a changing environment. Methods: The microlysimeter (ML)-measured evapotranspiration (FEML), reference evapotranspiration (ETo), volumetric soil moisture (VSM), and canopy leaf area index (LAI) were monitored in a Larix principis-rupprechtii plantation located in the semi-humid Liupan Mountains of Northwest China in 2019 (June-September) and 2021 (May-September). The response functions of the FE coefficient (the ratio of daily FEML to ETo) to the individual factors of VSM and LAI were determined using upper boundary lines of scatter diagrams of measured data. The framework of the daily FE (FEML) model was established by multiplying the response functions to individual factors and then calibrated and validated using measured data to assess the FE response to environmental and canopy conditions. Results: (1) The FE coefficient increased first rapidly and then slowly with rising VSM but decreased slowly with rising LAI. (2) The simple daily FE (FEML) model developed by coupling the impacts of ETo, LAI, and VSM in this study performed well for predicting FE. (3) The impacts of ETo, LAI, and VSM were quantified using the FE (FEML) model, e.g., at a given VSM, the impact of ETo on FE increased obviously with decreasing LAI; at a given ETo, the impact of LAI on FE increased with rising VSM. (4) In the two study years, when directly using the microlysimeter measurement, the real FE on the forest floor was overestimated when the VSM in microlysimeters was above 0.215 but underestimated below 0.215 due to the difference in VSM from the forest floor. Thus, the VSM on the forest floor should be input into the FE model for estimating the real FE on the forest floor. Conclusions: The daily FE of larch plantation is controlled by three main factors of environmental (ETo and VSM) and canopy conditions (LAI). The variation in daily FE on the forest floor can be well estimated using the simple FE model coupling the effects of the three main factors and by inputting the VSM on the forest floor into the model to avoid the errors when directly using the microlysimeter measurement with different VSMs from the forest floor. The developed FE model and suggested prediction approach are helpful to estimate the FE response to changing conditions, and to guide forest management practices when saving water by thinning is required.