The energy-limited water loss of an alpine shrubland on the northeastern Qinghai-Tibetan Plateau, China

Alpine shrubland on the northeastern Qinghai-Tibetan Plateau. Water provision ability is a pivotal ecological service of high-altitude alpine regions and is controlled by precipitation, evapotranspiration (ET), and soil water storage whereas the underlying ecohydrological processes remain highly unquantified. Here, we investigated continuous 19-year flux measurements to quantify the temporal patterns of ET and water budget (precipitation minus ET, P−ET), as well as 0-20 cm soil water storage change (ΔSWS). At a monthly scale, ET peaked in July (96.7 ± 26.4 mm, Mean ± S.D.) and averaged 41.7 ± 31.9 mm, whose variations were determined by the slope of the saturation vapor pressure curve at air temperature, air and soil temperatures, regardless of vegetation growth stage. P−ET averaged 18.3 ± 26.3 mm in August and September while stayed deficit during the other months. The variations in P−ET were controlled by precipitation in the May-October growing season whereas by ET in the non-growing season from November to April. ΔSWS peaked in May (28.8 ± 11.2 mm) and September (3.0 ± 2.7 mm) and almost accumulated to zero over the whole season. At annual scales, none of ET, P−ET, and ΔSWS changed significantly. ET averaged 512.2 ± 68.4 mm and exceeded precipitation (459.1 ± 58.4 mm), likely due to the lateral flow supply of uphill locations. The variations in ET were regulated directly by bulk canopy resistance and indirectly by net radiation. P−ET averaged −53.2 ± 95.4 mm and demonstrated a clear water deficit (−51.6 ± 21.0 mm) during the non-growing season. The variations of P−ET were driven jointly by precipitation and ET, with opposite but equivalent effects. The dominance of thermal conditions and energy availability on ET variability manifested an energy-limited feature of water loss in the alpine shrubland. The temporal patterns in P−ET elucidated that the alpine shrubland plays the water retention rather than water provision function through transforming variable precipitation input into stable ET loss. Pathways analysis showed that the composite thermal condition controlled the variations of monthly ET, including its indirect positive effect through atmospheric evaporative demand and negative effect through vapor transfer resistance. The observed thermal condition (Heat, including T a , T s , and △), atmospheric evaporative demand (Demand, including R n and VPD), vapor transfer resistance (Resistance, including r a and r s ), and moisture supply (Supply, including Rain and SWC) were classified into composite variables. Numbers neighboring the measured variables are their coefficients with composite variables. The thickness of the arrow is ten times the standardized coefficients. Red solid and black dashed arrows are positive and negative (significance levels are * P < 0.05, ** P < 0.01, and *** P < 0.001). R 2 represents the proportion of the variance explained by the model. Abbreviations: T a : air temperature; T s : topsoil temperature; △: the slope of the saturation vapor pressure curve with respect to T a ; R n : net radiation; VPD: satur a ted vapor pressure deficit; r a : aerodynamic resistance; r s : bulk canopy resistance; Rain: precipitation; SWC: 10 cm soil water content. • Variations of ET and P−ET were investigated by long-term eddy flux measurements. • Thermal conditions and bulk canopy resistance drove the variations of monthly and annual ET, respectively. • Monthly P−ET related to ET in the growing season and to P in the non-growing season. • Alpine shrubland plays a function of water retention rather than water provision to catchment.