www.iita.org R.O. Akinwale1, B. Badu-Apraku*2, and M.A.B. Fakorede3 *Email:b.badu-apraku@cgiar.org Evaluation of striga-resistant early maize hybrids and test locations under striga-infested and striga-free environments Conclusions The two test locations used by IITA for screening genotypes for Striga tolerance or resistance are unique and non-redundant in their evaluation of the genotypes. The hybrids TZEI 12/TZEI 25, TZEI 11/TZEI 127, TZEI 80/TZEI 2B, and TZEI 60/TZEI 87 showed outstanding performance and should be promoted for extensive testing in the regional hybrid trials and subsequent release for commercial production. Acknowledgement The authors gratefully acknowledge the financial support of the DTMA Project and technical assistance of IITA-DTMA staff for this research. References Badu-Apraku, B., and Akinwale, R.O. 2011. Cultivar evaluation and trait analysis of tropical early maturing maize under Striga-infested and Striga-free environments. Field Crops Research 121:186–194. Yan, W. and Holland, J.B. 2010. A heritability-adjusted GGE biplot for test environment evaluation. Euphytica 171:355-369. Yan, W., Kang, M.S., Ma, S., Woods, S. and Cornelius, P.L. 2007. GGE biplot vs. AMMI analysis of genotype-by-environment data. Crop Science 47: 596-605. Introduction The usefulness of a test environment is determined by its ability to discriminate sufficiently among genotypes and its genetic correlation with target environments. Yan and Holland (2010) reported that the most appropriate GGE biplot scaling method for simultaneous evaluation of genotype as well as test environment is the heritability- adjusted (HA) GGE biplot. Badu-Apraku and Akinwale (2011) also employed genotype-by trait biplot analysis to identify superior and maize cultivars based on multiple Striga-adaptive traits. The objectives of the present study were to (i) evaluate selected Striga-resistant hybrids for grain yield performance and stability based on multiple traits and HA-GGE biplot under Striga-infested and Striga-free conditions; and (ii) assess the discriminating ability and representativeness of the test locations used for evaluating genotypes for Striga resistance by the IITA maize program. Materials and Methods Thirty Striga-resistant single–cross hybrids plus two checks were evaluated under artificial Striga infestation and Striga-free conditions at 2 locations in Nigeria in 2008 and 2009. The hybrids were selected from 378 diallel crosses among 28 parent inbreds based on superior performance under Striga infestation. The trial was laid out as pair experiment using randomized incomplete block design with two replicates under artificial Striga infestation and Striga-free conditions at Mokwa and Abuja. Data on grain yield and other important agronomic traits were collected and subjected to analysis of variance, genotype-by-trait biplot and HA-GGE biplot analysis. Results Table 1. Mean squares of grain yield and other agronomic traits of early maize hybrids with checks derived from ANOVA combined across Striga-infested and Striga-free conditions at Mokwa and Abuja, Nigeria in 2008 and 2009. Source DF Grain yield, kg ha-1 Ears per plant ASI Plant height, cm Stalk lodging, % Ear aspect LOC 1 76769335** 0.03ns 0.2ns 62260** 132.0** 10.23** YEAR 1 230322680** 0.36** 42.8** 30074** 321.9** 100.96** REP(LOC*YEAR) 4 6732312** 0.10** 1.1ns 673* 5.1ns 1.57 STRT# 1 487449204** 4.11** 87.8** 118889** 461.3** 52.87* ENTRY 31 5610312** 0.18** 7.1** 2221** 3.0ns 3.04** LOC*YEAR 1 164075840** 0.05 21.16** 11381.63 7.0 21.81ns LOC*ENTRY 31 1150769ns 0.03ns 3.0** 581** 4.7ns 0.58ns YEAR*ENTRY 31 1283238ns 0.03ns 3.8** 429* 2.5ns 0.72ns STRT*ENTRY 31 2906526** 0.08** 1.8ns 522** 1.9ns 0.94** LOC*YEAR*ENTRY 31 1093478ns 0.02ns 3.1** 361ns 4.3ns 0.67* ERROR 348 1174426 0.03 1.8 287 3.6 0.60ns *,**, Significant F-test at 0.05 and 0.01 levels of probability, respectively. #STRT = Striga infestation treatment Figure 2. Average tester coordination view of the GGE biplot showing the discriminativeness and representativeness of the test environment (a) under Striga infestation; (b) Striga-free conditions. Abbreviations; MKS for Mokwa under Striga infestation, ABS for Abuja under Striga infestation, MKN for Mokwa under Striga-free conditions, ABN for Abuja under Striga-free conditions, 1 stands for year 2008 and 2 for year 2009. Figure 3. Average tester coordination view of HA-GGE