依据紫花苜蓿生物学特性，通过田间试验和广泛收集资料，构建了紫花苜蓿光合生产与干物质积累模拟模型。该模型包括光合作用、呼吸作用、叶面积消长、干物质积累、同化物分配和产量形成等过程，考虑了温度对光合作用的影响。计算得到紫花苜蓿不同生育时期的干物质转换系数( β )和同化物分配分配系数[C(d)I]，确定了主要紫花苜蓿品种的光合参数(a 和Pmax)。分别利用北京和太原不同年份和不同品种的试验资料对地上部生物量、产量和叶面积指数模型进行了检验。结果表明，模型对叶面积动态、地上部生物量和产量模拟效果较好，叶面积指数、地上部生物量、茎和叶生物量的决定系数分别为0.98、0.95、0.96和0.88（n=20），产量均方差(RMSE)为103 kg hm-2，相对均方差(NRMSE)为2.1% (n=102)。模型不仅具有较强的机理性，而且有较好的拟合性。
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. 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.
全 文 ：