种子是农业生产的基本资料，在玉米种子标准发芽率连续多年达标的基础上，对冷浸萌发率和种子整齐度进行比较，更能反映种子的田间出苗情况，对提高机械化精量播种效果具有参考意义。本文以8家种子公司2020年在甘肃省张掖地区生产加工的京科968玉米种子为材料，检测标准发芽率和冷浸发芽率，结果表明，不同公司生产的京科968玉米种子标准发芽率均在95%以上，达到单粒播标准（93%），批次间无显著差异，但冷浸发芽率在不同公司之间的差异达到显著水平，有5个批次的种子冷浸后发芽率显著降低。进一步分析检测到多个与冷浸发芽率显著相关的快检指标，其中R_同质性和B_同质性与冷浸发芽率正相关，相关系数达0.7以上，B_相异度、裂纹率、龟裂纹率与冷浸发芽率显著负相关，相关系数分别为-0.728、-0.807、-0.717。仅以裂纹率为自变量建立的种子冷浸发芽率回归方程达显著水平，决定系数为0.652。对粒型、尺寸、单粒体积、单粒比重和单粒重进行比较，发现不同公司生产的京科968玉米种子在上述指标上均存在显著差异，K-均值聚类表明有3个公司生产的京科968整齐度较高。

Seeds are the basic material of agricultural production. Considering that the standard germination percentage of maize seeds have been reaching the national standard for several years, cold soaking germination percentage and seed uniformity are the more practical indicators for field performance, and have reference value for improving the precision sowing effect. In this paper, Jingke 968 maize seeds produced and processed by 8 seed companies in Zhangye area of Gansu Province in 2020 were used as materials to test the standard germination percentage and cold soaking germination percentage. The results showed that the standard germination percentages of Jingke 968 maize seeds produced by different companies were more than 95%, more than the single kernel sowing standard (93%), and without significant difference. But there was significant difference in cold soaking germination percentage among different seed lots. The germination percentage of 5 seed lots decreased significantly after cold soaking. Further analysis showed that two texture features, namely R_homogeneity and B_homogeneity, were significantly positively correlated with cold soaking germination percentage, and the correlation coefficients were both more than 0.7. There was a significant negative correlation between B_dissimilarity, crack percentage, multiple crack percentage and cold soaking germination percentage, and the correlation coefficients were -0.728, -0.807 and -0.717, respectively. Only with the crack percentage as the independent variable, the regression equation of the cold soaking germination percentage reached a significant level. The decision coefficient of the regression equation was 0.652. Based on the comparison of seed shape, size, single seed volume, single seed specific gravity and single grain weight, it was found that Jingke 968 maize seeds from different companies had significant differences in the above features. K-means clustering showed that the seeds of 3 companies had a higher uniformity.

科技部十三五重点研发项目(2018YFD0100903)