Simulation Model On Development, Growth And Quality For Greenshouse Standard Cut Chrysanthemum

Chrysanthemum is one of the most famous flowers in China and it is also theimportant cut flower around the world. Standard cut chrysanthemum is planted ingreenhouse to fit market requirements for year round production. The model ofdevelopment, growth and quality of cut chrysanthemum can be used as tools for predictionand planning of production and greenhouse climatic control. In this study, aiming at mainproblems of untidy anthesis and low quality of chrysanthemum, experiments with differentcultivars, planting dates and greenhouse types were carried out in greenhouse in Shanghaiduring 2005 and 2006. Based on the effects of temperature and light on the development,growth and quality of chrysanthemum, a simulation model was developed to predict harvestdate, organ dry weight and main quality indices of standard cut chrysanthemum. The mainresearch contents are as follows:1. The development simulation of standard cut chrysanthemum in greenhouseBased on the effects of temperature, photosynthetically active radiation (PAR) and daylength on chrysanthemum development, the concept of physiological product of thermaleffectiveness and PAR (PTEP) was proposed and used to predict cut chrysanthemumdevelopment stages and harvest date. PTEP was defined as the needed accumulated productof thermal effectiveness and PAR (TEP) under the optimum temperature and light conditionfor a certain development stage of chrysanthemum. PTEP was calculated as follows:BD₁=TEPs/TEPsiBD₂=TEPh/TEPhiWhere PTEP (MJ·m^-2) is physiological product thermal effectiveness and PAR; TEP (MJ·m^-2) is product of accumulated thermal effectiveness and PAR; RPE is relativephotoperiod effectiveness. BD₁ and BD₂ are, respectively, the basic development factorsfrom planting to beginning of short-day treatment and from beginning of short-daytreatment to harvest date. TEPs and TEPsi are, respectively, the TEP needed for thestandard variety (basic development factor is 1) and variety i from the planting to short-day treatment; TEPh and TEPsi are, respectively, the TEP needed for the standard variety andvariety i during the short-day treatment to the harvest date; SD is model parameter. Themodel was validated with independent experimental data. The results showed that thesimulated values agreed well with the observed ones in different development stages of thestandard cut chrysanthemum. Based on the 1:1 line, the root mean squared error (RMSE),from cutting to planting, short-day treatment, bud visible and harvest stage, were 2.2, 2.9,1.2 and 3.1 days respectively. Compared with the model based on the Growth Degree Days(GDD), the prediction accuracy of the model developed is significantly higher than that ofthe GDD-based model (RMSE is 3.0, 12.5, 12.5 and 15.5 days, respectively).2. The simulation model of leaf area of greenhouse cut chrysanthemum Firstly, the relationship between the leaf number (N) and the PTEP after planting wasexpressed with the equation: N=64.11/(1+exp (-(PTEP-80.53)/35.38))Then, the relationship between the leaf length increasing rate (VLLi), the maximum ofleaf length (LL_MAXi) and the leaf order (i) were expressed as follows, respectively:The actual leaf length (LLi) was calculated as follows:â–³PTEP_ij=PTEP_j-PTEP_iWhere PTEPj is the PTEP accumulated from planting to the day j. PTEPi is the neededPTEP from planting to the time when leaf i unfolding. The leaf area per plant after theapperence of leaf i can be calculated as LA=sum from i=1 to N LA_m,i－LAoLA is the leaf area per plant; i is the leaf order. N is leaf number. LAmi is the leaf area ofvariety m with leaf i. LAo is the leaf area of old leaf removed before harvest. Thecoefficient of determination (R²) and the root mean squared error (RMSE) between the predicted and the measured leaf area index (LAI) based on the 1:1 line are 0.94 and 0.75,respectively.3. The simulation of biomass production and dry matter partitioningA model for predicting biomass production and dry matter partitioning were developed.The relationship between the biomass prouduction with the PTEP after planting wasexpressed with an expolinear equation: DWT= (C_m/r_m)×ln (1+exp (r_m×(PTEP-PTEPb)))Where DWT (g.m^-2) is plant total dry weight per square meter ground. C_m(g.m^-2.