关键词: 华北落叶松, 树干液流, 坡位, 半干旱区

Abstract: [Objective] It is key to clearly recognize variation in sap flow velocity in trees in various slope positions and its main environmental factors, so that to improve the calculation accuracy for forest transpiration by means of up-scaling from a plot/position to a whole slope or a watershed, especially in semi-arid areas featuring with a strong variation of soil moisture in slopes. [Method] Three plots of Larix principis-rupprechtii were set up respectively at the upper, middle and lower positions on the northwest-facing slope in Diediegou watershed of Liupan Mountains, and the sap flow velocity in trees, as well as the soil water potential and meteorological factors were monitored simultaneously throughout the growing season (Jun. to Sep.) in 2015. [Result] The result showed that there were significant differences in the soil water potential of 60 cm layer (Ψ_0-60, MPa) among slope positions, in an order of lower position (-0.210)>middle position (-0.410)>upper position (-0.511), and in the daily average sap flow velocity (mL·cm^-2 min^-1) in an order of lower position (0.045)>upper position (0.036)>middle position (0.034). Throughout the study period, the daily average sap flow velocity was significantly and positively correlated with the average daily air temperature (T_a), daily maximum temperature (T_max), daily minimum temperature (T_min), potential evapotranspiration (PET), average daily saturated vapor pressure deficit (VPD), average daily solar radiation density (R_s) and Ψ_0-60, but significantly and negatively with the daily average air relative humidity (RH) and precipitation (P). Moreover, the correlation coefficients with each other were significantly different among slope positions. The order was lower>upper>middle for the meteorological factors, whereas upper>middle>lower for the Ψ_0-60. According to the boundary line analysis, significant differences were also found in the degree of sap flow velocity variation in response to each single factor among slope positions. From the upper position to the lower, the degree decreased gradually for T_a, T_max, RH, VPD, and Ψ_0-60, whereas increased for the R_s with stepwise regression analysis, a multivariate linear model was established for each slope position between daily average sap flow velocity and environmental factors. It was clear that the selected factors for each model obviously differed at different slope positions. For example, the first and the second important factors were R_s and Ψ_0-60 successively in the model of the lower position. However, they turned to Ψ_0-60 and VPD at the upper position, and VPD and Ψ_0-60 at the middle position, respectively. [Conclusion] As a conclusion, the variation of sap flow velocity among slope positions is a consequence of joint contributions of soil moisture and meteorological factors affected by terrain. Therefore, it is necessary to consider the effect of position differences in soil moisture and meteorological factors when scaling up the value of sap flow velocity from a plot/position scale to a slope scale.

Key words: Larix principis-rupprechtii, sap flow, slope position, semi-arid areas