论文部分内容阅读
调查了福建省山区40个试点1168丘杂交水稻的产量结构,按照亩产由亩总粒数、结实率和千粒重构成,亩总粒数由亩穗数和每穗粒数构成的模式,应用PC-1500型微电脑,对结果因素与构成因素的数量变化关系进行多元分析,结果明确: (1)各构成因素对结果因素的作用力,主要决定于各自变异度的相对大小。通过调节变异度大的构成因素,可以大幅度改变结果因素的取值;而一个变异度不大的构成因素,即使其偏回归系数可能较大,也不可能大幅度改变结果因素的取值。因此,在生产实践中,加强产量结构的调查,着力于增加变异度大的构成因素的数量,是提高杂交水稻产量的秘诀。 (2)构成亩产的三个因素中,以亩总粒数(X3)的变异度及对产量的通径最大,40个试点平均,CV3=16,1％,P3=0.9835,P3:P4:P5=61:23:16,三类杂交组合平均,CV3=19.1％,P3=1.0255,P3:P4:P5=62:20:18;仅在少数试点,结实率(X4)的变异度及对产量的通径,接近于亩总粒数。因此,提高山区中低产田产量的关键,在于增加亩总粒数,扩大产量库,在个别条件下,还需同时注意提高结实率;千粒重(X5)的变异度及对产量的通径最小,40个试点平均,CV6=4.4％,P5=0.2582,反映出千粒重是组合(品种)的稳定性状,不能指望通过提高粒重来达到增产的目的。剩余因素的通径系数Pe6很小,产量通径分析的机误在3％以内,是相当可靠的。 (3)构成亩总粒数的两个因素——亩穗数(X1)和每穗粒数(X2),变异度及对亩总粒数的通径都较大。有的试点以亩穗数,有的试点以每穗粒数的变异度及对亩总粒数的通径较大,但在多数试点,二者的变异度及对亩总粒数的通径相近。40个试点平均,CV1=14.1％,CV2=12.8％,P1=0.8658,P2=0.8031,P1:P2=52:48。数据表明,随着产量水平的提高,亩总粒数的增加,亩穗数和每穗粒数多是双双增长的,二者对于增加亩总粒数具有近于同等的重要性。剩余因素的通径系数Pe3很小,亩总粒数通径分析的机误在5％以内,是相当可靠的。 (4)通过微电脑计算,建立了三类杂交组合各产量区间的亩产和亩总粒数的多元线性回归方程,复相关系数达到极显著标准,且多在0.95以上,为建立合理的产量结构模式和预测预控产量,提供了依据。
The yield structure of 1168 mound hybrid rice in 40 pilots in 40 mountainous areas of Fujian Province was investigated. According to the model that the total output of grain per mu, seed setting rate and grain weight per mu, -1500 micro-computer, the results of factors and changes in the composition of the relationship between the number of multivariate analysis, the results are clear: (1) the composition of factors on the outcome of the force, mainly depends on the relative size of the respective degree of variation. By adjusting the constituent factors of large variation, the value of the result factor can be greatly changed. However, a component with a small variation degree can not change the value of the result factor significantly even though its partial regression coefficient may be large. Therefore, in production practice, strengthening the investigation of production structure and focusing on increasing the number of constituent elements with large variation are the tips to increase the yield of hybrid rice. (2) Among the three factors that make up the mu production, the variation degree of the total grain number (X3) and the path to yield were the largest. The average of 40 pilot sites was CV3 = 16,1%, P3 = 0.9835, P3: P4 P5 = 61:23:16, the average of the three types of hybrid combinations was CV3 = 19.1%, P3 = 1.0255, P3: P4: P5 = 62:20:18. The path of production, close to the total grain number of acres. Therefore, the key to increase the yield of middle and low-yielding mountainous areas is to increase the total grain number of acres and expand the yield base. Under certain conditions, it is also necessary to pay attention to improving the seed setting rate. The variation of 1000-grain weight (X5) The average of 40 pilots, CV6 = 4.4%, P5 = 0.2582, reflects that 1000-grain weight is the stability trait of the combination (variety) and can not be expected to increase production by increasing the grain weight. The remaining factor of the path coefficient Pe6 is very small, the output path analysis of the machine error within 3%, is quite reliable. (3) The two factors that constitute the total grain number of MU - the number of ears per acre (X1) and the number of grains per head (X2), the variation degree and the path of the total number of grains per acre. Some experiments to the number of acres per panicle, some of the pilot to the variation of grains per spike and the total number of acres of large diameter path, but in most of the pilot, both the degree of variation and the total grain size of the path similar. The average of 40 pilots was CV1 = 14.1%, CV2 = 12.8%, P1 = 0.8658, P2 = 0.8031, P1: P2 = 52:48. The data show that with the increase of production level, the increase of the total grain number, the increase of the number of ears per panicle and the number of grains per spike, both increase nearly equally. The remaining factor of the path coefficient Pe3 is very small, total grain number of path analysis machine error less than 5%, is quite reliable. (4) Through the microcomputer calculation, the multiple linear regression equations of the total output of mu per mu and the total number of mu of the three types of hybrid combinations were established. The multiple correlation coefficients reached the extremely significant standard, and were more than 0.95. To establish a reasonable yield structure Model and predict the yield of pre-control, provided the basis.