Abstract

Combining ability of inbred lines is crucial information in maize hybrid breeding programs incorporating materials from various germplasm sources. This study was conducted to assess the gene action for grain yield and other agronomic traits for germplasm having varying resistance to Striga hermonthica and genetic. In a half diallel cross of ten parents, general and specific combining abilities for grain yield, plant and ear height, plant and ear aspect, ears and plants harvested, ear rot, husk cover, moisture and resistance to Striga hermonthica were determined. The grain yields of the single crosses were significantly higher for 1368STR x TZISTR1198, TZISTR1132 x CML442, TZISTR1174 x TZISTR1198 and TZISTR1199 x TZISTR1174. The importance of both GCA (50%) and SCA (50%) for grain yield, ear rot, ear texture and ears harvested were observed, but a preponderance of GCA was existed for AUSNPC, whereas plant and ear height, plant and ear aspect, and moisture content exhibited preponderant SCA. TZISTR1174, TZISTR1162, TZISTR1192, and CML442 were good general combiners for grain yield showing highly significant positive GCA effects of 0.40, 0.2, 0.17, and 0.22, respectively while lines TZISTR1199, TZISTR1192, TZISTR1174 and TZISTR1162 were good general combiners for resistance to Striga showing highly significant negative GCA effects of-646.99,-428.21,-338.00, and-76.51. These inbred lines could be exploited in hybrid breeding to develop high yielding Striga resistant maize varieties. Hybrids such as TZISTR1174 x CML312, TZISTR1192 x CML442and TZISTR1174 x 1368STR had significant positive SCA effects for grain yield whereas crosses like TZISTR1162×TZISTR1198, TZISTR1199×TZISTR1181, TZISTR1192×1368STR had highest negative significant SCA effects of-1453.19,-1058.28, and-808.252 for AUSNPC which can be used for direct production as single cross hybrids or developed further as three way cross hybrids.

1. Introduction

Maize has a remarkable place among cereals and it is used as human food, animal feeding and industry ^{1, 2}. The identification of parental inbred lines that perform superior hybrids is the most costly and time consuming phase in maize hybrid development ³. Maize breeding programmes designed for specific-end uses, improved maize genotypes tolerant to pests/disease, and development of commercial maize hybrids usually require a good knowledge of combining ability of the breeding materials to be used. Hence, the relevance of combining ability studies for successful maize breeding ⁴. Many breeders have used combining ability to solve many maize agronomic problems. These include breeding for higher grain yield and adaptation to tropical Africa ⁵, selecting for high heterosis and adaptation to different agro-ecologies ⁶, developing maize varieties with good ear height and uniform flowering days ⁷, and identifying suitable maize inbred lines for higher grain yield and improved agronomic traits ⁸. Plant breeders and geneticists often use diallel mating designs to obtain genetic information about a trait of interest from a fixed or randomly chosen set of parental lines ⁹. The diallel analysis is an important method to know gene actions and it is frequently used by crop breeders to choose the parents with a high general combining ability (GCA) and hybrids with high specific combining ability (SCA) effects ¹⁰. Large genotype × environment effects tend to be viewed as problematic in breeding because the lack of a predictable response hinders progress from selection ¹¹, and influence the environment and interaction between genotype and environment ¹². Breeders still contend, however, that dominance effects caused by genes with over dominant gene action are also important ¹³. Most of the literature about maize, one of the most extensively studied plant species, suggests that additive effects of genes with partial to complete dominance are more important than dominance effects in determining grain yield ¹⁴. The objective of this study was therefore to evaluate ten inbred lines and their crosses using a diallel to identify type of gene action controlling the inheritance for studied traits, estimate of combining ability effects for the inbred lines and identify superior crosses and inbred lines to improve the yielding ability in maize breeding programs.

2. Materials and Methods

Fifty six inbred lines were evaluated in farmers’ abandoned naturally Striga infested fields in Nakyere, Namutumba district of Eastern Uganda during 2016 B growing season using a 7×8 alpha lattice design with two replications. Ten inbred lines Table 1 of varying resistance to Striga hermonthica were selected from the preceding study of 2016B and all possible crosses were made among the inbred lines using 10 × 10 half diallel to generate 45 single-crosses during 2017A growing season.

