水稻品种产量试验的边际效应分析

农作物品种产量试验中边际效应是普遍存在的。本研究按照江苏省水稻品种区域试验的田间设计要求,选用8个品种对小区边际效应进行分析。结果表明:①每穗总粒数、每穗实粒数和小区行产量均有多个品种表现显著或极显著的边际效应,穗数只有扬4928、千粒重只有武育粳3号存在边际效应,但品种的边际效应主要是在边1行。②剔除小区边1行后,所有品种的产量均低于不剔除边1行的产量,即边际效应最终影响品种的产量,但边际效应没有影响品种的产量位次和显著性分析结果。③大穗和偏大穗型品种的每穗总粒数、每穗实粒数易表现出边际效应,而小区行产量的边际效应与不同穗数类型的相关性并不明确。

Study on Border Effect in Rice (Oryza Sativa L.) Yield Trials

Interference between plots is known to occur in cultivar trials for many cultivated species.This paper is based on an experimental study of border effect in rice yield trials using eight cultivars,according to the experiment designs in rice regional trials of Jiangsu province.The results are as follows: 1.For many rice cultivars,the border effect is significant or very significant on grains per spike,filled grains per spike and grain yield respectively,only Yang4928 occurs border effect on panicles,and Wuyujing 3 occurs border effect on 1 000-grains weight,but the border effect only occurs in the first row of plots,and there is no significance among the other rows of the plots.2.The grain yield with unharvested the first row of plots is lower than that of harvested the first row of the plots.The border effect distorts yield estimates in cultivar trials,but the bias can be neglected in rice yield trials.3.The cultivars with big panicle occurs border effect easily on grains per spike and filled grains per spike,and there is no relation between border effect of grain yield per row and different type rice cultivars in panicle