In the hilly region of Chinese Loess Plateau,the rest riction of the nature conditions and the effect of human activities lead to ecological degradation and ecological problems, such as soil and water loss. To improve the ecological fragility of the Loess Plateau, our government took a lot of measures of ecological restoration, and afforestation is one of the most important measures for ecological restoration. Although these measures had made certain achievements in regional ecological
environment, there also existed problems. Due to not consider the plant adaptation to the environment, the restoration in the Loess Plateau caused soil dry layer and low growth rate for plants. So to study the soil water consumption of restoration vegetation and the response of vegetations to soil moisture is helpful to understand the sustainable development of restoration ecosystem. So the ai m of this study was to
invest igated the adaptation of physiolo gical process of restoration vegetations to environmental factors, analyze the its water consumption in water balance model, and understand the ecological hydrolo gical process es of restoration plantation. Shrub species are one of the most important restoration species on the Loess Plateau. However, the study of water consumption and ecological hydrological effect of typical shrub species were not insufficient, lack ing the quantitatively estimates of
response to soil moisture and mechanism explanation.
We conducted our research in Yangjuangou catchment, Yanan, with three typical shrub plantation, which is Vitex negundo L., Hippophae rhamnoides L., and Spirea pubesce ns Turcz. The experiments were conducted in the main growing seasons during 2015 and 2016 (from June to September). Sap flow density ( was measured using heat balance method. The s tand transpiration was cal culated basing on the stand
sap flow density an d leaf area. In addition, environmental variables, including solar radiation (Rs), air temperature (Ta), precipitation (P), relative humidity (RH) and soil water content (SWC) above 200 cm. We also calculated the vapor pressure deficit (VPD), and potential evapotranspiration (ET 0 ) using micro meteorological variables.
During the study, the eco ph ysiological indicators of plants were also measured for three shrub plants, incl uding tem poral dynamics of leaf transpiration, stomatal conductance and leaf water p otential. The main results are as follows
（1 ) The influencing factors of Fd were different at different time scales. On hourly and daily time scale s , the main influencing factors for three shrub species were meteoro lo gical factors . Fd was significantly p ositi vely correlated with VPD, Rs and Ta P < 0.05), while Fd was significantly negative correlated with RH P < 0.05). On daily time s cale , the response of Fd to VPD and Rs showed exponential regression.The SWC in different soil layer had different effec t on Tr. The Tr in three shrub plantations were correlated with deep soil layer (120 200 cm) ( P < 0.05). Yet there is
no significant correlations between Tr and SWC in H. rhamnoides and S.pubescens P > except V. negundo plantation, which is correlated with SWC in topsoil layer (0 40 cm) P < 0.05). Monthly Tr in three plantations were correlated with leaf area index ( V. negundo ) and soil moisture H. rhamnoides and S.pubescens
( 2 ) Precipitation was the main source of s oil moisture. The responses of plant physiological processes to different precipitation amount were different. In this study, the response of sap flow density (Fd) to rainfall events depended on species and the rainfall patterns. The increase in daily Fd w as large with small rainfall events (0.2--2 mm) mainly ascribed to enhanced meteorological conditions on days after rain , although the soil moisture did not change significantly. When rainfall events were
2-- 10 mm, the response percentage of Fd reduced. This behavior mainly because rainfall would increase the air's relative humidity and decrease Ta. When rainfall amount exceeded 10 mm, SWC was the main factor for the response percentage of Fd. In addition, the Fd related hydraulic conductance increased in res ponse to rainfall event for H. rhamnoides , while leaf water potential were F or V. negundo and S.pubescens , the water potential at the soil root interface and leaf water potential
increased , while hydraulic conductance was stable .
(3) In addition to daytime transpiration, plants could adjust the nighttime sap flow density to adapt the drought conditions. Although the low fraction of nighttime Fd to daily Fd of three shrubs, the importance of nighttime sap flow for plant drought adaptation was signi ficant. T he ratios of nighttime Fd to daily Fd of three shrubs were increased with decreasing SWC. Nighttime Fd was importance in increasing predawn
leaf and xylem water potential, opening predawn stomatal conductance earlier . In addition, because of stem refilling during the night, transpiration in the morning used the refilling stem water.
（4 ) The components of water balance (Tr, canopy interception, runoff, soil evaporation) were all influenced by precipitation. During the main growing seasons in 2015 i.e., a relatively dry year, representing 41.47% of the long term meanprecipitation (P) and 2016 (i.e., a relatively wet year, with almost the same level of P as the long term mean rainfall) the total Tr values were 107.21 and 180.60 mm, respectively, fo r V. negundo ; 122.33 and 180.28 mm, respectively, for H. rhamnoides ;84.03 and 137.02 mm, respectively, for S.pubescens . Tr of three plantation w ere the main component of evapotranspiration in 2015 and in 2016 for V. negundo and H.rhamnoides , followed by soil evaporation and canopy interception . Soil evaporation was the main component in S. pubescens plantation in 2015. The observed surface runoff values were very low in both years, suggesting that the restored shrubs effectively reduced soil erosion. Duri ng the study period, water loss exceeded P during most months of the main growing seasons, and the discrepancy between water loss and P was greater in 2015. Thus, the change in net soil water storage was
negative, especially in the HR stand.
In summary, meteorological variables and SWC were the main factors for the transpiration, and then influencing the ecological hydrological processes. Plant could adjust their response of small rainfall events and nighttime sap flow density to enhance their adaptability to environment. However, mu l ti year observations should be conducted to further investi gate the long term hydrol ogical effects of the plantations on water cycling at stand, watershed and regional scales.