Harnessing native genetic variation for resistance to field and storage insects pests in tropical maize

Introduction

Maize, Zea mays L. (Gramineae: Poales) crop yields in sub-Saharan Africa remain low as compared to the potential yields realized in more developed countries (Araus et al., 2008; Bänzinger et al., 2000). Unfortunately, smallholder farmers bear the brunt of these low yields because they are resource constrained and also depend largely on the crop for food (Mugo et al., 2005). The reasons for the low yields are partly due to poor agronomic practices, lack of use of certified seed, and more so due to insect pests both in the field and at storage (Hiruy and Getu, 2018; Mucheru-Muna et al., 2021). Field pest stem borer, Chilo partellus Swinhoe (Lepidoptera: Pyralidae), is a major pest in maize production, which is widely distributed in all maize producing ecologies of sub-Saharan Africa (Guofa et al., 2001; Ong’amo et al., 2006). In Kenya, it is found in the lowland tropics and mid-altitude agro-ecologies (Ong’amo et al., 2006). Due to global warming and climate change, the spread of this pest is rapidly moving to the highland tropics (Guofa et al., 2001; Ong’amo et al., 2006).

Infestation by C. partellus starts after emergence all the way to the milk stage of the developing grain and up to 100% infestation has been reported (Mohamed et al., 2004; Mutyambai et al., 2014). The feeding action of larvae causes damage to leaves, which reduces the photosynthetic area of the maize plant (Mutyambai et al., 2014). The developing larvae tunnel within the maize stem, creating tunnels that affect the translocation of nutrients and water from the roots to the rest of the plant, and also interfere with the translocation of photosynthates from the leaves (Mohamed et al., 2004).

The second generation of C. partellus feeds on the developing grains, and this lowers the value of the grain, as well as opening entry points for secondary pathogens to infest the developing grain. Secondary effects of C. partellus infestation include increased stem lodging of maize plants, weak maize plants that are vulnerable to attack by disease causing pathogens, and increased mycotoxin infections on the young developing cob. Yield losses of 13.5%, amounting to an estimated US$90 million in Kenya, were attributed to this pest (De Groote, 2002).

On the other hand, storage pests, maize weevil, Sitophilus zeamais Motschulsky (Coleoptera: Curculionidae), is a major pest in maize storage. Maize grain yield losses due to this pest are estimated at 15%–30% in developing countries (Derera et al., 2014). Infestation starts in the field especially on the open tipped varieties (Abebe et al., 2009). The complex and merciless handing over of grain yield losses—from stem borers in the field, to maize weevil in storage—is a major threat to food security in Africa.

Breeders, in an effort to address this challenge, have developed maize hybrids with separate resistance to each of these insect pests: stem borer resistant (SBR) hybrids, and storage pest-resistant hybrids. There is, however, no record of hybrids with dual resistance to these pests.

Previous studies have extensively documented the biology, ecology, and management of stem borers in African maize systems. Stem borers, including C. partellus and Busseola fusca Fuller (Lepidoptera: Noctuidae), remain among the most destructive insect pests, causing significant yield losses across agro-ecologies (Kfir et al., 2001; Khan et al., 2011). Their impact is expected to increase under changing climatic conditions, which favor pest survival, reproduction, and geographical expansion (De Groote et al., 2020). To mitigate these effects, several management strategies have been developed, including host plant resistance, biological control, cultural practices such as intercropping, and transgenic approaches such as Bt maize (Khan et al., 2011). In addition to breeding, integrated pest management strategies have been widely promoted. The push–pull system, for example, has proven effective in reducing stem borer infestation while improving soil fertility and crop productivity (Khan et al., 2011). Improved storage technologies such as hermetic storage bags have also been effective in reducing post-harvest losses caused by maize weevil (Tefera, 2012). Despite these advances, most approaches target either field or storage pests independently.

Host plant resistance has been a key strategy in managing stem borers. Significant genetic variation for resistance has been identified in tropical maize germplasm, enabling the development of improved inbred lines and hybrids (Mugo et al., 2005). Resistance is largely quantitative, controlled by multiple genes, and associated with traits such as stem strength, leaf toughness, and biochemical defenses (Dhliwayo et al., 2005). Advances in molecular breeding have facilitated the identification of genomic regions associated with resistance, supporting marker-assisted selection and genomic selection approaches (Prasanna, 2012). However, resistance in most commercial varieties remains partial and can be influenced by environmental conditions.

Similarly, resistance to storage pests such as S. zeamais has been explored in maize breeding programs. Resistance mechanisms have been linked to kernel characteristics including grain hardness, pericarp thickness, and the presence of secondary metabolites (Kim and Kossou, 2003). Studies have shown substantial genetic variation for resistance, indicating the potential for genetic improvement (Dhliwayo et al., 2005). However, the inheritance of resistance is complex, involving both additive and non-additive gene effects which can complicate breeding efforts (Dhliwayo and Pixley, 2003).

