Determination of suitable planting date and stable genotype in cultivars and promising lines of rapeseed (Brassica napus L.) using GGE Biplot graphical analysis

The average yield of rapeseed in each country depends on climatic conditions, production methods and cultivar types. Proper planting dates provide adequate growth and development for crops and minimize the damage caused by stresses (Shirani-Rad et al., 2015). In the analysis of data collected from on-farm trials, environment can be defined as any management practices such as planting date, plant density, and fertilizer application that are recommended for crop producers (Balalić et al., 2012). Several methods have been proposed to analyze the genotype-by-environment interaction and to determine stable cultivars in different environments. However, each of these methods are based on certain statistical procedures and has their own advantages and disadvantages. In experiments, which are used to determine the interaction between genotype and environment, it is often difficult to establish the response patterns for the genotype-by-environment interaction without the aid of graphical representation of the data. The objective of this study was to use GGE Biplot graphical analysis and its efficiency to (1) determine the most suitable planting date for rapeseed 2) to determine the most stable genotype in the normal and delayed planting dates among the rapeseed cultivars and promising lines in the studied region.

The experiment was conducted over two cropping years (2014-2016) at the experimental farm of Hassan Abdali village, southwest of Zanjan, Iran (48°32'E, 36°37'N, altitude 1770 m). Each experimental plot consisted of four 6 m-long rows spaced 30 cm apart with 5 cm distance between plants on the rows. The experiment was arranged in split-plot based on a random complete block design with planting dates (D1: Oct. 7 as normal planting date, D2: Oct. 22 as semi-late planting date, D3: Nov. 6 as late planting date) in the main plots and genotypes G1 to G10 in the sub-plots with three replications. After eliminating the marginal effect, seed yield of each plot was estimated and finally seed yield/ha was determined. The outlier data detection and normality test of data were done before variance analysis using Grubbs’ test and Shapiro-Wilk test, respectively. Then the data were subjected to analysis of variance (ANOVA) using SAS software. GGE Biplot analysis was performed on the seed yield in order to determine the proper planting date and stable genotype among the cultivars and promising lines of rapeseed.

The results revealed that there were significant differences between genotype, planting date and their interaction (P<0.01). Seed yield of the investigated genotypes were declined from the highest yield of 4547 kg ha-1 in the first planting date (Oct. 7) to the lowest yield of 2118 kg ha-1 in the third planting date (Nov. 6). These results were similar with those of Singh et al. (2017) and Shirani-Rad et al. (2015). GGE Biplot analysis indicated that first two components explained 93 % of the total variation in seed yield in the three different planting dates. Based on which-won-where pattern, genotype G9 had the maximum seed yield (5437 kg ha-1) in the first planting date (Oct. 7) and genotype G7 had the maximum seed yield (2608 kg ha-1) in the second (Oct. 22) and the third (Nov. 6) planting dates. Based on vector view, the interaction of three different planting dates revealed that the first planting date (Oct. 7) with longer vector length and having more angle than the two subsequent planting dates had a greater effect on the genotypic distinction in terms of seed yield of the studied genotypes. On the other hand, the second and third planting dates had a smaller angle, indicating a positive and high correlation with each other.

Having an understanding of visual patterns is very important in analyzing agricultural data as compared to numerical results and numerous statistical tables. The GGE Biplot software, with a user-friendly graphical interface, can analyze various types of two-way data and provide a quick and complete understanding of the relationships between genotypes, environments, and their interactions. It can be concluded that GGE Biplot was a good method in identifying the suitable planting date and stable genotype in normal and late planting dates.