Water Consumption And Evapotranspiration Model Of Rain-fed Jujube Plantations In The Loess Plateau

In semiarid hilly areas of Chinese Loess plateau,limited water resources are the main constraints to the agricultural development under dry climate.In the past several decades,unreasonable afforestation leaded to sever soil desiccation,and the dried soil layers have been noted in many plantations.As a drought tolerant plant,jujube(Ziziphus jujuba Mill.)has long growing history and is still commonly planted in the region today.Thus well understanding the law of jujube water consumption will contribute to improve soil and water management and maintain the sustainability of jujube plantation.Our study focus on researching the water use strategies of jujube orchard under rainfed condition.Field measurements including trunk sap flow,tree growth indexs,soil moisture,evaporation and meteorological data were performed by using thermal dissipation method(TDP),Winscanopy analysis equipment,time domain reflectometry(TDR)probe sensors,neutron moisture gauge and weather station over three jujube growth season from 2012 to2014.Firstly,the response of sap flow to meteorological factors under different soil moisture conditions was explored to reveal the regulation ways of jujube to drought.Secondly,the performance of three evapotranspiration models—Shuttleworth-Wallace(SW)model,the Penman-Monteith(PM)model and the Penman-Monteith equation with a variable surface canopy resistance(PMv)model were compared using Bayesian analysis and suitable models were selected at daily and hourly time scales based on model accuracy and uncertainty.These works lay the foundation for irrigation schedule in some parts of the region.For other parts of the region without irrigation condition,the effectiveness of branch removal pruning on water conservation were evaluated.The main findings are as follows:(1)For jujube trees,relative sap flow(SF_R)was mainly driven by solar radiation(_sR)and vapor pressure deficit(VPD).The sensitivity of sap flow to meteorological factors significantly decreased with increasing water stress conditions.Daily relative sap flow was linearly correlated with solar radiation but logarithmically correlated with vapor pressure deficit,variable transpiration and reference evapotranspiration under the two soil moisture conditions.Hourly analysis showed that relative sap flow linearly increased with increasing trends in meteorological factors under sufficient soil water conditions(?_e>0.3).However,peak values only occurred after initial linear increase with increasing meteorological factors under soil water stress conditions(?_e<0.3).(2)The variation characteristics of relative sap flow in the two time periods(that are,change in relative sap flow with meteorological factors for the period 13:00–15:00 hrs and15:00–23:00 hrs)were largely different under both soil moisture conditions.The difference in the variation characteristics of relative sap flow between the two soil moisture conditions mainly occurred under the first time period(13:00–15:00 hrs).The results of the study were critical for accurate refinement of transpiration models.For instance,the study suggested that it was important to separately simulate relative sap flow processes in relation to meteorological factors for the two time periods.It was also important to take into account two soil moisture conditions investigated in this study.(3)Compared with prior distribution,the calibration using the Bayesian approach generally reduced uncertainties associated with the three ET model parameters.While some parameters varied with the seasons or the soil water conditions,the significant differences were noted among the model posterior distributions at the hourly and daily time scales.Thus for long-term simulation of ET,it was critical to take into account the effects of changes in season,time scale,and/or soil water condition on model performance.(4)The three models not only met the evaluation criteria at the daily time scale,but also had acceptable uncertainty bounds for the 2012 and 2013 jujube growth seasons in the Loess Plateau region of China.Furthermore,differences in accuracy and uncertainty between the PM and SW models were minimal after using the Bayesian approach.Where only climatic data were available,the PM_V model was recommended for estimating daily ET in jujube plantations within the study area due to its simplicity and accuracy.Where sufficient field data were available,the less complex PM model was preferred over the SW model for estimating daily ET in jujube plantations in the semiarid Loess Plateau region of China as both models performed equally well.(5)The three models were less accurate at the hourly than at the daily time scale for the calibration(2013)and validation(2012)periods using Bayesian analysis.Model errors mainly occurred at growth stage 1 for the SW and PM models and at the?<9%soil water conditions for the PM_V model.Irrespectively,the three models met the evaluation criteria for the 2013 calibration period while only the SW model met the criteria for the 2012validation period.Based on the accuracy and uncertainty bounds,the SW model was preferred for estimating ET in jujube plantations within the Loess Plateau region of China at the hourly time scale.(6)In rainfed jujube plantations in Chinese semiarid Loess Plateau Region,branch removal pruning was a good treatment in terms of improving soil water storage as well as maintaining a reasonable productivity.Moreover,it obviously improved the tolerance of jujube trees to negative water potential and xylem cavitation.Compared with other pruning methods,branch removal is labor-saving,once it was applied in the first year,the improvements it brings to the orchard can last for many years.Since the soil evaporation accounted for a large proportion in evapotranspiration of jujube orchard,it is important to try combining branch removal pruning with mulching,that would bring a synergetic effect on water conservation.And among all treatments,removal of two(out of three)branches was recommended regarding the obvious benefits it makes for soil and water ecology.(7)Branch removal pruning significantly reduced transpiration per unit ground area,making it more sensitive to environment conditions,while increased transpiration per unit leaf area,making it less sensitive to soil moisture condition,but not influence the isohydric stomatal control of jujube trees.Branch removal pruning also obviously increased the values of canopy conductance at VPD=1 kPa,enables isohydric trees to maximize carbon assimilation under low VPD and soil drought conditions.Moreover,it reduced the vessel diameter of the trunk,thus improving the tolerance of jujube trees to negative water potential and xylem cavitation.