Recent studies have shown that multiple insect pests often occur sequentially or simultaneously in maize production systems, resulting in cumulative yield losses that frequently exceed those caused by individual pests alone (Prasanna et al., 2021). Consequently, there is increasing recognition that maize improvement requires resistance mechanisms that operate across both field and post-harvest stages of the production cycle. Most resistance efforts in tropical maize have focused on individual insect pests. For example, substantial progress has been made in developing germplasm resistant to stem borers, including populations with resistance to multiple borer species such as B. fusca and C. partellus (Job et al., 2022; Mugo et al., 2005; Tefera et al., 2016). Similarly, breeding efforts targeting post-harvest insect pests have identified maize genotypes with resistance to S. zeamais and other storage pests (Meireles et al., 2025; Mwololo et al., 2012; Sserumaga et al., 2021). Other studies have combined insect resistance with agronomic traits such as stalk strength, lodging resistance, drought tolerance and yield stability to enhance overall field performance (Badu-Apraku and Fakorede, 2017; Murenga et al., 2015). For instance, Job et al. (2022) evaluated maize hybrids for resistance to both stem borers and fall armyworm, whereas Sserumaga et al. (2021) assessed tropical maize testcrosses for grain yield and maize-weevil resistance, demonstrating continued progress in addressing specific pre-harvest and post-harvest insect challenges. However, these studies primarily addressed resistance within a single production phase and did not simultaneously target insects occurring before and after harvest.

To date, limited information is available on the genetic basis and breeding potential of combining resistance to both field insect pests and storage pests within the same tropical maize germplasm. In addition, the genetic relationships between resistance mechanisms expressed during crop growth and those expressed during grain storage remain poorly understood. Therefore, the present study addresses an important knowledge gap by investigating the inheritance and combining ability of resistance to C. partellus and S. zeamais simultaneously. The objective of this study was to identify maize lines that can combine to generate hybrids with combined resistance to both C. partellus and S. zeamais and contribute to food security among smallholder farmers in sub-Saharan Africa.

Materials and methods

Germplasm

The parental inbred lines used in this study were obtained from the International Maize and Wheat Improvement Centre (CIMMYT). The levels of resistance to both field and storage pests were evaluated and documented by earlier research efforts (Beyene et al., 2012; Mwololo et al., 2012). These parental lines were selected from previous CIMMYT breeding and screening programs based on their demonstrated resistance to either the stem borer (Chilo partellus) or the maize weevil (Sitophilus zeamais). As a result, they were treated as fixed and the results of the study pertain to this germplasm set.

A total of 40 maize inbred line parents were obtained from two populations in the maize breeding program at CIMMYT-Kenya (Table 1a). Twenty from SBR population and 20 from storage pest resistant (SPR) population. These parents had varying levels of resistance to stem borer (C. partellus) in the field and maize weevil (S. zeamais) in storage. The 40 inbred lines were divided into groups of five parents based on resistance levels. Eight sets of 25 hybrids each were then formed using a 5 x 5 North Carolina II mating design. The SBR lines were used as female parents and SPR lines were used as males (Table 1a). A total of 200 single crosses were expected (25 from each of the eight sets) (Table 1b). However, 10 crosses were unsuccessful leading to a total of 190 hybrids with sufficient seeds for evaluation. In addition, five commercial check hybrids were used alongside these test hybrids in trials (Table 1b).

Table 1a. List of 40 maize inbred lines and their role in North Carolina II design mating scheme.

†Cpr/Szs, Chilo partellus resistance but Sitophilus zeamais susceptible; Cps/Szr, Chilo partellus susceptible but Sitophilus zeamais resistance.

Table 1b. Number of experimental hybrids derived from North Carolina II mating design within eight sets evaluated across four environments in Kenya.

#ENV, Environment.

Experimental sites

The nurseries were planted at Kiboko (Environment 0) in 2013A, while the evaluation sites were at Kirinyaga, Embu and Kakamega sites during the 2013B and 2014A seasons. Season by site interaction was classified as one environment (Table 2).

Table 2. Experimental sites and environments for nursery and evaluation of hybrids in Kenya.

†masl, meters above seas level; ENV, environment.

Design of the experiment

The nurseries

Female lines were planted in blocks of twenty rows per entry with five rows of male per entry. This allowed one male row for every four female rows. The males were planted in one row planting staggered 3 times in intervals of -5, 0, + 5 days, corresponding to the female planting. These males were planted adjacent to each of the female blocks, each comprising 20 rows. The plots were 5-meter rows per entry with row-to-row distance of 75 cm and hill to hill distance of 25 cm. Two seeds were sown and later thinned to one plant per hill. Different color tags were used to label females and males. Chemical stem borer control was applied on seedlings at the 6-leaf, and 10-leaf stages. Routine agronomic practices of weeding, fertilizer application and rogueing off-types were practiced to ensure clean, uniform, and healthy plants. Shoots were covered as they emerged in all plants.

