Abstract Alfalfa (Medicago sativa L.) is the predominant forage legume species in Canada and is considered a prioritized option for sustainable cropping under climate change. Crop growth models provide an opportunity to explore the potential impacts of climate change on alfalfa and for evaluating potential adaptation options. For this study, six experimental datasets in eastern Canada were used to parameterize the newly adapted CSM-CROPGRO-Perennial Forage Model (CSM-CROGRO-PFM) in simulating alfalfa regrowth and to identify areas for further model improvement needed for climate change assessments in the northern agricultural regions of North America. Estimated air temperatures under snow cover were used successfully to drive the CSM-CROPGRO-PFM model for simulating alfalfa regrowth in eastern Canada. The simulated values of aboveground biomass across all sites and years were acceptable with a root mean square error (RMSE) of 936 kg dry matter (DM) ha−1 and a normalized RMSE of 24%. A sensitivity analysis of the model revealed that with no change in the number of harvests per year, the simulated annual herbage yield (harvestable biomass) declined with increasing temperature, increased with elevated atmospheric CO2 concentration, and changed little with increased precipitation. However, the increase in the number of harvests made possible by warmer temperatures may increase the simulated annual herbage yield. Although most alfalfa physiological processes were successfully simulated, some additional model functions may be required to further improve the simulation of alfalfa regrowth for climate change studies conducted in Canada. These functions include quantifying plant density decline and its relationship with biomass in post-seeding years, estimating temperatures surrounding alfalfa crowns during the overwintering period, and simulating herbage nutritive attributes.

中文翻译：

---

使用 CSM-CROPGRO-多年生草料模型模拟加拿大东部的苜蓿再生和生物量

摘要 紫花苜蓿 (Medicago sativa L.) 是加拿大的主要牧草豆科植物，被认为是气候变化下可持续种植的优先选择。作物生长模型为探索气候变化对苜蓿的潜在影响和评估潜在的适应选择提供了机会。在这项研究中，加拿大东部的六个实验数据集被用于参数化新调整的 CSM-CROPGRO-多年生草料模型 (CSM-CROGRO-PFM) 以模拟苜蓿再生，并确定需要进一步改进模型的领域，以便在加拿大进行气候变化评估。北美北部农业区。雪盖下的估计气温被成功用于驱动 CSM-CROPGRO-PFM 模型，以模拟加拿大东部的苜蓿再生。所有地点和年份的地上生物量模拟值均可接受，均方根误差 (RMSE) 为 936 kg 干物质 (DM) ha-1，归一化 RMSE 为 24%。该模型的敏感性分析表明，在每年收获的数量没有变化的情况下，模拟的年牧草产量（可收获生物量）随着温度的升高而下降，随着大气 CO2 浓度的升高而增加，随着降水的增加而变化不大。然而，由于气温升高而增加的收获数量可能会增加模拟的年度牧草产量。尽管大多数苜蓿生理过程都被成功模拟，但可能需要一些额外的模型函数来进一步改进对加拿大进行的气候变化研究的苜蓿再生的模拟。