QTL Mapping Of Leaf Area Index And Chlorophyll Content In Multi-spectral Wheat Based On Unmanned Aerial Vehicle Remote Sensing

Wheat（Triticum aestivum L）is one of grain food crops with the most widely in the world.In China,it is also an important food ration,which plays a vital role in food security.In recent years,with the global warming and frequent extreme weather,drought stress has seriously restrains the sustainable production of wheat yield.A important indexes of wheat drought resistance and yield is leaf area index（LAI）and chlorophyll content（CC）.A set of 309 lines of spring wheat‘Worrakatta’בBerkut’recombinant inbred line（RIL,at F₆）as materials,the phenotype s of leaf area index（LAI）at blooming,filling and mature stages,flag leaf chlorophyll content（CC）at heading,blooming and filling stages,were measured and Unmanned aerial vehicle image acquisition under drought stress and normal irrigation,respectively.The inversion models of LAI and CC were established by decision tree C4.5 algorithm,and two sets of values about measured values and predicted values of LAI and CC were obtained.The values were combined with wheat 50K SNP chip for QTL mapping and exploring candidate genes,respectively.in order to provide reference for rapid and nondestructive phenotypic data acquisition in the field and genetic mechanism research of LAI and CC.The main results are as follows:1.The results of phenotypic data analysis showed that LAI and CC were significantly decreased under drought stress compared with normal irrigation.The maximum values were reached at the filling stage and the blooming stage under normal irrigation and drought stress,respectively,indicating that drought stress could promote the increase of LAI in a short time.CC reached its maximum value at blooming stage under both water treatments.LAI and CC of both parents and RIL line showed great differences at each growth stage in the two treatments,and there was obvious superparent separation.RIL line B057 LAI,and at the same time on the expression of the CC value were higher,while B092 on the expression of LAI and CC value are lower,shows that they were stable in response to drought stress and could be used as the parents of drought-resistant varieties2.QTL mapping results of measured data are showed that two and one QTL related to LAI were detected at blooming stage and mature stage,respectively,and were distributed on chromosomes 5BS,2BL and 1BL under normal irrigation.QLAI.xjau-5BS,QLAI.xjau-2BL.1 and QLAI.xjau-1BLcan explain 6.8%～8.2%of.phenotypic variation.CC-related QTL QCC.xjau-1DS was detected at both heading and blooming stages,located on the 1DS chromosome,accounted for 5.3%～5.8%of the phenotypic variation.One QTL for LAI was detected at mature stage under drought stress,QLAI.xjau-2BL.2 located on the 2BL chromosome,accounts for 13.8%of the phenotypic variation.Seven candidate genes related to LAI and CC were screened from the QTL locis found in the study,including two F-box family proteins,one MYB related gene,one GATA related gene,one coding abscisic acid receptor gene,one BTB/POZ related gene and one WUS related gene.These genes are involved in regulating crop growth and development and signal transduction,as well as in response to drought and other stress responses.This study provides reference information for gene discovery and molecular breeding of leaf area index and chlorophyll content in wheat.3.Use of decision tree C4.5 algorithm of LAI and CC modeling,validating and forecasting,and the accuracy of the model was evaluated based on the size of determination coefficient R2.The results showed that the R2 of LAI and CC from 0.639 to 0.883,and the predicted values of LAI and CC showed consistent dynamic changes with the measured values under the two water treatments in each growth period,indicating that the modeling and prediction of LAI and CC by this algorithm was feasible.Four LAI related locis and two CC related locis were found through QTL mapping of the predicted values.There were 3 new locis,including QLAI.xjau-2BL-pre.1 and QLAI.xjau-2BL.1 at the filling stage,which accounted for 3.6%and 3.3%of the phenotypic variation,and QCC.xjau-3AL-pre at the heading stage,accounted for 3.6%of the phenotypic variation.Three overlaps were found between the predicted and measured values,which were QLAI.xjau-2BL-pre.1 and QLAI.xjau-2BL.1 at blooming stage,respectively,accounted for 3.1%and 6.8%of the phenotypic variation.QLAI.xjau-2BL-pre.2 and QLAI.xjau-2BL.2 at mature stage accounted for 3.3%and 13.8%of phenotypic variation.CC-associated loci QCC.xjau-1DS-pre and QCC.xjau-1DS at heading stage accounted for 2.5%and5.8%of the phenotypic variation.The number of overlapping locis accounted for 60%of the measured QTL number,indicating that it was reliable to obtain LAI and CC phenotype data from UAV multi-spectrum.