Pollen was harvested and bulked from at least four plants in the male rows. As many females as possible were hand pollinated, to ensure large number for seeds for multi-location field-testing. Harvested ears from each female were kept separate. The rotten ears were eliminated, and the seeds from each female were dried, shelled, weighed and protected with a suitable chemical pesticide.

Evaluation of F1 crosses for resistance to stem borer and storage pests’ resistance

The experimental model integrated field and laboratory evaluations to assess combined resistance to C. partellus and S. zeamais in maize germplasm. The hybrids were evaluated in a 13 x 15 α-lattice design with three replications at Kirinyaga University College, Embu and Kakamega sites in Kenya during the 2013B and 2014A seasons (Table 2). The plots were two rows per entry, with same hill to hill distance and row length as described for the nurseries. The plots were partitioned into two sections 3 weeks after planting using a nylon rope to separate the section for infesting from the one to be protected. The partitioned genotypes were artificially infested with C. partellus neonates as described by Tefera et al. (2010). The plants were allowed to develop up to maturity. At harvest the hybrids were evaluated for resistance to S. zeamais under controlled storage conditions.

Data collection and analysis

Data on leaf damage were collected 2 weeks and 4 weeks after infestation using a scale of 1–9, where 1 is clean and 9 is heavily damaged (dead hearts) (Beyene et al., 2012). The second scoring which is the determinant score was subjected to analysis while the first rating was used to ascertain that the insects inflicted damage on leaves and therefore the infestation was effective. At harvest, grain yield, exit holes, and cumulative tunnel length data were collected. Grain yield was then adjusted to 12.5% moisture (Bänzinger et al., 2000), using the formula:

Grain Yield (t/ha) = [Grain Weight (kg/plot) x 10 x (100−MC)/(100−12.5)/(Plot Area)],

Where MC = Grain Moisture Content. Standard heterosis was calculated using formula:

100 × ((F1− MC)/MC).

Where: F₁​ = experimental hybrid means

MC = mean of commercial check hybrids.

The harvested grain was screened for post-harvest insect pest resistance data using the protocol described by Tefera (2012) (post-harvest insects’ resistance screening and data collection). This data includes percent undamaged kernels and grain weight loss.

Data analysis was conducted in accordance with the North Carolina Design II nested within sets evaluated across environments as originally formulated by Hallauer and Miranda (Hallauer and Miranda, 1981, Hallauer et al., 2010) which partitions genetic, environmental and interaction variations to cleanly isolate General Combining Ability (GCA) and Specific Combining Ability (SCA) effects. The analysis of variance was done using PROC MIXED model of SAS, for all the traits (SAS Institute, 2014).

The following elements of the model were treated as random: sets, environments, blocks within replications within sets, all interactions of the environments with male, female and male x female within sets. The following elements of the model were treated as fixed: male, female and male x female within sets. The male, female and male x female within sets represents GCA associated with males, GCA associated with females and SCA effects within sets, respectively.

The full linear model is indicated below:

Y_ijklpq = µ +S_p+g_fi (S_p)+g_mj (S_p)+h_ij (S_p)+E_q+r_kl(SE)_pq+(SE)_pq+(gfE)_iq(S_p) +(gmE)_jq(S_p) +(gfgmE)_ijq(S_p) +e_ijklpq

where:

Y_ijklpq = observed value of the trait for the cross between the (i^th) female and (j^th) male in the (k^th) replication, (p^th) set, and (q^th) environment;

mu = overall grand mean;

(S_p) = effect of the (p^th) set;

(g_fi (S_p)) = GCA effect of the (i^th) female parent nested within the (p^th) set;

(g_mj(S_p)) = GCA effect of the (j^th) male parent nested within the (p^th) set;

(h_ij (S_p)) = SCA effect of the cross between the (i^th) female and (j^th) male parent nested within the (p^th) set;

(E_q) = effect of the (q^th) environment;

(r_kl(SE)_pq) = effect of the (l^th) block nested within k^th replication nested within the p^th set and q^th environment;

((SE)_pq) = set × environment interaction;

((gfE)_iq (S_p)) = interaction between the GCA effect of the (i^th) female parent and the (q^th) environment, nested within the (p^th) set (Female GCA × Environment);

((gmE) _jq (S_p)) = interaction between the GCA effect of the (j^th) male parent and the (q^th) environment, nested within the (p^th) set (Male GCA × Environment);

((gfgmE)_ijq (S_p)) = interaction between female GCA, male GCA, and environment nested within the (p^th) set (Female GCA × Male GCA × Environment);

(e_ijklpq) = residual experimental error.

Under the North Carolina II mating design, female and male parents were evaluated separately for their general combining ability (GCA) namely GCA male and GCA female effects within sets, while the female × male interaction represented the SCA effects. Individual GCA effects for female and male parents and SCA effects for individual crosses were estimated from the fitted linear model.

