Introduction

Maize is the third most vital cereal crop globally in terms of cultivation area and overall production1. Its rich grains include approximately 70% starch, 10% protein, and 8.5% dietary fiber2. This nutritional profile underscores maize grains widespread utility, ranging from edible oil and human consumption to animal feed and biofuel production3. Globally, maize cultivation spans an impressive 203.47 million hectares, yielding about 1.163 billion tons of grain, highlighting its critical role in the agricultural sector and its contribution to national economies1. However, countries like Egypt face a significant challenge, as domestic production fails to meet the demands of a burgeoning population, necessitating annual maize imports of around ten million tons1. The gap between maize consumption and production is widening, a situation aggravated by the unpredictable and changing climate. Addressing this issue, developing high-yielding maize hybrids adaptable to variable sowing dates has become increasingly important, particularly under current climate change and population growth4. Achieving higher maize production hinges on adopting optimal agricultural practices, including the strategic use of high-yielding hybrids and the precise timing of sowing5.

Sowing dates are recognized as a critical factor affecting maize yield6. Optimal sowing times improve crop profitability through higher yields and allow for better utilization of climatic resources, thus enhancing plant development and synchronizing key growth stages with favourable weather conditions7. On the other hand, late sowing can result in significant yield losses exacerbated by the unpredictability of weather conditions8. Late sowing poses risks to grain yield by reducing the sink strength and plant potential to produce and fill grains during the critical grain-filling period. This reduction is attributed to diminished source capacity, which impacts the production and transport of assimilates to the developing grains, affecting their growth and development9. Numerous studies have highlighted the adverse effects of delayed planting on maize yield, underscoring the need to develop resilient and high-yielding varieties to ensure sustainable maize production under the current climate challenges9,10,11.

In arid environments like Egypt, sowing significantly influences the agronomic performance of maize12. Optimal sowing, typically in mid-May, benefits from favourable temperatures and high levels of incident radiation, which support key developmental stages such as flowering and grain filling. In contrast, late sowing, beginning at the end of July, is characterized by declining temperatures and reduced solar radiation during critical growth periods of flowering and grain filling13. These suboptimal conditions primarily decrease number of kernels and kernel weight. This indicates the importance of selecting genotypes that enhance resilience to environmental stress and improve maize performance under late sowing conditions14. The distinct ecological challenges associated with timely and late sowing dates offer a valuable framework for assessing the genetic potential of maize inbred lines and identifying traits suited to varying sowing conditions in arid regions.

Developing superior maize hybrids demands selecting appropriate inbred lines, optimal trait combinations, and effective breeding strategies15. Exploring combining abilities and the inheritance patterns that influence key traits is essential16. The diallel analysis is a powerful method for evaluating general (GCA) and specific (SCA) combining abilities, shedding light on the roles of additive and non-additive genetic variances in trait expression17,18. Assessing the combining abilities of diverse inbred lines across different sowing dates is crucial for developing hybrids suited to various environmental conditions19. Moreover, evaluating molecular genetic distances among inbred lines using SSR markers can efficiently inform hybrid selection and breeding strategies, avoiding extensive field trials. SSR markers, known for their codominant inheritance, allele diversity, and high-throughput capabilities, are instrumental in assessing genetic diversity and relatedness within maize germplasm2,20. This study aimed to explore the genetic diversity among local and exotic maize inbred lines using SSR markers, evaluate the performance of F1 crosses under different sowing conditions, analyze GCA and SCA effects, elucidate the inheritance patterns of key traits, and investigate trait relationships under both timely and late sowing scenarios.