Seed of the successful crosses were harvested and single crosses were evaluated in a 9×5 alpha lattice replicated two times in three farmers’ abandoned naturally Striga infested fields at Nakyeere in Namutumba district; Ngerekyomu in Tororo district (Eastern Uganda) and Kinyamaseka in Kasese district (western Uganda) during 2017 B growing season. The hybrids were planted in two row plots measuring 5m in length. Planting was done at a spacing of 75cm by 25cm at the rate of two seeds per hole. 8 g of Diammonium Phosphate (DAP) were banded below the maize seed. The maize seedlings were thinned to one per stand at 14days after crop establishment. Low fertilizer dosage (50kg/ha NPK 20-10-10) was applied by broadcast to minimize the likelihood of nitrogen (N) suppressing Striga emergence ¹⁵. Hand weeding was done to remove all other weeds other than Striga. Cypermethrine was applied to control fall army worm and stem borer infestation.

Maize plant and ear heights were measured at maturity by measuring five plants selected randomly in each plot using a graduated measuring stick. Husk cover was scored for when ears were fully developed using a score rating of 1 to 5, where 1: husk tightly arranged and protracted beyond the tip of the ear and 5: ear tips fully exposed as described by CIMMYT ¹⁶ while that of ear rot was rated on the scale of 1 to 5 as described by CIMMYT ¹⁶, where; 1: little or no visible ear rot, and 5: extensive visible ear rot. Plants and ears harvested included the total number of respective plants and ears harvested per plot. Ear aspect scored on a scale of 1 to 5, where 1: best, 3: average, and 5: poorest ear aspect. Certain factors such as ear size, insect damage, grain filling, and uniformity of cob size, grain colour and texture were also considered while scoring. At physiological maturity the ears were harvested, dehusked and weighed for each genotype to determine the field weights. Furthermore, each genotype was sampled to obtain grain used for estimation of the moisture content (%). Grain yield (t ha^-1) was calculated for each genotype as follows:

Where MC = grain moisture content (%).

Striga related traits assessed included Striga count/m² at 8,10, and 12 weeks after crop emergence, Striga vigour (using a scale of 0-9), where 0= no emerged Striga plants and 9= very vigorous Striga plants (average height >40cm with >10 branches) ¹⁷, plant damage scores (using a scale of 0-9), where 1=Normal plant growth, no visible symptoms and 9 = Complete leaf scorching of all leaves, causing premature death or collapse of host plant and no ear formation ¹⁸, area under Striga number progress curve (AUSNPC) and area under Striga severity progress curve (AUSVPC) ¹⁹. The AUSNPC was calculated as follows;

where n is the number of Striga assessment dates, Yi the Striga number at the i^th assessment date, ti the days after planting at the i^th assessment date, t is 0, and Y is 0.

The analysis of variance (ANOVA) for all traits under study was carried out using Genstat release 14.1 statistical package ²⁰. The 10×10 half-diallel analysis was executed to estimate general combining ability (GCA) and specific combining ability (SCA) effects using Griffing's diallel analyses, Model 1 (fixed genotype effects), Method IV (Crosses only) ²¹, according to model; Y_ijk=µ + g_i + g_j + s_ij + e_ijk., where; Y_ijk: Observed measurement for the ij_th cross in the k_th replication/environment combination, µ: Overall mean, g_i and g_j: GCA effects for the i_th and j_th parents respectively, s_ij: SCA effects for the i_th and j_th parents, e_ijk: Error term associated with the ijth cross evaluated in the kth replication/environment combination. The interaction terms were used to test for the significance of the corresponding main effect ²². The environments and replications within environments were considered random and therefore tested against the residual error term. Mean squares of parents were estimated from the GCA effects while that of single-crosses were obtained from the SCA effects of the diallel analysis. These were further used to estimate GCA: SCA ratios ^{23, 24}.

