Environmental Response And Modeling Of Stomatal Conductance Of Apple Trees In The Jinzhong Basin

Stomatal conductance(gs)is an important parameter to measure the rate of material exchange between plants and the external environment through stomata,affecting the photosynthetic physiological activities of leaves.The analysis of the response of gs to environmental factors has positive significance for explaining the process of crop nutrient utilization and scientific optimization of the environment.Taking apple trees in Jinzhong basin as an example,the gray relational degree method was used to study the relationship between gs and photosynthetically active radiation(PAR),air temperature(Ta),leaves temperature(Tl),leaves surface relative humidity(RHl),air relative humidity(hs),CO2 concentration(Ca),saturated water pressure deficit(VPD)and other environmentalfactors on the basis of analyzing the diurnal variation of gs,combining with specific trend changes to analyze the specific response relationship between gs and each environmental factor,and the results were tested for significance.The Jarvis,Ball-Woodrow-Berry(BWB),BallBerry-Leuning(BBL)and the optimal stomatal conductance(OS model)were used for gs numerical simulation,and we also modified them by introducing the difference of CO2 concentration(ΔC)and leaf temperature(ΔT),respectively.The main conclusions were as follows:(1)The diurnal variation trends of gs during the main growing seasons of apple trees in different years were basically the same,showing a "bimodal" curve,with only differences in the time node and change amplitude of the peak values.According to the grey correlation analysis,PAR is the most core environmental factor affecting the change of gs,and the correlation degree is also the highest.In addition,Ta and Tl were secondary factors to PAR,and environmental factors such as VPD,hs,RHl and Ca had low correlation with gs.In 2019,the values of correlation degree between the main influencing factors and gs were PAR(0.860),Ta(0.699),Tl(0.657),and in 2021,the values were PAR(0.726),Tl(0.697),Ta(0.691),respectively.The response relationship of gs to environmental factors in different years showed that gs increased first,and then decreased with the increase of PAR,temperature factors(Ta and Tl),humidity factors(RHl and hs)and Ca.Then with the increase of VPD,gs decreased.The significance test results between gs and various environmental factors showed that PAR,Ta and Tl were significantly correlated with gs in different years,and the correlation coefficients were PAR(0.546),Ta(0.520)and Tl(0.518)in 2019,and PAR(0.676),Tl(0.653)Ta and(0.571)in 2021,respectively.However,the correlation between RHl,hs,Ca,VPD and gs was poor,and VPD was negatively correlated with gs.(2)In 2019 and 2021,the Jarvis,BWB,BBL and OS models were used to simulate the gs of apple tree leaves.Results showed that the fitting accuracy of Jarvis model was better than other three models.In 2019,the accuracy index coefficients of Jarvis model were El(0.637),dl(0.740)and MAE(0.047),respectively.In 2021,they were E1(0.436),d1(0.609)and MAE(0.024).The fitting accuracy of the improved BBL model(E1=0.321,d1=0.682,MAE=0.125;In 2021,E1=0.415,d1=0.599,MAE=0.029)was better than BWB model(E1=0.319,d1=0.568,MAE=0.146 in 2019;In 2021,E1=0.319,d1=0.529,MAE=0.031),indicating that the BBL model was more suitable for gs simulation of apple tree leaves than the BWB model after improvement.The fitting accuracy of OS model was relatively poor,which may be related to the environmental conditions of low soil moisture in the experimental area.According to the parameter sensitivity test and analysis of Jarvis model,gs has the highest sensitivity coefficient values for PAR(0.192,0.183)in different years.(3)The CO2 concentration difference ΔC and leaves temperature difference ΔT were introduced to modify the four gs models.Results showed that the modified Jarvis-C,BWBC,BBL-C,OS-C,Jarvis-T,BWB-T,BBL-T and OS-T models have better fitting effect than the original expressions.In terms of overall fitting accuracy,the Jarvis-T model had significantly improved the fitting accuracy compared with the original expression in the numerical simulation of gs on different observation days and different years in one year,so it could be used to simulate the diurnal and inter-annual variations of gs of apple tree leaves in Jinzhong basin.