The Predicted Genetic Architecture for Number of Pods per Plant in Cowpea in Phosphorus Environments

Aims: The aim of the study was to analyze quantitative trait loci for pod number per plant in cowpea under different phosphorus environments.

Study Design: On the field, the experimental design was a split-plot with two replicates. The main plots were two phosphate levels: 0 P and 30 Kg P ha-1 (Triple super phosphate, TSP), while the 118 RILs and the two parents constituted the sub-plots randomized in a 12 x 10 α-lattice design. The experimental design for the pot experiment was a factorial randomized complete block design with two factors, and two replications. The factors were phosphorus levels (0 and 30 mg P per Kg soil) and genotypes.

Place and Duration of Study: The study was conducted at two sites. The first site was at the IRAD (Institut de la Recherche Agricole pour le Développement) research station, in Nkoemvone, in the HFZ of Cameroon while the second site was at Nkometou, a village in the Yaoundé neighborhoods, still within the HFZ of Cameroon.

Methodology: A RIL F11 population consisting of 118 lines derived from a cross between ‘58-77’ and ‘Yacine ’ using the single seed descend method was used in the study. The line ‘58-77’ (female parent,) is a black small-seeded local cultivar from Senegal resistant to pests and diseases with many pods per plant while ‘Yacine ’ (male parent) also from Senegal has large brown seeds but with very few pods per plant. Evaluation of cowpea RILs was done on low nitrogen plots both in the field and screen house and data collected on number of pods per plant. Analysis of Variance (ANOVA) was performed with the software SAS version 9.2 (SAS Institute Inc., Cary, NC, USA 2008).Marker genotype data for 118 RILs of the 58-77 x Yacine population were generated from the Illumina GoldenGate assay of 1,536 genome-wide SNP markers derived from EST sequences. The software WinQTL Cartographer 2.5. was used for composite interval mapping. QTL mapping was also performed using QTLnetwork 2.1 that uses a model that includes the effects of multiple QTL, epistasis, QTL-by-environment interactions and epistasis-by-environment interactions.

Results: Win Cartographer identified a total of eight QTL for Npod in all eight environments while QTLnetwork identified the following three main QTL (M-QTL) for Npod across the eight environments: qNpod2.1, qNpod5 and qNpod8. In total, three digenic epistatic interactions were detected for Npod across the eight environments. All three digenic pairs had epistasis main effects, and epistasis by environment interaction effect [aae] affects in one environment.

Conclusion: This study shows that, two QTL with epistasis effect were found to also have significant additive by environment effects. This means that the usual estimates of QTL effects could be confounded by epistatic interactions and result in biased estimation unless epistatic effect are isolated.