Results

Molecular diversity among parental inbred lines

Screening with twenty-one SSR markers revealed genetic diversity among the seven maize inbred lines. Thirteen markers showed polymorphism, facilitating the assessment of genetic variability. The allele count per locus ranged from 2 (phi96100, phi453121, phi112) to 5 (phi024, umc1014), averaging 3.31 alleles per locus. Heterozygosity (He) varied from 0.25 to 0.96, averaging 0.58, while the Polymorphic Information Content (PIC) spanned from 0.22 (phi96100) to 0.68 (umc2332), with an average of 0.50 (Table 1). Marker Index (MI) values ranged from 0.43 (phi96100) to 3.2 (umc1014), and the Resolving Power (Rp) averaged 1.95, with a range from 1.7 to 2. Genetic distances between inbred lines, as determined by SSR markers, varied from 0.47 to 0.91, averaging 0.76. Notably, the greatest genetic distance was observed between L101 and L105 (0.91), while the smallest was between L101 and L102 (0.47) (Table 2). Based on genetic distances, a dendrogram clustered the genotypes into two major groups: L102, L101, and L103, and the other containing L106, L104, L105, and L107 (Fig. 1).

Table 1 Description of the SSR markers analyzed in this study.
Full size table
Table 2 Genetic distance among the assessed parental inbred lines derived from SSR marker analysis.
Full size table
Fig. 1
figure 1

UPGMA dendrogram of seven parental maize inbreds based on genetic similarity from SSR markers

Full size image

Analysis of variance

The combined analysis of variance (ANOVA) revealed significant effects of sowing date, hybrids, and their interactions on all measured traits, highlighting the influence of environmental conditions and genetic variability on the expression of agronomic traits (Table 3). The year effect was significant only for plant height and number of kernels per row, suggesting that these traits were more sensitive to inter-annual variations in environmental factors. The ANOVA results for general combining ability (GCA) and specific combining ability (SCA) indicated significant contributions of both additive and non-additive genetic effects across all traits, emphasizing the importance of evaluating these parameters for hybrid performance. Interaction between GCA and sowing date was significant for ear length and number of kernels per row. In contrast, the interaction between SCA and sowing date was significant for several traits, reflecting the complexity of genotype-by-environment interactions under varying sowing conditions. The ratio of GCA to SCA mean squares was less than one for all traits except ear position, indicating that non-additive genetic effects predominated in determining most traits. This suggests that heterosis and specific combining ability play a critical role in the performance of the evaluated hybrids, particularly under the environmental stresses associated with late sowing date. The observed significance of GCA × sowing date and SCA × sowing date interactions highlighted the potential to select hybrids with specific adaptations to environmental conditions, offering a valuable strategy for improving maize performance in arid and variable environments.

Table 3 Mean squares and combining abilities for studied traits under timely and late sowing dates.
Full size table

Performance of F1 hybrids

All traits significantly improved under timely sowing, with increases noted across ear height, plant height, ear length, ear diameter, rows/ear, kernels/row, and grain yield by 20.4, 18.9, 12.9, 10.8, 14.9, 21.3, and 23.8%, respectively compared to late sowing (Fig. 2). Significant genotype differences were observed under both sowing conditions. Under timely sowing, plant height varied from 214.0 (L105 × L106) to 280.0 cm (L101 × L105), while under late sowing, it ranged from 168.0 cm (L101 × L107) to 220.0 cm (L101 × L105) (Fig. 3). Compared to the commercial hybrid, six hybrids, L101 × L102, L101 × L107, L102 × L106, L104 × L106, L105 × L106, and L105 × L107, were noticeably shorter than the commercial hybrid under both sowing dates. Likewise, ear height varied from 121 cm (L101 × L107) to 156 cm (L105 × L107) under timely sowing while from 95.5 cm (L101 × L107) to 124.33 cm (L105 × L107) under late sowing. Eight hybrids possessed substantially lower ear height under timely sowing than the commercial hybrid SC-10, while 16 hybrids showed the same characteristic under late sowing. Ear position exhibited variation ranging from 50.76% (L103 × L107) to 66.46% (L105 × L107) under timely sowing conditions, whereas it ranged from 50.83% (L102 × L104) to 65.11% (L105 × L107) under late sowing conditions. Ear length varied from 17.10 cm (L101 × L107) to 22.83 (L101 × L105) under timely sowing, while it ranged from 16 cm (L103 × L105) to 21.33 cm (L101 × L105) under late sowing.

Fig. 2
figure 2

Comparative boxplots of agronomic trait variation in F1 hybrids under timely and delayed sowing conditions. PH (plant height, cm); EH (ear height, cm); EP (ear position); EL (ear length, cm); ED (ear diameter, cm); NRE (number of rows/ear); NKR (number of kernels/row); and GY (grain yield).

