Abstract:

Sap flow velocity (Js,cm/s) of Schima superba was measured using Granier’s thermal dissipation probe with meteorological factors monitored simultaneously by automatic weather station in Qianyanzhou ecological station. The results showed that significant difference existed in the diurnal fluctuations of sap flow under three typical weather conditions. Diurnal variations of sap flow displayed widepeaked curves on sunny days with starting time at 6:30 and peaking time at 12∶30. Diurnal variations of sap flow displayed singlepeaked curves on overcast days with starting time at 7∶30 and peaking time at 14∶30. While on rainy days, no distinct diurnal variations of sap flow occurred during the daytime. On the whole, sap flow velocity on sunny days were obviously higher than those on overcast days and rainy days due to obviously higher solar radiation strength,air temperature and lower relative humidity. Patterns of diurnal variations of sap flow varied among different growing seasons, mainly in the average level of sap flow velocity, the initiating time in the early morning and the declining time in the late afternoon. The average sap flow velocity followed the 〖JP2〗order of July (0001 663) > August (0001 56) > June (0001 472) > September (0001 314)〖JP〗 > May (0001 216) > April (0001 101) > October(0000 847). Sap flow velocity peaked in June, followed by July and August. The starting time of sap flow in June was 0.5-1 hours earlier than that in May, July, August and September. The initiation of sap flow was the latest in April and October. Sap flow peaked earliest in July, followed by May, June and August and latest in September and October. The initiation and peak of sap flow lagged behind average net radiation (ANR) but synchroised with canopy relative humidity (CRH) and canopy air temperature (CT). The variance analysis results showed that in the cases with and without the time lag effect, there existed significantly positive correlations between Js and ANR, while significantly negative correlations between sap flow velocity and average net radiation. There was no significant correlation between sap flow velocity and CT in both cases. The correlation strength between meteorological factors and Js were ANR (0677*〖KG-*2〗*) > CRH (-0417*〖KG-*2〗*) > CT (0088) and ANR (0752*〖KG-*2〗*) > CRH (-0325*〖KG-*2〗*) > CT (0158), respectively, indicating that when the time lag effect was considered, sap flow velocity was more dependent on average net radiation and canopy relative humidity. Multivariate linear model were used to quanitfy the integrated effects of three meteorological factors on sap flow velocity. The regression models and all parameters were significant at 1% level. The determination coefficients between comprehensive meteorological factors and sap flow velocity were 093 and 095, respectively