Modeling The Functional And Structure Coupling Model G-ORY For Rice And Its Applications

Rice(Oryza sativa L.)is a major grain crop in China.Reasonable plant type and breeding is of great importance to improve rice yield.Building functional and structure model of rice can combine rice growth and development with its structure characteristics,which provide digital tools for rice breeding and field management.In this dissertation,three rice varieties(TQ,YIL55 and PAY1)with contrasting aerial architecture(e.g.,growth habit,plant height,number of tillers)were chosen and planted at two markedly different latitudes_Beijing(c.40°N)and Sanya(c.180N).We measured canopy architectures and geometric features of the three varieties in situ using a 3-D digitizer,then simulated and evaluated the growth and development process of the three varieties at two sites by ORYZA2000.Combining field experiment data,meteorological data with above work,a rice functional-structure coupling model,G-ORY,was built at GroIMP platform.Distributed computation and its visualization have been realized in the embedded module,the results of which greatly improved ORYZA2000 light module.Then,we evaluated the light distribution of the three rice varieties in the two latitudes using G-ORY model,the objective of which was to provide scientific basis for rice plant type selection.The main conclusions are as follows:(1)The static 3-D structure model of rice was reconstructed using 3-D dimensional digital technology.The azimuth angle and inclination angle of leaf and stem of three varieties were identified and classified in three key growth stages and two latitude regions by statistical senses.First,the results showed that the azimuth angle distribution at base of blade(AAB)and stem(AAS)of TQ are uniform at jointing stage(JS),and dispersed to east-south at heading stage(HS),then migrated to 240°～60° at filling stage(FS).The AAB and AAS of PAY 1 exhibited a similar pattern with TQ after jointing stage,but before that were more decentralized.AAB and AAS of YIL55 were more uniform in the whole growth periods than TQ and PAY.Second,the angle between stem and leaf(ASL)and stem inclination angle(SIA)of YIL55 were larger,and evenly distributed in 50°,30° and 20° at JS,HS and FS respectively.SIA of TQ and PAY1 distributed in 0°-30° and 0°-35° at JS respectively,and distributed in 0°～10°after HS.Third,compared with Beijing,SIA of Sanya is more evenly distributed.(2)The dynamic structure model of rice was built based on the statistical data of angle distribution(e.g.,AAB,AAS,ASL and SIA)at GroIMP platform.Photosynthetically active radiation(PAR)could be dynamically simulated under different solar positions.The dynamic changes of morphological structure of rice organs(e.g.,length,width and diameter)with development stages could be also computed.In addition,the 3-D rice can be dynamically visualized in OpenGL module at GroIMP.(3)We calibrated and evaluated ORYZA2000 for the three rice varieties,using data from two field experiments carried out in Beijing and Sanya in growing seasons from 2012-2014.A trial-and-error approach was applied to calibrate extinction coefficient for leaves(KDF).First,the results showed that,the three varieties did not differ significantly in terms of specific leaf area(SLA)and dry matter partitioning factors of panicle(FSO)between Beijing and Sanya.But,development rate(DVR),dry matter partitioning factors of leaves(FLV)and stems(FST)over the same phenological stages were all found differed.Second,the KDF for the two erect genotypes(TQ and PAY1)was the same as the default value of the model.However,the KDF for the semi-prostrate genotype(YIL55)for the period from emergence to flowering was lower than that for the other two varieties.Third,accuracy evaluations of simulated and measured organ dry matter and leaf area index(LAI)indicated that the predictive performance of ORYZA2000 was sufficiently accurate for simulating the biomass of the genotypes in this study(d>0.95,NRMSE range between 6-25%),but simulated LAI was generally underestimated.(4)A rice function-structure-coupled model,G-ORY,was built by integrating the static 3-D structure model,ORYZA2000 model,field experiment data and meteorological data at GroIMP platform.The results indicate that,in comparison with ORYZA2000,G-ORY has more ample information when simulating light distribution of rice canopy.For example,it can express the distribution difference between similar plant type(i.e.,TQ and PAY1)and reflect the heterogeneity of canopy light distribution of YIL55 at different canopy heights(i.e.,the PAR intercepted by the top canopy is larger than bottom).The comparison of the light distribution among varieties show that the PAR of each phytomer of PAY1 was uniform in early development stage,and the PAR of TQ and YIL55 were concentrated on canopy top.In late development stages,PAR for three rice varieties was all concentrated on of canopy top,and intercepted by the flag leaf of PA1 is 175.54 kJ>TQ(89.71 kJ)>YIL55(44.35kJ).The bottom of canopy of TQ and PAY1 could be absorbed more PAR than YIL55 because of the erect growth habit.(5)Quantitative evaluating the simulation results of G-ORY in PAR,CO2 assimilation of the canopy(GPCDT),organ biomass(BIO),LAI and organ morphology structure are shown as follows.First,compared with the simulation result of ORYZA2000,the simulated values of PAR,GPCDT,BIO and LAI of different rice varieties in two latitudes are at the same level and have a good agreement with each other(the average NRMSE is 19±7%).It demonstrated that the G-ORY model can accurately calculate canopy light distribution and photosynthesis production,and it can reflect the dynamic change process of above ground organ biomass and LAI over the course of development.The morphology of organs simulated by G-ORY is at same level with measured values(i.e.,the NRMSE is 23%and the average relative error was less than 20%for each relative phytomer number).(6)We discussed and analyzed the effect of plant type and latitude factors on the canopy light distribution and the photosynthesis using G-ORY.The result showed that the plant type of PAY 1 is the best(i.e.,the maximum daily PAR is 10.92 kJ m-2 d-1),which indicated it can absorb more light energy under same light environment,then TQ(the maximum daily PAR is 10.00 kJ m-2 d-1)and YIL55(the maximum daily PAR is 7.18 kJ m-2 d-1).Compared with growing in Sanya,the PAR absorbed by canopy of three varieties growing in Beijing have more photosynthesis because of higher temperature,more sunshine hours and lower solar elevation.Especially for TQ and PAY1,the maximum PAR increased 35.82%and 10.44%respectively in Beijing compared with Sanya.On the other hand,YIL55 is suitable for planting in high latitude areas(i.e.,Sanya),because it’s prostrate growth habit could intercept more solar energy.