Full size image
Fig. 3
figure 3

Comparative performance of developed 21 F1 hybrids and check hybrid (SC-10): plant height (A), ear height (B), ear position (C), and ear length (D) with Least Significant Difference (LSD) Indicators at P < 0.05 on the top of the columns.

Full size image

Similarly, ear diameter (ED) fluctuated from 4.80 cm (L101 × L107) to 5.80 cm (L102 × L104) under timely sowing and from 3.80 cm (L101 × L107) to 5.20 cm (L102 × L104) under late sowing (Fig. 4). Two crosses, L101 × L105 and L104 × L105, had substantially higher values of ear length than the check hybrid. Moreover, the number of rows/ear (NRPE) varied from 12.65 (L104 × L106) to 16.50 (L103 × L105) under timely sowing conditions and from 10.0 (L104 × L105) to 15.20 (L103 × L105) under late sowing conditions. Number of kernels/row (NKPR) varied from 31.0 (L101 × L107) to 46.50 (L104 × L107) under timely sowing conditions, while it ranged from 23.50 (L104 × L106) to 40.50 (L104 × L107) under late sowing conditions. Two crosses, L101 × L105 and L104 × L107, possessed significantly greater values than the check hybrid. Moreover, grain yield ranged from 7.30 (L104 × L106) to 11.20 (L101 × IL105) ton/ha under regular sowing date. Otherwise, it fluctuated from 6.10 (L104 × L106) to 9.09 (L101 × L105) ton/ha under late sowing. Noticeably, four hybrids, L101 × L103, L101 × L105, L104 × L105, and L104 × L107, under regular sowing date, and two hybrids, L101 × L105 and IL104 × IL107 under late sowing date, significantly out-yielded the check hybrid SC-10.

Fig. 4
figure 4

Comparative performance of developed 21 F1 hybrids and check hybrid (SC-10): Ear diameter (A), number of rows per ear (B), number of kernels per row (C), and grain yield (D) with Least Significant Difference (LSD) Indicators at P < 0.05 on the top of the columns.

Full size image

Association among studied traits and assessed hybrids

PCA differentiated late from timely sowing with minimal overlap (Fig. 5). The first two components explained 83.4% of the variance (69.1 for PC1 and 14.3% for PC2). PC1 divided maize hybrids into negative and positive sides of PC1 based on the sowing date (Fig. 5). The evaluated agronomic traits, except ear position, were related to the maize hybrids on the positive side PC1. This signified that the hybrids situated on the PC1 positive side reflected high agronomic performance, specifically H4 (L101 × L105), H22 (SC-10), H3 (L101 × L104), H8 (L102 × L104), and H18 (L104 × L107). Otherwise, the hybrids were on the extreme negative side of PC1 had lower performance under late sowing, specifically H6 (L101 × L107), H16 (L104 × L105), H19 (L105 × L106), H17 (L104 × L106). The close vectors indicate a robust positive interrelationship among their traits. Consequently, grain yield positively correlated with all assessed traits except ear position. All traits except ear position are significantly associated with the first component, ear position, which is a distinct behavior, suggesting its independence from sowing date effects. All traits except for ear position were represented by the first PC, where they had higher loading on PC1 and were located together in the right region of the biplot, distributed alongside timely sowing. Meanwhile, PC2 represented the ear position as it had high loading. It was found in the lower center of the biplot and away from both sowing dates, reflecting the relatively similar value under late and timely sowing. These results demonstrated that all traits except for ear position were suitable for distinguishing between late and timely sowing. Spearman coefficient correlations confirmed significant associations between most traits and grain yield, except for ear position under both sowing dates and all over the two conditions (Fig. 6). In general, the results suggested consistent trait correlations across both sowing dates. Density plots highlighted the impact of late sowing on traits like ear height, with varying degrees of effect observed across other characteristics. This refined results section offers a structured and concise overview of the findings, emphasizing the impact of genetic diversity and sowing dates on maize hybrid performance and trait correlations.

