Adaptability and phenotypic stability of soybean genotypes regarding epicotyl length using artificial neural network and non-parametric test

Genetic improvement together with statistics has contributed to the growth of importance soybean in Brazil. One contributions been launching new cultivars national market, which requires, its legal procedures for registration and protection, verification several tests, one them being distinguishability test. Several studies have reported that some phenotypic characters are potential this distinction, is length epicotyl. In work, objective was identify genotypes present low or high average, highly stable throughout analyzed environments adaptability different environments. Two groups experiments were conducted a greenhouse measure epicotyl plants submitted (planting season). The data obtained using analysis individual variance, joint Scott-Knott test stability through Artificial Neural Network non-parametric It can be concluded showed average length, wide poor responsiveness environmental improvements over seasons TMG 1175 RR (in V2), BMX Tornado BG 4272 BRS283 V2 V3) FT-Cristalina V3). BRSMG 752 S V3), 4185 BRSGO 7560 behaved as medium, adaptability. BRS 8381, 4185, MG/BR46_Conquista, 850 GRR, Valiosa 4277 recommended favorable