Abstract: [Objective] The responses model of leaf stomatal conductance to environmental factors of Populus pruinosa in different periods constructed would be very helpful to elucidate stomatal regulation behavior of P. pruinosa, and to further simulate the dynamics of leaf photosynthesis and to develop a new water-carbon coupling cycle model in an extreme arid terrestrial ecosystem. P. pruinosa has been declining in recent years because of the increasingly worsening ecological environment, mainly caused by increased human water consumption. Up to now, the adjustment mechanisms of its stomatal conductance (G_s) are not clear. Our study is to elucidate current understanding of the mechanism that underlay the responses of stomatal conductance to variable environmental factors, and thereby to build up a model that expresses the relationship between stomatal conductance and environmental factors. This study would help us to further understand the photo-physiological characteristics of P. pruinosa and provide valuable information for protection of this vulnerable species. [Method] P. pruinosa, a constructive species of desert riparian forests in an extreme arid region in northwest China, was used as experimental material in this study. The leaf gas exchange parameters and environmental factors were measured with Li-6400 portable photosynthesis system during June to September in 2012 and 2013. The dynamic characteristics and the relationship between stomatal conductance and environmental factors were analyzed based on field observation data. Further, Jarvis and Leuning-Ball models were used to simulate the dynamic process of leaf stomatal conductance, and applicability of the two models in extreme arid region was compared. [Result] The diurnal courses of stomatal conductance of P. pruinosa were a single peak curve in growth season, there were obviously differences in peak values, time and amplitude in every months. Especially, peak time appeared the earliest and largest amplitude in September, and minimum amplitude of peak value in June. The leaf stomatal conductance was sensitive to photosynthesis active radiation, vapor pressure deficit and air temperature. The leaf stomatal conductance increased with photosynthesis active radiation, atmospheric CO₂ concentration and air humidity, and decreased with increase of the vapor pressure deficit and air temperature. Statistical analysis showed that photosynthesis active radiation, vapor pressure deficit and air temperature significantly affected stomatal conductance of whole day and forenoon, while the stomatal conductance in afternoon was affected obviously by atmospheric CO₂ concentration and air humidity. Stomatal conductance of P. pruinosa in different periods was regulated by the different environmental factors. The fitted models of stomatal conductance of P. pruinosa in different periods were simulated and constructed with two representative stomatal conductance models, Jarvis model could explain on average 69.1%, 62.2%, and 63.3% of variation and Leuning-Ball model could explain on average 53.5%, 30.6%, and 44.5% of variation in the observed stomatal conductance at whole day, forenoon and afternoon, respectively. The sensitivity and fitting effect of Jarvis model was better than that of Leunning-Ball model at different periods. The fitting effect of Leunning-Ball model in afternoon was better than that in forenoon, indicating that the environmental factors that affected stomatal conductance were different in different periods. The validations of Leuning-Ball linear and Jarvis non-linear models based on field data of leaf stomatal conductance indicated that Jarvis model was better estimation of stomatal conductance than Leuning-Ball model, and Jarvis model could improve the simulation effect of stomatal conductance. [Conclusion] The environmental factors of different periods affecting stomatal conductance of P. pruinosa were obviously different in growth seasons. The sensitivity and fitting effect of Jarvis non-linear stomatal conductance models were better than that of Leunning-Ball linear model at different periods, it had better applicability in the extremely arid-desert region. The relationship among leaf G_s and environmental factors in extremely arid Tarim basin could be expressed as:G_s=PAR(0.001T_air²+ 0.013 T_air -0.090)/((260.443+PAR)(-0.219+VPD)).

Key words: extreme arid region, Populus pruinosa, stomatal conductance, environmental factors, simulation