Effects Of Groundwater Depth On Physiological Indexes Of Populus Euphratica And Its Response Of Spectral Characteristic

Populus euphratica forms a natural green corridor along the rivers of the Tarim Basin in Xinjiang,maintaining the stability of the desert riparian ecosystem in the desert-oasis transition zone.In this study,a monitoring research plot was set up in the National Public Welfare Forest Reserve in the lower reaches of the Yarkand River.The spectral reflectance and physiological index density of Populus euphratica leaves under different groundwater depths were obtained,and the spectral characteristics and physiological index density of Populus euphratica leaves under different groundwater depths were analyzed.At the same time,used the PROSPECT-SAIL model to simulate the Populus euphratica canopy spectrum and screen the sensitive spectral indices.Three machine learning algorithms were used to establish the Populus euphratica canopy physiological index estimation model,and completed the mapping of Populus euphratica physiological index density based on hyperspectral remote sensing.The spatial change characteristics of physiological indicators of the Populus euphratica were clarified in the Tarim River,and carried out the health evaluation analysis of the Populus euphratica to provide scientific basis for the precise management of the natural Populus euphratica forest and the conservation of the desert riparian ecosystem.The setting of groundwater depth range in this study corresponds to drought stress.The deeper the groundwater level,the more severe the drought stress.The main research results are as follows:(1)The chlorophyll density,carotenoid density and the ratio of chlorophyll to carotenoid of Populus euphratica in the groundwater depth of 8-10 m were significantly lower than those under normal conditions,while the ratio of chlorophyll a to chlorophyll b was significantly higher,indicating that the photosynthetic intensity and stress tolerance of Populus euphratica were decreasing under drought stress.SOD,POD,MDA,SSC and Pro increased with the increase of groundwater depth,so as to resist the damage of excessive reactive oxygen radicals,regulate the intracellular osmotic potential or produce secondary metabolites,so as to improve the drought resistance of Populus euphratica.ABA,Z and ZR showed a continuous increase trend with the increase of groundwater depth.ABA,as a plant hormone,could inhibit the growth of Populus euphratica under drought stress.ABA could also prevent the decrease of the activities of POD and SOD,and effectively regulate the metabolic balance of reactive oxygen species.ABA had complex relations with many biochemical components,and would continue to increase under drought stress.High Z and ZR concentration could promote the development of canopy leaves and enhance the photosynthetic adaptability of canopy leaves.On the one hand,the hormones needed to be balanced under drought stress,and on the other hand,the morphology and physiology of P Populus euphratica could be adapted to drought stress to improve drought resistance.Due to drought stress,the specific leaf weight of leaves increased,which requires a lot of elements to change the leaf structure.The increase of Fe²⁺,Mn²⁺and Cu²⁺indicated that Populus euphratica tended to maintain normal nutrient metabolism and enhance its resistance.With the increase of groundwater depth,the density of Zn²⁺decreased significantly,which indicated that the absorption of other ions,such as Fe²⁺,Mn²⁺and Cu²⁺,inhibited the absorption of Zn²⁺by Populus euphratica.On the other hand,the decrease of absorption of Zn²⁺indicated that some physiological and metabolic functions of Populus euphratica would be affected if the stress was further aggravated.The absorption of K⁺and Ca²⁺increased while the absorption of Na⁺decreased,indicating that the cells tended to maintain a high K⁺/Na⁺,thus reducing intracellular osmotic pressure.(2)The red valley position and near-infrared reflectance of Populus euphratica leaves with groundwater depth of 6-8m and 8-10m were significantly different from those with groundwater depth gradient.The spectral curves of Populus euphratica leaves with different groundwater depths could be distinguished by using the first and second derivative spectra resampled at 5nm and 10nm and the spectral similarity parameters SID and SAM.The continuum removal spectrum showed two distinct absorption valleys.In the absorption valley near 500nm,?₅₀₀(reflectance)and W₅₀₀(absorption width)showed an upward trend with the increase of groundwater depth,while?₅₀₀(absorption position),Tan₅₀₀(slope),S₅₀₀(symmetry)and SAI₅₀₀(absorption index)showed a downward trend.In the absorption valley near 675nm,?₆₇₅ and SAI₆₇₅ showed an upward trend with the increase of groundwater depth,while H₆₇₅(absorption depth),Tan₆₇₅,AL₆₇₅(left area),AR₆₇₅(right area),A₆₇₅(total area)and W₆₇₅ showed a downward trend.Tan₅₀₀ and W₆₇₅ had the highest correlation with chlorophyll density.With the increase of groundwater depth,the absorption intensity of blue edge,yellow edge,red edge,green peak and red valley decreased,and the amplitude of yellow edge and red edge decreased.The correlation between red valley parameters and most physiological indexes was very significant.Absorption characteristic parameters Tan500,W675 and Red Valley parameters were sensitive factors to the changes of leaf physiological characteristics of Populus euphratica.(3)The canopy spectra of Populus euphratica were simulated by PROSPECT-SAIL model.NDVI,RVI and DVI spectral indices were constructed by the canopy spectra.According to the construction principles of spectral indices,the optimal spectral indices for inversion of the biochemical components of Populus euphratica canopy were screened.Spectral index was used as a modeling factor to construct SVM,RF and ANN models for estimating canopy physiological indexes of Populus euphratica.Fourteen index models of SVM estimation model passed accuracy verification,including Chl a,Chl b,Chl a+b,Car,Ant,N,SLA,Ca,ABA,2ipr,Z,ZR and 2ip.Therefore,it is proved that the predictive ability of SVM is the best among the three machine learning algorithms,and it also provides a basis for the selection of model methods in image-based estimation modeling.(4)Based on AHSI hyperspectral images,SVM model was used to estimate and model the physiological indexes of Populus euphratica,the results showed that the validation accuracy of the estimation model for 17 indexes,including Chl a,Chl b,Chl a+b,Car,Ant,Chl/Car,N,SLW,SLA,Fe,Mn,Zn,Ca,ABA,GA₃,Z and ZR met the modeling requirements.Comprehensive of these indicators,the regional Populus euphratica forest health assessment and mapping were completed.