Results

Gene action

The results showed significant differences (p ≤ 0.05) among hybrids for grain yield under protected and insect infested conditions as well as insect resistance parameters such as leaf damage scores (LDS), grain weight loss and undamaged kernels. The ANOVA revealed that significant differences existed for grain yield, LDS and undamaged kernels for test hybrids (Table 3a). Significant differences were also observed among male and female parents within sets for grain yield and insect resistance-related traits. The male x female interaction within sets was significant for grain yield under protected (Gyp) and infested (Gyi) conditions and for undamaged grain (M-uDgrn), whereas no significant differences were observed for LDS or percentage weight loss (MWtls) (Table 3a). The variance component analysis showed significant male and female effects within sets across all traits (Table 3b). However, the male x female variance component was significant for grain yield under insect infestation, undamaged grain and grain weight loss, but not for grain yield under insect protected conditions and leaf damage score (Table 3b).

Table 3a. Analysis of variance for grain yield, leaf damage scores, percent grain weight loss and grain damage for 190 experimental maize hybrids in Kenya across four environments.

‡Data: DF, degrees of freedom; Gyp/i, grain yield (t/ha), protected/infested; LDS, leaf damage scores; MWTLS, percent grain weight loss; M-uDgrn, percent undamaged grain under weevil infestation.

*Adjusted DF due to missing crosses.

Table 3b. Variance components for grain yield, leaf damage scores, percent grain weight loss and grain damage for experimental maize hybrids in Kenya across four environments.

‡Data: Gyp/i, grain yield (t/ha), protected/infested; LDS, leaf damage scores; MWTLS, percent grain weight loss; M-uDgrn, percent undamaged grain under weevil infestation.

General combining ability (GCA) effects associated with both male and female parents were highly significant (p ≤ 0.01) for all traits indicating involvement of additive gene effects. SCA effects were significant for grain yield under protected and insect infested conditions, and undamaged kernels indicating the role of non-additive gene effects for these traits. However, the SCA effects were not significant for leaf damage and grain weight loss (Table 3a).

The relative contributions of female, male, and female × male effects to the total sum of squares varied among traits, indicating differences in the genetic control of trait expression. For grain yield under protected conditions (Gyp), male effects accounted for the largest proportion of the total variation (60.34%), followed by female effects (25.30%) and female × male interactions (14.37%). A similar pattern was observed for grain yield under insect infestation (Gyi), for which male effects accounted for approximately 59.51% of the total variation, while female and female × male effects each contributed approximately 20.2% (Table 4).

Table 4. Hybrids sum of squares for grain yield, leaf damage scores, percent grain weight loss and grain damage across four environments in Kenya.

Data: Gyp/i, grain yield (t/ha), protected/infested; LDS, leaf damage scores; MWTLS, percent weight loss; M-uDgrn, percent undamaged grain under weevil infestation; SS, sum of squares; % Cross SS, percentage cross sum of squares.

For leaf damage score (LDS), female × male effects contributed the largest proportion of the variation (40.60%), followed by male effects (38.08%), whereas female effects accounted for 21.23% (Table 4). A similar but less pronounced contribution of both male and female × male effects was observed for mean weight loss (MWTLS). Male effects accounted for 45.48% of the total variation, followed by female × male interactions (33.40%) and female effects (21.12%).

In contrast, variation in mean undamaged grain (M-uDgrn) was predominantly explained by female effects, which accounted for 65.48% of the total variation, compared with 16.05% for male effects and 18.47% for female × male interactions (Table 4).

General combining ability of parents

General combining ability (GCA) effects associated with both male and female parents were highly significant (p ≤ 0.01) for all traits (Table 3a). The GCA effects of parents within sets are shown in Table 5. Positive GCA values were favorable for grain yield and undamaged grain while negative GCA was desired for LDS and grain weight loss (Table 5). Female parent 3 had positive and favorable GCA effects for grain yield under protected and infested treatments, as well as positive and favorable effects for undamaged grain. The same parent 3 had favorable negative effects for LDS due to C. partellus infestation, and negative and favorable effects for weight loss when infested with S. zeamais. Similar results were observed with the same parent when used in set 5, where favorable GCA effects were obtained for grain yield, for LDS, for weight loss in maize weevil infestation and for undamaged grain under maize weevil infestation (Table 5). The male parent 22 exhibited favorable GCA effects for all the traits in consideration in set 2. However, when parent 22 was used with different females in set 5, the GCA effects were not favorable for undamaged grains. Similar results were observed with parent 36 in sets 4 and 6 (Table 5).

Table 5. General combining ability of maize inbred lines within sets for grain yield, leaf damage scores, undamaged grain and grain weight loss.

Data significance: ***means data is significant at p ≤ 0.001; ***data is significant at p ≤ 0.01.

