| According to the characteristic of growth and development of Cymbidium responding to temperature and light, Cymbidium Hiroshima Golden Cup'Sunny Moon'and Cymbidium Great Flower 'Marie Laurencin'were used as experimental material to study the growth and development simulation of Cymbidium in the greenhouse by the method of TEP. The mathematical simulation model of Cymbidium development, growth, dry material production and partitioning based on production of thermal effectiveness and PAR was initially established through different varieties and seeding age tests. The main results are as follows:(1) The development simulation model of Cymbidium was established. The results indicated that the accumulated production of thermal effectiveness and PAR and the number of actual days required to complete vegetative growth stage, flower bud differentiation stage, flower bud development stage, flowering stage of two-year-old Cymbidium were respectively 834959047.56,38932786.95, 351740790.33,32157558.21μmol.m-2 and 206,46,139,19d for'Sunny Moon',842531090.35, 53299351.14,359509584.64,44167788.32μmol.m-2 and 224,43,142,19d for'Marie Laurencin'. The accumulated production of thermal effectiveness and PAR needed of'Marie Laurencin'were more than 'Sunny Moon'to complete the same development stages. By testing, the prediction accuracy of the model to the observed and simulated days required to complete vegetative growth stage, flower bud differentiation stage, flower bud development stage and flowering stage of two-year-old Cymbidium were respectively 85.4%,89.1%,97.9%,85.0% for'Sunny Moon' and 85.9%,90.2%,98.1%,90.5%for 'Marie Laurencin'.(2) The growth prediction model of Cymbidium was established. The results showed that the quadratic or cubic curve equations well fitted the relationship between Cymbidium plant height, pseudobulb diameter, number of unfolding leaf and accumulated production of thermal effectiveness and PAR, the optimal regression equation all reached a significant level by test. The decision coefficient (R2) of predicted values and actual measured values of the Cue ball, Son ball, Sun ball plant height, pseudobulb diameter, the number of unfolding leaf of the two varieties were respectively higher than 0.98,0.98,0.95; the regression estimation standard error (RMSE) were respectively lower than 2.06cm, 0.85mm,1.21; the relative prediction error (RSE) were lower than 4.8%,2.8%,8.8%. The model simulation of the growth index had higher precision.(3) The simulation model for dry matter production of Cymbidium was established. The relationship between the total dry weight per plant and the accumulated production of thermal effectiveness and PAR could be well fitted by exponential, quadratic or cubic curve equations, the best fitting equation achieved a significant level by test. The decision coefficient (R2) of simulated values and observed values of total dry weight per plant of the seedlings, one-year-old seedlings and two-year-old seedlings of the two varieties were higher than 0.98, the regression estimation standard error (RMSE) and the relative prediction error (RSE) were lower than 12.99 g per plant and 10.1% respectively.(4) The dry matter partitioning simulation model of Cymbidium was established. The changes of aboveground part, leaf and flower dry matter partitioning index with the accumulated production of thermal effectiveness and PAR could respectively be fitted by quadratic or cubic curve equations, quadratic or exponential curve equations, logarithmic or quadratic curve equations and the regression equations all reached to significant level by test. The decision coefficient (R2) of simulated and observed values of root, pseudobulb, leaf and flower dry weight were higher than 0.98,0.97,0.93,0.98, the regression estimation standard error (RMSE) and the relative prediction error (RSE) were lower than 3.55,3.09,7.90,6.66 g per plant and 8.6%,10.7%,11.8%,23.4% respectively.
|
PDF Full Text Request