A Dynamic Model Simulating Photosynthetic Production and Dry Matter Accumulation for Alfalfa (Medicago sativa L.)

ZHU Yu-Jie,FENG Li-Ping,YI Peng,YANG Xiao-Guang,HU Yue-Gao

DOI: https://doi.org/10.3321/j.issn:0496-3490.2007.10.018

2007-01-01

ACTA AGRONOMICA SINICA

Abstract:Photosynthetic production and dry matter accumulation are major determinants of the final yield in crop production.Early simulation model of photosynthetic production and dry matter accumulation for alfalfa is relatively simple and much depending on experience.On the base of the eco-physiological processes and biological characteristics of alfalfa(Medicago sativa L.),a simulation model for photosynthetic production and dry matter accumulation was established by the field experiment data and widely collected data.This model included the sub-models of photosynthesis,respiration,leaf area dynamic,dry matter production,assimilate partitioning,and yield.The model took into account of the effect of temperature.The model was parameterized with data from the literature and local experiments.Data from Beijing were used for calculation of the coefficients of dry matter allocation [C(d)i] at different development stages.Data from literature were used for calculation of the coefficients of dry matter conversion(β).The photosynthetic parameters(a and Pmax) for six alfalfa cultivars were determined based on the data from the field experiments at Beijing and Taiyuan of Shanxi province.The verification and validation of the model were conducted.Data from Beijing were used for validation of leaf area dynamics and above-ground biomass.Data from Beijing and Taiyuan were used for validation of crop yields of 6 cultivars.The model simulated appropriately the growth of the crop.The adjusted linear correlation coefficient(R2) values between simulated and measured leaf area,above-ground biomass,stem biomass,and leaf biomass were 0.98,0.95,0.96,and 0.88(n=20),respectively.The root mean squared error(RMSE) between simulated and measured yields was 103 kg ha-1,and the normal root mean squared error(NRMSE) was 2.1%(n = 102).The results indicate that the model has not only good mechanism in model building,but also good simulation capability.