The highest and positive favorable GCA for grain yield were observed in female parent 12 (1.37 and 0.85) in set 7, while the lowest was in parent 1 (-1.46 and 1.23) for protected and infested respectively.!!!Leaf damage scores highest negative GCA was contributed by parent 6 (-0.36) in set 2, while parent 36 contributed the highest positive GCA effect (7.6 and 7.05) for undamaged kernels in sets 6 and 4 respectively. Overall, favorable GCA obtained translated to yield grain of between 37% infested to 45% protected respectively, leaf damage scores (42%), and weight loss (45%) and, undamaged grain (53%) for the female parents. The male parents contributed favorable GCA for grain yield (53% and 55%) protected and infested respectively, 45% for leaf damage scores, 43% weight loss, and, 48% of undamaged kernels (Table 6).

The SCA effects were highly significant for grain yield under both protected and infested conditions, LDS and maize weevil damage (Table 3a). Variance components for GCA males, GCA females, and SCA were all significant for all traits except grain yield under insect protection and LDS (Table 3b). The SCA sum of squares ranged from 14%–40% for all traits (Table 4). The SCA effects of hybrids are shown in Table 6. The following four hybrids: 17x 40, 8 x 40, 2 x 32, 12 x 32 displayed positive and significant SCA effects for grain yield under both protected and insect infested conditions. The following hybrids (7 x 40 and 12 x 34) had favorable SCA effects for yield under protected conditions, while the hybrids 19 x 40, 4 x 34, and 15 x 34 exhibited favorable SCA effects under insect infested conditions. The hybrids 17 x 40, 12 x 32, and 7 x 40 showed favorable SCA for both yield and undamaged grains. Hybrids 3 x 40, 12 x 23, and 14 x 32, displayed favorable SCA for undamaged grains (Table 6).

Table 6. Specific combining ability of maize lines with contrasting resistance to stem borers and maize weevil for yield and insect traits.

‡Data: SCA, specific combing ability; Gyp/i, grain yield (t/ha), protected/infested; LDS, leaf damage scores; MWTLS, percent grain weight loss; M-uDgrn, percent undamaged grain under maize weevil infestation.

Seven hybrids (8 x 40, 4 x 34, 14x 34, 3 x 34, 11 x 32, 12 x 21, and 11 x 34) showed significant and negative SCA effects which is favorable for LDS. The parent 34 was involved in four of these hybrids. Four hybrids (17 x 40, 12 x 34, 2 x 32, and 7 x 40) displayed favorable SCA effects for grain weight loss. The hybrids 17 x 40, 2 x 32, and 7 x 40 demonstrated their superiority for most of traits as they combined high grain yield and insect resistance traits (Table 6).

Heterosis and value for breeding

Heterosis was observed for grain yield, LDS, weevil weight loss, and, undamaged grains in the test hybrids (Table 7). Heterosis for grain yield when protected ranged from 26%–41% for the best five hybrids, and 26%–37% when infested. The worst affected five hybrids displayed -45% to -61% reductions in grain yield (Table 7).

Table 7. Percentage standard heterosis of maize hybrids for grain yield under protection and infestation with Chilo partellus across four environments in Kenya.

Inbred line parents of hybrids that appear at least 2 times among the best 5 hybrids in this table are highlighted in bold.

Heterosis for LDS for the best five hybrids was -22% to -17%, while weight loss, registered -45 to -32%. Conversely, the worst five hybrids had a reduction of 15%–22% increase in LDS, while weevil weight loss had an increase of 25%–33% (Table 7).

Heterosis values for insect pests’ resistance parameters for test hybrids

The best performing hybrid, single cross 17 x 40, had percent heterosis of 29%–37% both when protected and infested (Table 7). The same hybrid had grain yields of 154% higher above the population mean. Other hybrids that had high heterosis and reasonable value for breeding are single cross 19 x 40 and single cross 3 x 25. Conversely there are some hybrids which performed badly with the worst single cross 6 x 36, yielding 24% grain yield below the population mean (Table 8).

Table 8. Grain yield percentage over and above the population mean (100) of best ten hybrids across four environments in Kenya.

†Data: Gyp/Gyi, grain yield (t/ha) protected/infested; LDS, leaf damage score; MuDKN, undamaged grain under maize, Mwtls, weight loss due to maize weevil.

Weevil infestation; Mwtls, weight loss under maize weevil infestation. †Means with the same letter are not significantly different.

Mwtls; weight loss due to maize weevil.

Inbred line parents of hybrids that appear at least 2 times among the best hybrids, and the parents of the worst hybrid 6x6 in this table are highlighted in bold.

