Response of cassava genotypes to four biotic constraints in three agro-ecologies of Nigeria

Eight improved cassava ( Manihot esculenta check for this species in other resources Crantz) genotypes and one local variety were grown over three years in three agro-ecological zones of Nigeria, to study their response to natural infestations of African cassava mosaic disease (ACMD), cassava bacterial blight (CBB), cassava anthracnose disease (CAD) and cassava green mite (CGM). Additionally, genotypes with stable resistance using the Additive Main Effects and Multiplicative Interaction (AMMI) statistical model was identified. Environments, genotypes and genotype x environment (G x E) interactions were highly significant (P< 0.0001) for the pests, implying that the G x E effects were sufficiently high to mask differences among genotypes. The local variety, TME1, was more tolerant (30% severity) to cassava green mite across sites, compared to other clones, while clone 30001 exhibited the highest susceptibility to the pest (54%). Genotypes (30001, TME1 and 30573) gave lowest severity scores for ACMD. Clones used in the study, especially U/41044, 4(2) 1425 and 63397, showed tolerance to CBB and CAD with severity ratings as low as 2.0. Across test sites, CGM had the highest severity scores (3.12) averaged over 3 years in Ilorin (in the Southern Guinea savanna ecology). The lowest severity scores of this disease were observed in Ubiaja and Onne. The severity of ACMD and CBB was highest in Ibadan and Ubiaja and lowest in Onne and Owerri. Cassava anthracnose disease was more severe in Ibadan (score of 2.5) and least severe in Mokwa (1.12). For CGM the highest severity was recorded in the Southern Guinea savanna (score 2.88) and lowest in the humid forest zone. Correlation coefficients calculated between pairs of biotic constraints revealed that cassava bacterial blight was negatively correlated (r=-0.78) with the green mite but positively correlated with anthracnose disease (r=0.94). The AMMI model selected AMM11 for CGM and CBB, and AMM14 and AMM13 for ACMD and CAD, respectively, as the best predictive models since these models had the smallest actual root mean square prediction differences (0.60598 and 0.37646 for CGM and CBB, respectively and 0.37297; and 0.40929 for ACMD and CAD, respectively. Genotypes 30572, 63397, 50395 and 4(2)1425 with low ACMD scores and low IPCA1 scores were the most stable to environmental changes as they combined resistance with stability of response. Such genotypes would be most appropriate in breeding programmes.