(MJ.m^-2)^-1) is the maximum growth rate in the linear phase; rm ( g.g^-1(MJ.m^-20-1)is the maximum relative growth rate in the exponetial phase. PTEPb (MJ.m^-2) is the PTEPneeded for canopy development before all radiation is intercepted. PTEP was thephysiological product thermal effectiveness and PAR after planting. The dry matterpartitioning model can be expressed as follows: PIS=0.7669 + 0.0016×PTEP - 5×10^-6×PTEP² PIR=1-PIS PIL=0.371 + 0.2579×exp (-0.5×((PTEP-41.22)/46.89)²) PIST=1-PIL-PIFWhere PIS, PIR, PIL, PIF and PIST are the dry matter partitioning index of shoot, root, leaf,flower and stem respectively. Independent experimental data were used to validate themodel. The results showed that the simulated values agreed well with the observed ones.Based on the 1:1 line for biomass production, dry weight of leaf, stem and flower, and freshweight of shoot per plant, the coefficient of determination (R2) and root mean squared error(RMSE) between the simulated and the measured values were 0.97, 0.97, 0.97, 0.83 and0.94, and 67.56g.m^-2, 22.4g.m^-2, 28.2g.m^-2, 15.4g.m^-2, 10.37g-plant^-1, respectively. For thebiomass production, the prediction accuracy of the PTEP-based model is significantlyhigher than that of the photosynthesis process driven crop growth model (R² and RMSEwere 0.82 and 274.68g.m^-2, respectively).4. The simulation model of quality of standard cut chrysanthemum in greenhouseBased on the response of quality of standard cut chrysanthemum to temperature andight, the model for predicting the leaf numbe(N), plant height(H), stem diameter(Ds), internode length (Ln)and the flower diameter (Df) was developed: N=64.11 / (1+exp (-(PTEP-80.53)/35.38)) H=105.58/(1+exp(-(PTEP-74.15 )/28.26)) Ds=8.24/(1+exp(-(PTEP-42.83)/51.03) Ln=0.75+0.019×PTEP-8.5×10^-5×PTEP²According to the validated results of the model with independent experiment data, thecoefficient of determination (R²) based on the 1:1 line and relative prediction error (RSE)for leaf number, plant height, stem diameter, intemode length and flower diameter were:0.99, 0.98, 0.92, 0.87 and 0.88, and 5.5ï¼…, 5.9ï¼…, 4.1ï¼…, 11.2ï¼…, 12.4ï¼…, respectively.5. Simulation the effects of number of stems on quality of standard cut chrysanthemumExperiments with different cultivars, stems, planting density and planting dates werecarried out during 2005 and 2006 to quantitatively investigat the effects of number of stemson quality of standard cut chrysanthemum. A dynamic model for predicting leaf area indexof multi-stem standard cut chrysanthemum was developed using the experimental data. Theleaf area model was expressed with the equations as follows: LAI=LAI_max/(1+exp(-(t-tb)/LAI_max)) LAI_max=4.57+0.04×NWhere LAI_max is the maximum LAI of plant with different stems; t_b is the apparent time lostduring canopy development before all radiation is intercepted; N is the number of stems perground area. The relationship between leaf area index (LAI (i, L)) and the intercepted PARof canopy (PAR (i)) was described as follows: PAR(i,L)=PAR(i)×(1-exp(-k×LAI(i,L)))PAR(i) is the total PAR above the canopy; k is the light extinction coefficient, Therelationship between the shoot dry weight per ground area (DWS) and the canopyintercepted PTEP (PTEPi) can be described as follows: DWS=81.16+11.96×PTEPiThe relationship between the quality index (Y) and the PTEPi is described by as follows: Y=Y_max×(1-exp(-R×PTEPi/Y_max))Where Y_max and R is the maximum and increasing rate of quality indices of cut chrysanthemum respectively; a0 is model parameter. The model was validated withindependent experimental data, and the predicted results agreed well with the observed ones.The coefficient of determination (R²) based on the 1:1 line and relative prediction error(RSE) for shoot fresh weight per stem, plant height, stem diameter, leaf number, and flowerdiameter were: 0.95, 0.96, 0.94, 0.91 and 0.81, and 16.1ï¼…, 10.1ï¼…, 12.8ï¼…, 13.4ï¼…, 15.9ï¼…,respectively.In this study, the model was developed for predicting the different development stages,dry matter production, dry weight of each organ and main quality indices of standard cutchrysanthemum by inputting the data of average air temperature, PAR in the greeenhouseand the cultivar parameter. From the results, it can be concluded that the model can givesatisfactory prediction of harvest date, the biomass production and quality of cutchrysanthemum, and be used in optimizing climate management for greenhouse standardcut chrysanthemum production.