Discussion

Gene action

The study indicated that there is native genetic variation for resistance to field and storage insect pests in this set of tropical inbred lines and hybrids which can be exploited in the breeding program through hybridization and selection. The presence of significant GCA and SCA for grain yield, LDS and undamaged kernels indicates that both additive and non-additive gene action was important for conditioning combined resistance to C. partellus and S. zeamais in maize hybrids in the current study. The predominance of male effects, accounting for 60% of sum of squares of the crosses, for grain yield under both protected and insect-infested conditions suggested that the GCA of lines used as male played the major role in determining hybrid grain yield. The females sum of squares accounted for 25% of the crosses indicating that they also played a role in conditioning the hybrid grain yield and insect resistance. Total GCA accounted for 60% of the hybrid sum of squares for LDS while SCA effects accounted for 40% indicating that genes with both additive and non-additive effects were crucial in governing hybrids resistance to stem borer infestation. A different trend was observed for grain weight loss due to maize weevils where the total GCA effects accounted for 77% of the cross sum of squares implying that additive gene action had more significant role for insect resistance in hybrids. The GCA associated with female lines accounted for 66% of the sum of squares of the crosses while the male counterparts accounted for 16% sum of squares of the crosses, indicating that female parents played a more significant role in conferring grain resistance to maize weevil damage. Although the SCA accounted for 18% of the crosses sum of squares, it was significant for undamaged grain indicating that the role of non-additive gene action cannot be ignored in breeding hybrids for maize weevil resistance. Similar findings were reported when investigating resistance of early-generation maize inbred lines and their hybrids, where both additive and non-additive gene action were found to contribute to weevil resistance (Dari et al., 2010). Overall, the results suggest that both hybridization and selection would be effective to develop productive and insect resistant maize hybrids. Breeding for insect resistance and yield can be achieved in the same hybrid, with the right parents. This is because hybrids such as 17 x 40, 2 x 32, and 7 x 40 demonstrated superiority for most traits, as it combining high grain yield with insect resistance.

These findings can be used when making selections for breeding for insect resistance through recurrent selection and pedigree methods. The presence of non-additive gene action places emphasis on the need to consider dominance effects. Comparable results were reported in studies on genetic resistance to pink stem borer in maize (Butrón et al., 2009).

It can be concluded that there was a balanced contribution of both the male and female parent for insect resistance. This makes it complicated to deduce that additive gene action was more predominant over non-additive as earlier studies suggested (Menkir and Ayodele, 2005). These findings are also in agreement with previous work on breeding for yield potential, stress tolerance, and yield stability in maize (Derera, 2005). Breeding for insect resistance and yield can be achieved in the same hybrid, with the right parents, the correct selection procedure and the right model.

Mean performance of eight sets

Set 1 F1 hybrids gave the lowest grain yield when under protected and infested treatment. The female parents from SBR were resistant to C. partellus but susceptible S. zeamais, while the male parents from SPR population were susceptible to C. partellus but resistant to S. zeamais. The challenge of high resistance to the insect pests but low yields observed in set 1 could mean that there is presence of tightly linked genes for insect resistance with undesired low yields.

The parental combination for set 7 gave the F1 hybrids most resistant to both C. partellus and S. zeamais. Ironically, the parents, though drawn from different populations—females from SBR and males from SPR—had similar attributes in terms of resistance, with both being susceptible to C. partellus, but resistant to S. zeamais. The F1s from this set showed higher yields and combined resistance to the two pests. This can be explained as having some linkage disequilibrium for C. partellus resistance, which breaks when crossed, and the resistance expressed in the F1 progeny. This also explains the importance of screening maize germplasm for resistance to other biotic stresses even when they have not been developed for resistance to the particular stress. The SPR population was screened for C. partellus and the SBR population screened for S. zeamais resistance. This is useful for sourcing for resistance genes to insect pests. The findings corroborate previous studies demonstrating the presence of available sources of maize weevil resistance (Mwololo et al., 2012). These resistance sources can be explored for combined resistance. Other researchers reported resistance to each of the pests in hybrids without compromising yield (Beyene et al., 2012; Butrón et al., 2012).

General combining ability

Female parent 3 exhibited positive general combining ability (GCA) effects for grain yield, undamaged kernels, leaf damage score, and weight loss in sets 1 and 5, while male parent 22 showed positive GCA effects in Set 2. These results suggest that these parents contribute desired additive genetic effects for resistance to both C. partellus and S. zeamais, making them valuable parental lines for developing hybrids with combined insect resistance. The consistent positive trends in GCA across multiple traits indicate that these parents could serve as reliable donors of resistance alleles in maize breeding.

Male parents 22 and 38 also exhibited positive GCA effects for all evaluated traits except undamaged kernels. This indicates that although these parents possess substantial breeding value for most resistance-related traits, they may require complementary crossing with parents exhibiting strong GCA for undamaged kernels to achieve balanced improvement across all target traits. Such complementary parent selection is an important strategy for combining maize weevil and stem borer resistance.

