Cowpea data for three environments in the main rainy season in northern Ghana for 2016 cropping season.

The research hypothesis was that, F2-derived F8 Recombinant inbred lines could perform differently and even yield better than the parental checks under main growing season conditions in the three agro ecologies of northern Ghana.
Specifically, the study was designed to identify promising inbred lines with high yield potential, and then identify inbred lines with stable mean yield performance either with specific adaptation to a particular environment or across the three environments that will subsequently be recommended to farmers in the Guinea, Sudan and Transitional agro ecologies of Ghana
The experimental design used at each test location was a randomized complete block design with four replications. The seeds were planted according to the conditions for each location but were thinned to one plant per hill. Plot size was 3 m long; each plot contained 5 rows of 10 plants per row; with plant spacing of 60 cm between rows and 20 cm within rows with the number of entries being 24. Thus, each experimental unit consisted of 50 plants per plot, and each block contained 24 plots giving the total plots as 96 plots for the whole experiment for each location.
Data were recorded on plot bases for all three locations. Days from planting to first flowering for each plot was recorded, the date to 50% flowering data was recorded when half of the plants per plot produced flowers. Based on this information, the days to 50% flowering were estimated. At harvest, number of pods per plant, number of seeds per pod and hundred seed weight were taken as average of five randomly selected plants within a plot excluding the border plants. The weight of hundred seeds (g) for each treatment was determined by the use of an electronic balance. Data on grain yield was recorded on plot bases using three middle rows of 10 plants (30 plants per plot) in grams extrapolated to t/ha
The data for each location were subjected to Analysis of Variance (ANOVA) using GenStat statistical package 12th edition. Combined analysis of variance across locations for grain yield and yield components were also carried out to determine the interactive effects of genotypes by environment. The additive main effect and multiplicative interaction (AMMI) and the genotype, and genotype by environment interaction were concurrently determined using the Breeding management software (BMS) and GenStat.
Analysis of variance for each location as well as combined analysis of variance across all the locations revealed that there were significant differences among genotypes and their environments for grain yield and yield components indicating the presence of variability in the inbred lines as well as diversity of growing conditions and locations.

本研究提出的假设为：在加纳北部三个农业生态区的主生长季条件下，F2衍生的F8重组自交系（Recombinant inbred lines）的表现可能存在差异，甚至相较于对照亲本具备更高的产量。
具体而言，本研究旨在筛选具备高产潜力的优良自交系，进而鉴定出具有稳定平均产量表现的自交系——这类自交系要么可适应特定环境，要么可在三种试验环境中保持稳定表现，后续将向加纳几内亚农业生态区、苏丹农业生态区及过渡带农业生态区的农户推广应用。
各试验点均采用随机完全区组设计（Randomized Complete Block Design），设置4次重复。试验种子依据各试点的种植条件进行播种，随后间苗至每穴保留1株幼苗。小区长度为3米，每个小区包含5行，每行10株植株；行间距为60厘米，株距为20厘米，试验共设24个参试材料。由此可知，每个试验单元的每个小区包含50株植株，每个区组包含24个小区，单试点整个试验的总小区数为96个。
所有试点均以小区为单位收集试验数据：记录每个小区从播种至始花期的天数；当小区内半数植株开花时，记录50%开花期的日期，并据此估算至50%开花的天数。收获期，选取小区内除边界植株外的5株随机植株，取其平均值作为单株荚数、每荚种子数及百粒重的测定结果。使用电子天平称量各处理的百粒重（单位：克）。籽粒产量以小区为单位测定，选取小区内中间3行（每行10株，共30株植株）的产量，以克为单位记录，并折算为吨/公顷（t/ha）。
采用GenStat统计软件12th版对各试点的数据进行方差分析（Analysis of Variance, ANOVA）。同时针对籽粒产量及产量构成因素开展跨试点联合方差分析，以解析基因型与环境的互作效应。借助育种管理软件（Breeding management software, BMS）与GenStat软件，同步计算加性主效应与乘性互作（AMMI）模型的相关参数，并估算基因型及基因型×环境互作效应。
各试点的方差分析及跨所有试点的联合方差分析结果均显示，基因型及其所处环境在籽粒产量及产量构成因素上存在显著差异，表明供试自交系存在遗传变异，且各试点的种植条件与环境亦存在多样性。