biplot showing the ranking of hybrids based on yield performance (a) under Striga infestation; (b) under Striga-free conditions. ‘E’ marked the exact position of the environments used for the ranking of the genotypes. See Fig 1 for codes of hybrids. Figure 4. Ranking of hybrids based on multiple traits (a) under Striga infestation (b) under Striga-free environments. Ideal genotypes are located inside the innermost concentric circle. ‘E’ marked the exact position of the traits used for the ranking. See Fig. 1 for hybrid codes. Figure 1. Polygon view of heritability-adjusted GGE biplot showing which genotype is best for which environment. Hybrid codes: H1 = TZEI 2 x TZEI 83; H2= TZEI 2B x TZEI 60; H3= TZEI 11 x TZEI 12; H4= TZEI 11 x TZEI 14; H5= TZEI 11 x TZEI 23; H6= TZEI 12 x TZEI 14; H7= TZEI 12 x TZEI 23; H8= TZEI 14 x TZEI 23; H9= TZEI 2 x TZEI 80; H10= TZEI 188 x TZEI 2B ;H11= TZEI 2 x TZEI 21; H12= TZEI 2 x TZEI 2B; H13= TZEI 83 x TZEI 2B; H14= TZEI 128 x TZEI 14; H15= TZEI 7 x TZEI 2B; H16= TZEI 106 x TZEI 2B; H17= TZEI 188 x TZEI 98; H18= TZEI 2 x TZEI 81; H19= TZEI 2 x TZEI 98; H20= TZEI 106 x TZEI 87; H21= TZEI 80 x TZEI 2B; H22= TZEI 136 x TZEI 11; H23= TZEI 136 x TZEI 12; H24= TZEI 136 x TZEI 14; H25= TZEI 136 x TZEI 23; H26= TZEI 11 x TZEI 127 H27= TZEI 11 x TZEI 25; H28 = TZEI 12 x TZEI 25 H29 = TZEI 23 x TZEI 25; H30= TZEI 60 x TZEI 87; CHK1= TZE COMP4C3 (Resistant OPV); CHK2=8338-1 (Susceptible). Figure 2. Average tester coordination view of the GGE biplot showing the discriminativeness and representativeness of the test environment (a) under Striga infestation; (b) Striga-free conditions. Abbreviations; MKS for Mokwa under Striga infestation, ABS for Abuja under Striga infestation, MKN for Mokwa under Striga-free conditions, ABN for Abuja under Striga-free conditions, 1 stands for year 2008 and 2 for year 2009. a b Fi . 3. Average tester coordination vi w of HA-GGE biplot showing the ranking of hybrids based on yield p rf rmance (a) under Striga infestation; (b) under Striga-free conditions. ‘E’ marked the exact position of the environments used for the ranking of the genotypes. See Fig 1 for codes of hybrids. a b Fig. 4. Ranking of hybrids based on multiple traits (a) under Striga infestation (b) under Striga-free environments. Ideal genotypes are located inside the innermost concentric circle. ‘E’ marked the exact position of the traits used for the ranking. See Fig. 1 for hybrid codes. b a Picture of a resistant maize hybrid in Striga infested plot Picture of Striga plant on susceptible maize hybrid. 1. International Institute of Tropical Agriculture (UK) Limited, Carolyn House, 26 Dingwall Road, Croydon, CR9 3EE, UK. 3. Department of Plant Science, Obafemi Awolowo University, Ile-Ife, Nigeria. ba ba b a Fig. 4. Ranking of hybrids based on multiple traits (a) under Striga infestation (b) under Striga-fre environ I eal genotypes are located insid the innermost concentric circle. ‘E’ marked the exact position of the traits used for the ranking. See Fig. 1 for hy es. b a b • Significant location effects under the two research conditions indicated that the locations were distinct from each other (Table 1). • The HA-GGE biplot revealed that TZEI 11/TZEI 25 (H27) was the best (vertex genotype) under Striga infestation at both Abuja (ABS) and Mokwa (MKS) whereas under Striga-free conditions, TZEI 12/TZEI 25 (H28) was the best at Mokwa (MKN) and TZEI 106/TZEI 87 (H16) at Abuja (ABN) (Fig. 1). • The two test locations possessed high discriminating ability (long vectors), indicating that both were important in the evaluation of the hybrids and that neither of them could be considered as redundant when screening genotypes for Striga resistance and yield potential (Fig. 2). • More promising genotypes (entries that fell inside the innermost concentric circle in Figs 3and 4) were identified using multiple traits under Striga infestation than selection based on yield performance only, suggesting that grain yield alone may not suffice as a precise predictor of Striga resistance (Fig. 3 and 4). • Based on both HA-GGE and GT biplot analyses, TZEI 12/TZEI 25 was identified for its outstanding performance under Striga-infested and Striga-free conditions. Furthermore, TZEI 11/TZEI 127 and TZEI 80/TZEI 2B were identified as promising under Striga-infested conditions and TZEI 60/TZEI 87 under Striga-free environments by the two biplot methods.