Female parent 12 displayed good GCA for grain yield, suggesting that it is an excellent donor for improving productivity. However, because its contribution to insect resistance was comparatively limited, its greatest value would be realized when crossed with parents possessing strong additive effects for resistance traits. Likewise, parents 6, 31, 33, 36, and 38 exhibited desired GCA effects for leaf damage score, weight loss and undamaged kernels, indicating that they represent useful sources of additive resistance genes for incorporation into maize breeding (Fasahat et al., 2016). Several parental lines exhibited positive GCA effects for some traits but undesired effects for others. This reflects the quantitative nature of insect resistance and yield-related traits, which are polygenic and may have negative correlations among traits. These parents can be combined with complementary parents to pool desired alleles and maximize additive genetic gain for improving both insect resistance and grain yield in hybrid combinations (Fasahat et al., 2016).

The contribution of both male and female x male effects for grain weight loss indicates that both additive and non-additive genetic effects contributed to variation in storage insect damage with male effects making a relatively larger contribution to general combining ability. The predominance of female effects indicates that genetic differences among female parents contributed substantially to variation in the proportion of undamaged grain. This result further suggests that selection of appropriate female parents may be particularly important for improving resistance to storage insect damage as measured by undamaged grain. Overall, the predominance of male GCA effects suggests an important contribution of additive genetic effects.

Specific combining ability

The SCA effects showed considerable variation among the parental combinations for grain yield and insect resistance traits, indicating an important contribution of non-additive gene action to hybrid performance. For grain yield under insect protection (Gyp), SCA effects ranged from -0.46 to 1.02. This range is within the level of variation reported in stem borer resistance breeding materials. Murenga et al. (2015) reported grain yield SCA effects of 0.19 to 1.06 at Embu and 0.21 to 0.80 at Kakamega among selected maize testcrosses evaluated under Busseola fusca infestation. Similarly, Gakunga et al. (2012) detected significant SCA effects for grain yield among crosses involving Chilo partellus-resistant maize inbred lines. These results together with the present finding, show that particular parental combinations would make a substantial contribution to hybrid yield especially when the hybrids are exposed to insect pressure. Under insect infestation, grain yield SCA effects ranged from -0.07 to 1.02. The positive effects observed for several crosses indicate that some combinations were able to maintain better yield than expected from their parental performance under insect pressure. Gakunga et al. (2012) also reported significant SCA for grain yield under C. partellus infestation while Murenga et al. (2015) observed significant SCA for grain yield under Busseola fusca infestation. The occurrence of favorable SCA under both protected and infested conditions suggests that specific parental combinations contribute to maintaining yield while also responding differently to insect pressure.

For leaf damage, SCA effects ranged from -0.30 to 0.27. The negative effects are desirable because they indicate combinations with lower leaf damage than expected. The magnitude was smaller than reported by Murenga et al. (2015), who recorded SCA effects of approximately -2.43 to -1.29 at Embu and -2.43 to -1.30 at Kakamega for leaf damage among selected testcrosses under Busseola fusca infestation. The difference in magnitude reflects differences in the parental germplasm, infestation levels and environments. Nevertheless, the favorable direction of the SCA effects was consistent, supporting the essentiality of specific parental combinations in improving stem borer resistance. The maize weevil traits showed a wider range of SCA effects from -3.08 to 3.44 for MWTLS and -6.57 to 4.17 for M-uDgrn. This wide variation implies that there are substantial differences among crosses in their response to maize weevil damage. Dhliwayo et al. (2005) reported significant SCA effects for maize-weevil resistance among 14 southern African maize inbred lines, demonstrating that specific parental combinations contributed to Sitophilus zeamais. Sodedji et al. (2018) also reported significant SCA variation for maize weevil resistant traits including ranges of -30.50 to 34.80 for F1 insects emerged and -17.24 to 21.71 for grain damage. However, the significant contribution of SCA confirms that hybrid performance was influenced by non-additive effects, underscoring the importance of evaluating both parents’ combining ability and specific cross performance. This agrees with Hallauer et al. (2010) who reported that the expression of grain yield in maize is determined by the combined action of additive and non-additive genetic effects with their relative importance varying according to the germplasm under evaluation. A recent systematic review of combining ability studies in sub-Saharan Africa by Oladiran et al. (2026) reached a similar conclusion showing that although additive gene action predominates for many agronomic traits, significant SCA effects are consistently associated with improved performance under biotic stress and enhanced hybrid adaptation. In general, these findings imply that breeding for combined resistance to C. partellus and S. zeamais should integrate the selection of parents with superior GCA and identification of hybrids combinations with favorable SCA to maximize both grain yield and insect resistance.

Hybrids resistance to insect pests

Variation among test hybrids existed. This is evidenced by the observed differences in genotypes to infestation by C. partellus and S. zeamais, an indication that the hybrids were diverse in terms of response to infestation. These differences were registered as levels of LDS, percentage of weight loss and percentage of undamaged grain after infestation. Use of such parameters as measures of resistance has been documented by previous authors when screening maize germplasm for resistance to insect pests (Derera et al., 2014; Mwololo et al., 2012). LDS have been tested and validated as a measure of resistance and susceptibility to stem borers in Kenya (Tefera, 2012).

