OUMA, EVANS OCHIENG (2016-05-23)

Low available phosphorus (P) remains a major limitation to maize (Zea mays L.) productivity in low P soils. Selecting maize hybrids that acquire and use P efficiently could reduce the rising costs of inorganic P. The objectives of this study were to: (i) develop P-efficient experimental maize hybrids for low P soil conditions (ii) determine the genetic effects of maize P efficiency traits under low P acid and non-acid soils (iii) determine environmental influence and stability of maize P efficiency traits in low P soils (iv) identify single nucleotide polymorphic markers (SNPs) linked to the major quantitative trait loci (QTLs) associated with P efficiency loci in a maize linkage map. A total of 30 experimental hybrids were developed using North Carolina mating design II and evaluated together with 2 checks for tolerance to low P at high P (36kgP/ha) and low P (6kgP/ha) conditions across four locations using RCBD replicated three times. For each trait, both additive and non-additive genetic effects were estimated. Environmental variation was determined using the Genotype and Genotype x Environment Interaction (GGE) and Additive Main Effect and Multiplicative Interaction (AMMI) models across 8 environments. Yield stability and superiority were determined using Finlay and Wilkinson model (FW) and Wrickes ecovalence (wi). Two hundred and twenty eight F2 individuals and 239 SNP markers were used in QTL analysis. Mean grain yield (GYLD) was significantly lower (2.49 t/ha) in the low P treatments compared to the high P sites (4.78 t/ha). Relative yield reduction (RYR) was comparable across the four locations and ranged from 42.5 - 47.7% except at Sega where it was higher (59.4%). Mean Agronomic Efficiency (AE) was 44.8 kg grain kg-1 P applied across the locations. Eighteen out of the 32 experimental hybrids exhibited AE above the locational mean > 44.8 kgkg-1. Mean phosphorus efficiency ratio (PER) of 546.7 kgkg-1 of P was obtained across the four locations with Migori exhibiting the highest mean (556.5 kgkg-1 of P. For most of the traits, greater variation was accounted for by dominance followed by epistatic and additive genetic effects in both acid and non-acid soils. The magnitude of both additive and non-additive gene effects were always greater in non-acid compared to acid soils pointing to the detrimental effects of soil acidity on gene action. The AMMI Anova showed significant effects for genotype (G), environment (E) and GEI. For GYLD, the differences among the (E) accounted for more than half (67.6%) of the total variation while the G and GEI accounted for 11.6% and 10.3% respectively of the variation indicating the existence of mega environments. 26% of the new hybrids were more stable than the commercial hybrid (H515). Based on FW model, genotypes 1, 27, 21 and 23 were considered as superior and ideal hence can be used as reference genotype in further testing. A full multi-QTL model analysis suggested six QTLs (2 QTLs each for GYLD plant height (PHT) and ear height located on chromosomes 1, 3, 4 and 8. The two QTLs for GYLD increased yield under low P soils by 173 kg/ha while the 2 for PHT increased PHT by 18.14 cm. This study has developed potential maize hybrids that can significantly improve yields in low P soils in western Kenya. The new QTLs identified will be useful for improving maize productivity in low P soils of western Kenya. v



EVANS PhDThesis Final.pdf

Files in this item