摘要：
科学合理地确定适宜的试验小区规格,对于经济、高效地获取具有较高精度和代表性的试验数据具有重要现实意义.该研究以确定滴灌条件下适宜的试验小区规格为目标,采用幂函数建立土壤含水率方差与不同小区面积之间的相关关系,计算土壤水分异质性指数以表征试验小区内土壤含水率的相关程度;利用土壤含水率异质性指数,采用Hatheway法确定小麦滴灌试验小区适宜的面积和重复数;根据不同方向上土壤含水率异质性指数研究形状对试验结果的影响,确定试验小区适宜的长宽比.结果表明:增大试验小区面积可以提高试验数据的准确性,但不同面积区间数据离散程度的降低幅度变化较大:面积由1 m2增大到50~100 m2,灌水定额30 mm处理田块间土壤含水率方差降低77.4%~82.6%,灌水定额45 mm处理降低78.6%~83.7%,面积从100 m2增加到500 m2,方差降低幅度显著变小,灌水定额30 mm处理降低17.4%,灌水定额45 mm处理降低16.3%;增加试验重复数可以增加试验对处理间土壤含水率差异的区分能力,试验小区面积在50~100 m2时,重复数由2增加到3能检测出的处理间土壤含水率差异由23.2%~26.5%提高到13.1%~15.0%,表明增加试验重复数可以检测出试验处理间更小的土壤含水率差异,提高试验精度.通过研究滴灌试验小区土壤水分异质性指数与滴灌试验小区规格之间的关系,综合考虑试验精度、代表性和田间实际操作得出以下结论:若试验为高、低水2处理时,试验布设宜为3重复,每重复小区面积为50 m2,可在80%的概率下检测出15%的真实差异;若试验为高、中、低3处理时,试验布设宜为3重复,每重复小区面积为100 m2,可在80%的概率下检测出12%的真实差异.滴灌试验小区适宜的形状为沿滴灌带方向布设的长方形,长宽比在1:1到5:1之间可使试验小区所得数据更具代表性.%The objective of this research was to improve the accuracy and representativeness of the experimental plot date by determining the optimum plot area and the replication number for the wheat drip irrigation. The heterogeneity of soil moisture in experimental plots was provided by Smith heterogeneity index. The relationship between the plot area and the repetition number of the wheat drip experimental plot was measured by Hatheway method. The effect of shape on the accuracy of experimental data was determined by studying the soil moisture heterogeneity index in the horizontal direction of drip tape and the vertical direction of drip tape in experimental plot. The results showed,when the plot area was increased, the variance of soil water in the field was reduced and the relationship between them was in power function . For the plot form 1 m2 to 50-100 m2, the variance decreased significantly. The full irrigation treatment variance was decreased by 0.79-0.84, and the deficit irrigation treatment variance was reduced by 0.79-0.84. For the plot more than 100m2, the increase of plot area had a limited effect on the variance reduction. To determine the optimum size of the experimental plot you need take the factors as below into consideration, as the number of treatments increased, the real difference between treatments decreased, and the required size for the experimental increased. When the size of the plot increased to a certain value, improving the identification ability of test difference by increasing the area of experimental plot was limited and adding the repetition number to improve the distinguishing ability of the difference between treatments was necessary. For 50-100 m2 of experimental plots, when the repetition was increased, the experimental accuracy could be improved with the same area to distinguish the smaller real difference between treatments. By increasing the number of repeats from 2 to 3, the differences of soil moisture between the treatments could be detected which increased from 23.2%-26.5% to 13.1%-15.0%. The recognition ability of the real difference among the treatments could be significantly improved. Hatheway method was used to calculate the optimal experimental plot of the full and deficit irrigation treatment. Due to its larger coefficient variation of soil moisture and heterogeneity index, the deficit irrigation treatment required larger testing area. In order to facilitate the layout of the field experiment, the unified standard of the plot was set up with the optimum specification of deficit irrigation treatment. When the experiment was divided into full and deficit irrigation treatment, the testing plot should be 3 replicates, and the size of each repeat plot should be 50 m2.When the irrigation experiment was divided into three treatments, the experiment layout should be 3 replicates, and the size of each replicate plot was 100 m2. For the experimental plot shape, according to the relationship between the variance of field block and the size of experimental plot, the Power function was fitted by whether the length is greater than the width to calculate the soil moisture heterogeneity index in horizontal the direction of drip tape and the vertical direction of drip tape. About the difference between two directions, the heterogeneity index of deficit irrigation treatment was greater than that of full irrigation treatment. The heterogeneity index in the horizontal direction of drip tape was greater than that in the vertical direction of drop tape. Therefore, the result of rectangular area arranged along the direction of the drip tape was more accurate (smaller deviation) than other shapes. The rectangular experimental plot with aspect ratio between 1:1 and 5:1 could balance the variability of both directions and made the experimental plot more representative. This paper provided a scientific basis for determining the specifications of drip irrigation experimental plot and the theoretical support for the standardization of irrigation experimental layout.