Resistant maize varieties were reported to have lower LDS compared to susceptible varieties (Beyene et al., 2012). Using this parameter as a determinant of C. partellus resistance, the test hybrids in this study possessed reasonable resistance as indicated in the results where 8.9% were highly resistant, 41.6% resistant, 38.4% moderately resistant, and, 11% susceptible. The resistance distribution in the population followed the normal distribution curve, an indication that genetic variation existed among the test hybrids (Kim and Kossou, 2003).

Similarly, weight loss and grain damage have been used in previous studies involving breeding for post-harvest resistance in maize. The varieties with resistance to S. zeamais had low weight loss compared to susceptible maize varieties (Abebe et al., 2009; Mwololo et al., 2012). Grouping of the test hybrids into resistance groups also revealed that maize weevil resistance was normally distributed.

The parameters have, therefore, been found to discriminate and separate genotypes into resistant and susceptible groupings based on levels of damage and weight loss. Another study showed that percent grain weight loss was a conservative indicator of resistance in maize varieties (Matewele, 2014). It is this indicator that is preferred for the current study so that it can discretely discriminate hybrids that are truly resistant to S. zeamais from those that are susceptible.

In the current study combination of high yields when infested, low LDS, low weight loss due to weevil infestation and high percentage of undamaged grains suggests that the hybrids in sets 3, 7 and 8, had combined resistance to C. partellus and S. zeamais insect pests. The best check performance when protected and the drop in yield when infested emphasize the importance of incorporating insect resistance in breeding programs. Further, at storage the minimum undamaged grain percentage is an indication that yield loss at storage is of great economic importance.

The findings of this study show also that breeders can select for combined resistance from within their breeding program and enhance the resistance of insects in the existing germplasm. It can therefore be concluded that favorable alleles of combined resistance to C. partellus and S. zeamais can be fixed through selection.

Heterosis and value for breeding

High percentage values for increased vigor in test hybrids are an indication of progress in insect pest resistance. The high reduction in leaf damage indicates good resistance genes combination. It is not without a challenge though, because some of the test hybrids demonstrated increased LDS as well as reduced grain yield. This is an indication of the variation existing in the test hybrids from which breeders can select what is desirable and discard what is not favorable. These findings are consistent with earlier work demonstrating the existence of good sources of resistance to both maize weevil and stem borers in Kenya (Mwololo et al., 2012). Using these sources, some hybrids, for example, single cross 17 x 40 and 19 x 40 had good heterosis and performed well above population mean. It can be deduced that inbred line 40 had favorable genes for grain yield, even when infested. The hybrids can be said to be tolerant because despite the moderate levels of resistance observed, they yielded well. These hybrids can be recommended for advancement in mid-altitude areas.

Conclusion

The study successfully achieved its objective of combining resistance to C. partellus and S. zeamais in elite maize germplasm through breeding and selection. Significant genetic variability was observed among the test hybrids for insect resistance and grain yield, indicating the presence of exploitable genetic variation for simultaneous improvement of these traits. The relatively large contribution of female × male interactions confirms significant SCA effects indicating an important role of non-additive genetic effects in the expression of resistance to stem borer damage. This suggests that the performance of specific parental combinations was important in determining leaf damage response, beyond the average performance of the individual parents. The hybrids that displayed favorable SCA effects were identified and will be advanced in the breeding program. Several hybrids, particularly those derived from sets 3, 7, and 8, expressed combined resistance to C. partellus and S. zeamais while maintaining superior grain yield across environments, demonstrating that resistance to both field and storage insect pests can be effectively incorporated into high-performing maize hybrids. Among these, four hybrids demonstrated superior performance for grain yield under insect protection and infestation (17 x 40, 19 x 40, 3 x 25, 2 x 23). These best performing hybrids can be advanced for preliminary yield trials and subsequently released for production. Significant genetic variation among the test hybrids, together with significant GCA and SCA effects, indicated that both additive and non-additive gene actions contribute to the inheritance of resistance and grain yield. The identification of superior female and male general combiners, together with specific hybrid combinations exhibiting favorable SCA effects, provides valuable parental resources for developing combined insect resistant maize hybrids. The observed heterosis and positive breeding gains for grain yield and reduced insect damage further confirmed the effectiveness of conventional hybrid breeding in improving multiple traits simultaneously. Overall, the study confirms that the genetic improvement of tropical maize for combined resistance to C. partellus and S. zeamais is feasible, providing valuable parental lines and hybrid combinations for future breeding programs aimed at enhancing food security in sub-Saharan Africa.

Generative AI statement

The author(s) declared that generative AI was not used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Footnotes

Edited by: Salvatore Arpaia, Energy and Sustainable Economic Development (ENEA), Italy

Reviewed by: S. B. Suby, Indian Council of Agricultural Research (ICAR), India

Adamu Masari Abubakar, Ahmadu Bello University, Nigeria