3. Results and Discussion

Table 2 presents the combined analyses of variance. The observed significant mean squares of location and genotypes for Striga and maize agronomic traits indicated that the three environments were distinct and that there were genetic variations among the single-cross hybrids, suggesting that selection of such traits for further improvement was feasible. Similar findings were reported by Badu-Apraku et al., ²⁵. The significant G×E interaction for Striga and maize agronomic traits suggested differences in expression of traits of the set of hybrid genotypes across the locations. The expression of almost all traits was influenced by the environmental differences further suggesting the need to develop specific varieties for specific environments to take into account the high influence of the environment on the expression of traits. Similar results were reported by Olakojo et al., ¹⁵ when they assessed the performance of newly developed Striga lutea (Lour) tolerant maize genotypes including seven Striga tolerant open pollinated maize varieties. The significant mean square estimates of GCA observed indicated the important role of additive genes in the inheritance of such traits. Derera et al., ²⁶ and Vivek et al., ²⁷ reported that resistance to phaeosphaeria leaf spot (PLS) was predominately additive.

The traits with significant mean squares for SCA indicated that the non-additive gene effect contributed significantly to the inheritance of such traits and thus, selection of such traits for further improvement could be achieved through recurrent selection, and backcrossing methods. There were significant differences among GCA and SCA effects for grain yield; both GCA (50%) and SCA (50.%) were equally important for this trait. The association of both GCA and SCA with grain yield concurs with other findings ^{28, 29, 30}. Vivek et al., ²⁷ also reported that in addition to GCA, SCA was also very important in the inheritance of resistance to PLS in maize. Badu-Apraku et al., ³¹ further observed significant mean square of GCA and SCA for maize agronomic traits which suggested presence of additive and non-additive gene actions and proposed the use of hybridization, backcrossing and recurrent selection methods to develop synthetics, populations and hybrid varieties. The observed significant mean square of GCA×E interaction indicated variations in the combining abilities of the inbred lines and emphasized the need for testing the inbred lines under different environments with the view to assess performance and stability. Similar observations were made by Menkir et al., ³² and Badu-Apraku et al., ³³ in a similar study. The observation that GCA x location interaction was highly significant and greater than SCA x location interaction also agreed with other authors ^{28, 34, 35}. The lack of significant mean square estimates of SCA×E interaction for some traits suggested that expressions of such traits among the single cross hybrids were consistent across environments and therefore, good selection progress for improvement of such traits was feasible under any environment as similarly observed by Machado et al., ²⁹.

Table 3 presents the GCA effects. The observed significant positive and negative GCA effects for GY for the inbred lines TZISTR1199, TZISTR1174, TZISTR1181 and CML312 indicated the possibility of transmitting favorable alleles from the parental lines to their hybrid combinations for improved GY ^{36, 37}. Badu-Apraku et al., ³⁸ observed preponderance of GCA effect over SCA effect and suggested that the additive gene action was more important than the non-additive gene action in modulating the expression of GY and other agronomic traits. Similar observations were also made by Xingming et al., ³⁹. Parents TZISTR1199, TZISTR1174 and TZISTR1162 were considered as best general combiners for PHT and EHT due to their respective significant positive GCA effects and could be selected for increasing yield in their hybrid combinations. Tall maize genotypes with high ear positioning or placement might offer opportunity for more ears to develop on the nodes below and ultimately increasing final yield ^{40, 41} even though they may be susceptible to lodging ^{42, 43}. Number of ears per plant is an indicator for increased grain yield while genotypes with long tipped off husk covers of maize cob provide maximum protection of the ear against birds’ damage, fungal infection and early germination of kernel when moisture and conditions suitable for germination occurred in the field. Similar results were reported by Abrha et al., ⁴⁴.

Parents TZISTR1181, TZISTR1198, CML312, 1368STR and TZISTR1132 had very high GCA effects for area under Striga number progressive curve indicating susceptibility ^{45, 46}, while TZISTR1199, TZISTR1192, TZISTR1174 and TZISTR1162 had low GCA effects indicating good resistance to Striga hermonthica. Similarly, GCA effect for area under Striga severity progressive curve and Striga vigor were generally low. The least values were recorded in TZISTR1199, TZISTR1192, TZISTR1174 and TZISTR1162 indicating good resistance to Striga hermonthica, hence higher resistance level in the parents confirming its near immunity status ⁴⁷. Kim ¹⁸ reported low GCA effects for Striga hermonthica emergence and host-plant response for most resistant maize inbred lines and high GCA effects for the susceptible. Omanya et al., ⁴⁸ and Hausmann et al., ⁴⁹ reported strong genetic control for AUSNPC in the field. They observed that the parameter was a useful measure of progressive Striga development in the field. However, Hausmann et al., ⁵⁰ additionally found that individual Striga emergence count was also under genetic control from experiments conducted in pots.

