Biophysical controls on nocturnal sap flow in plantation forests in a semi-arid region of northern China
Recent studies have recognized the importance of nocturnal sap flow (Qn) in affecting forest carbon and water budgets and responses to climate change at stand, regional, and global scales. However, biophysical controls on Qn are not fully understood, and their implications for land surface and vegetation models are unclear. We measured growing season sap flow of two widely distributed afforestation species, Pinus tabuliformis and Acer truncatum, in a middle-aged and a young monoculture forest stand, respectively, in a semi-arid mountainous area of northern China. We found a convergence in Qn between the two species and in the proportion of Qn to the total sap flow (12.2–15.0%) across species and ages. Total growing season Qn was higher for middle-aged stands than for young stands because of larger diameters at breast height of older stands. Nighttime vapor pressure deficit (VPDn) influence on Qn of young stands was soil moisture dependent. Nighttime wind speed indirectly controlled Qn through enhancing VPDn in young stands and directly promoted Qn in middle-aged stands with relatively low tree densities. For each species, both increased and decreased soil water content were able to promote Qn in stands with relatively dry soils, which might be due to enhanced nighttime water recharge for capacitance refilling and for avoiding hydraulic failure under prolonged water stress, respectively. Total effects of these three environmental factors explained less than 55% of the Qn variations. This study highlights uncertain physiological influences of VPDn on nighttime stomatal water loss, the nighttime water loss induced by wind, region-specific patterns of nighttime water recharge, and the importance of biotic controls on Qn. Our findings help to improve the existing VPD-based method for partitioning nighttime transpiration and water recharge at tree and stand levels, and suggest the importance of incorporating nocturnal sap flow into large-scale models.