Stability analysis of maize cultivars adapted to tropical environments using AMMI analysis

Abstract

Genotype × environment interactions complicate selection of superior genotypes for narrow and wide adaptation. Eighteen tropically-adapted maize cultivars were evaluated at six locations in Nigeria for 2 yrs to (i) identify superior and stable cultivars across environments and (ii) assess relationships among test environments. Environment and genotype × environment interactions (GEI) were significant (P < 0·05) for grain yield. Environments accounted for 63.5% of the total variation in the sum of squares for grain yield, whereas the genotype accounted for 3.5% and GEI for 32.8%. Grain yield of the cultivars ranged from 2292 kg ha -1 for DTSTR-W SYN2 to 2892 kg ha -1 for TZL COMP4 C3 DT C2 with an average of 2555 kg ha -1 . Cultivar DT SYN2-Y had the least additive main effect and multiplicative interaction (AMMI) stability value of 7.4 and hence the most stable but low-yielding across environments. AMMI biplot explained 90.5% and classified cultivars and environments into four groups each. IWD C3 SYN F3 was identified as the high-yielding and stable cultivar across environments. ZA15, ZA14, BK14, BK15 and IL15 had environment mean above the grand mean, while BG14, BG15, LE14, LE15, IL14, LA14 and LA15 had mean below the grand mean. ZA, BK, BG, LE and LA were found to be consistent in ranking the maize cultivars. However, Zaria, Birnin Kudu, and Ilorin were identified as the best test locations and could be used for selecting the superior maize cultivars. The identified high-yielding and stable cultivar could be further tested and promoted for adoption to contribute to food insecurity in Nigeria.

Evaluation of maize grain yield and yield stability by AMMI analysis

“…These results highlight the necessity of multi-year and multi-location testing, before hybrid could be recommended for a certain region. This is in accordance with OYEKUNLE et al (2017b).…”

“…AMMI model found its application in testing of interactions genotype x environment (GxE), genotype x year (GxY), or genotype x environment x year (GxExY) in different crops: sugar beet (ĆIRIĆ et al, 2017), onion (PAVLOVIĆ et al, 2017, spring barley (PRŢULJ et al, 2015), barley (MIROSAVLJEVIĆ et al, 2014, triticale (KAYA and OZER, 2014), soybean (SOUSA et al, 2015), perennial ryegrass (LAKIĆ et al, 2015), rice (BOSE KUMAR et al, 2014), and of course maize (MITROVIĆ et al, 2012;NZUVE et al, 2013;SHIRI 2013;KRIŢMANIĆ et al, 2014;ОYEKUNLE et al, 2017a;2017b;BADU-APRAKU et al, 2011).…”

Research Article Environmental stratifications for soybean cultivar recommendation and its consistency over time.

“…For the early and medium groups, this approach provided three of the more stable clusters and for the late-maturity genotypes, four stable environmental strata were produced (Table 4). These results agree with those obtained by other authors, in environmental stratification studies for sorghum grain yield (Rakshit et al, 2012), barley (Mohammadi et al, 2009;Mortazavian et al, 2014), and maize (Oyekunle et al, 2017). The AMMI1 model offers greater repeatability of the clusters, concerning more parameterized AMMI models, which incorporate more axes of the GxE interaction (Gauch, 1990(Gauch, , 1992Ebdon and Gauch, 2002).…”

Mean Performance and Stability in Multi‐Environment Trials I: Combining Features of AMMI and BLUP Techniques

“…Studies with a relatively low percentage of explanation in IPCA1 were observed in trials with Brassica spp. (38.6–54.6% [Bocianowski et al, 2019]); sugarcane (32.9‐35% [Ramburan et al, 2011]); maize (26.4% [Balestre et al, 2009)] and 40.8% [Oyekunle et al, 2017]); and wheat (43.5% [Bornhofen et al, 2017]; 32.1% [Tigabu et al, 2017]; 33.1‐38.2% [Veenstra et al, 2019]). In this sense, biplots with low GEI pattern recovery must be interpreted cautiously in trial networks with prevalence of crossover interaction, since only the simple part of the GEI can be represented in the first main components and the complex part of the GEI may be discarded.…”