Table 4 presents the SCA effects. The significant positive effects of SCA for GY for TZISTR1192 x 1368STR, TZISTR1192 x CML442, TZISTR1132 x CML442, TZISTR1174 x CML312 and 1368STR x TZISTR1198 across research locations suggested that the parents contributed favorable alleles to these single-cross hybrids. GCA effects of GY for some inbred lines were low and negative but resulted in increased SCA effects which meant that it was feasible to identify such inbred lines to develop hybrids with high yielding abilities, test and select based on increased SCA effects, even if GCA was more important in identifying potential inbred lines ⁵¹. The hybrid TZISTR1199 x TZISTR1174 showing a good SCA effect for GY was identified as single-cross tester based on assumptions described by Vivek and Pixley ⁵². Tall maize genotypes are important not only for increase grain yield but also for high biomass production for silage production ^{53, 54}. Hybrids 1368STR x TZISTR1181, 1368STR x CML442, TZISTR1162 x CML312, CML442, TZISTR1174 x CML312, TZISTR1174 x TZISTR1198, TZISTR1132 x TZISTR1174, TZISTR1192 x 1368STR and TZISTR1199 x CML442 showed good specific combining ability for plant height and could be released for silage production in intensive livestock production agro-systems. Hybrids with large number of ears per plant and long husk cover off the tip of the cob are desirable for increased yield and cob protection from several diseases and pests. Therefore hybrids such as TZISTR1199 x TZISTR1181 and TZISTR1199 x TZISTR1162 which had significantly high positive SCA effects could be considered as best combiners for ears per plant and husk cover, respectively.

Significant SCA effects recorded for some Striga related characters indicated differential response of the crosses to these Striga traits. Non-additive gene action played significant role in the inheritance of resistance to Striga in most of the crosses. Inbred lines TZISTR1199, TZISTR1192, TZISTR1174 and TZISTR1162 were identified as good combiners whose crosses had the lowest SCA making them useful in resistance to Striga breeding of maize. Kim ⁵⁵ reported that the highest level of resistance to Striga hermonthica was obtained from crosses involving two resistant parents. The results also suggested that the genes for resistance might be recessive since Striga hermonthica resistance appeared more common in resistant x resistant crosses compared with resistant x susceptible crosses. Related results were reported by ⁵⁶ when studying genetic analysis of resistance to MSV in dwarf maize germplasm. Hung and Holland ⁵⁷ also reported similar findings in their diallel analysis of resistance to Fusarium ear rot and Fumonisin contamination in maize. Kim ¹⁸ reported a negative SCA effect of-1.0 for Striga tolerant rating while studying the genetics of S. hernonthica tolerance in maize. Storey and Howland ⁵⁸ also observed that heterozygotes between resistant and susceptible lines reacted to infection in a manner intermediate between the parents where neither allelomorph was fully dominant.

4. Conclusion

The additive gene action was important in controlling Striga resistance indicating that resistance could be effectively improved through selection. There was significant genetic control for PH, EH, EHS, and EA with preponderance of additive genetic effects indicating that improvement could be achieved through selection. Maize parental lines TZISTR1199, TZISTR1192, TZISTR1174 and TZISTR1162 displayed negative GCA effects for resistance to Striga hence could be used as sources of resistance genes to Striga and could be used to introgess resistance to popular susceptible maize varieties. The parental lines TZISTR1174, TZISTR1162, TZISTR1192, TZISTR1132 and CML442 could be used as sources of genes for grain yield increment in a breeding programme.

Acknowledgements

National Agriculture Research Organization (NARO) through its Competitive Grant Scheme and the Muljibhai Madhvani Foundation Scholarship Programme provided financial support for this study. International Institute of Tropical Agriculture, International Maize and Wheat Improvement Centre (CIMMYT), and National Crop Resources Research Institute (NaCRRI) provided the germplasm used in the study.

References