Crop models can be used to explain yield variations associated with management practices, environment and genotype. This study aimed to assess the effect of plant densities using CSM-CROPGRO-Soybean for low latitudes. The crop model was calibrated and evaluated using data from field experiments, including plant densities (10, 20, 30 and 40 plants per m2), maturity groups (MG 7.7 and 8.8) and sowing dates (calibration: 06 Jan., 19 Jan., 16 Feb. 2018; and evaluation: 19 Jan. 2019). The model simulated phenology with a bias lower than 2 days for calibration and 7 days for evaluation. Relative root mean square error for the maximum leaf area index varied from 12.2 to 31.3%; while that for grain yield varied between 3 and 32%. The calibrated model was used to simulate different management scenarios across six sites located in the low latitude, considering 33 growing seasons. Simulations showed a higher yield for 40 pl per m2, as expected, but with greater yield gain increments occurring at low plant density going from 10 to 20 pl per m2. In Santarém, Brazil, MG 8.8 sown on 21 Feb. had a median yield of 2658, 3197, 3442 and 3583 kg/ha, respectively, for 10, 20, 30 and 40 pl per m2, resulting in a relative increase of 20, 8 and 4% for each additional 10 pl per m2. Overall, the crop model had adequate performance, indicating a minimum recommended plant density of 20 pl per m2, while sowing dates and maturity groups showed different yield level and pattern across sites in function of the local climate.

中文翻译：

---

使用 CSM-CROPGRO-Soybean 评估低纬度大豆成熟度组的播种日期和植物密度

作物模型可用于解释与管理实践、环境和基因型相关的产量变化。本研究旨在评估使用 CSM-CROPGRO-Soybean 对低纬度地区植物密度的影响。使用来自田间试验的数据校准和评估作物模型，包括植物密度（每平方米 10、20、30 和 40 株植物）2)、成熟组（MG 7.7 和 8.8）和播种日期（校准：2018 年 1 月 6 日、1 月 19 日、2 月 16 日；评估：2019 年 1 月 19 日）。该模型模拟物候的偏差低于 2 天的校准和 7 天的评估。最大叶面积指数的相对均方根误差从 12.2% 到 31.3% 不等；而粮食产量则在 3% 到 32% 之间变化。校准后的模型用于模拟位于低纬度的六个地点的不同管理方案，考虑到 33 个生长季节。模拟显示每平方米 40 pl 的产量更高2，正如预期的那样，但在低植物密度从 10 到 20 pl/m 时出现更大的增产增量2. 在巴西圣塔伦，2 月 21 日播种的 MG 8.8 的中位产量分别为 2658、3197、3442 和 3583 公斤/公顷，每平方米 10、20、30 和 40 pl2, 导致每增加 10 pl 每平方米相对增加 20%、8% 和 4%2. 总体而言，作物模型具有足够的性能，表明推荐的最低植物密度为 20 pl/m2，而播期和成熟组在不同地点之间表现出不同的产量水平和模式，这取决于当地的气候。