Fig. 5
figure 5

PCA biplot for trait variation in maize hybrids under different sowing conditions (a), bar chart detailing the percentage contribution of principal components to overall variance, and trait contribution bar charts with a dashed line demarcating significant contributions (c and d). H1 (L101×L102), H2 (L101×L103), H3 (L101×L104), H4 (L101×L105), H5 (L101×L106), H6 (L101×L107), H7 (L102×L103), H8 (L102×L104), H9 (L102×L105), H10 (L102×L106), H11 (L102×L107), H12 (L103×L104), H13 (L103×L105), H14 (L103×L106), H15 (L103×L107), H16 (L104×L105), H17 (L104×L106), H18 (L104×L107), H19 (L105×L106), H20 (L105×L107), H21 (L106×L107), H22 (SC-10). EH ear height, PH plant height, Ep ear position, ED ear diameterm, NKR number of kernels per row, EL ear length, NRE number of rows per ear, GY grain yield.

Full size image
Fig. 6
figure 6

Spearman correlation matrix for agronomic traits under optimal and late sowing conditions. PH plant height (cm); EH ear height (cm); Ep ear position; ED ear diameter (cm); EL ear length (cm); NKR number of kernel/row; NRE number of rows/ear; and GY grain yield. Timely sowing is marked in green while late sowing is in red colors.

Full size image

Heatmap and genotypic ranking

Two distinct clusters of hybrids were identified, with one group showing more similarity among its hybrids (Fig. 7). Hybrids, L103 × L104; L101 × L106; L102 × L107; L101 × L107; L104 × L105; L105 × L106; L105 × L107; L102 × L105; and L104 × L106 were grouped in one cluster where they are more similar to each other. Hybrids, L101 × L102; L101 × L103; L101 × L104; L101 × L105; L102 × L103; L102 × L104; L102 × L106; L103 × L105; L103 × L106; L103 × L107; L104 × L107; L106 × L107; and SC-10 were in the other cluster. Hybrids L104 × L107, L101 × L105, L101 × L103, L103 × L105, and SC-10 were the least affected by late sowing because they had small numbers of average rank (blue color) in the heatmap. Specific hybrids, like L104 × L107 and L101 × L105, showed resilience to late sowing effects on grain yield, as indicated by rankings (Fig. 8).

Fig. 7
figure 7

Heatmap of hierarchical clustering for maize hybrids based on tolerance indices. Blue indicates a high rank (less affected by late sowing), while red indicates a lower rank (more affected by late sowing). PH plant height (cm); EH ear height (cm); Ep ear position; ED ear diameter (cm); EL ear length (cm); NKR number of kernel/row; NRE number of rows/ear; and GY grain yield.

Full size image
Fig. 8
figure 8

Ranking of assessed maize hybrids by resilience to late sowing based on tolerance indices. Small numbers of average ranks mean less affected by late sowing (near graph bottom). H1: L101×L102, H2: L101×L103, H3: L101×L104, H4: L101×L105, H5: L101×L106, H6: L101×L107, H7: L102×L103, H8: L102×L104, H9: L102×L105, H10: L102×L106, H11: L102×L107, H12: L103×L104, H13: L103×L105, H14: L103×L106, H15: L103×L107, H16: L104×L105, H17: L104×L106, H18: L104×L107, H19:L105×L106, H20:L105×L107, H21: L106×L107, H22: SC-10.

Full size image

General combining ability effects (GCA)

Positive GCA effects were observed for most studied traits, with specific inbreds identified as superior combiners for certain traits under different sowing conditions (Table 4). The GCA effects showed that L106 had the highest significant and negative GCA effects for plant and ear height. The best desirable combiner for ear position was L103, which exhibited negative and desirable GCA effects under both sowing dates. Otherwise, the uppermost positive and significant GCA effects for ear length were expressed by L103 and L104 under timely sowing and L101 under late sowing. Likewise, L102 and L104 exhibited positive and significant GCA effects for ear diameter under both dates. Inbred lines, L102 and L105, under timely sowing and L102, under late sowing, showed positively significant GCA effects for number of rows/ear. The superior GCA effects for number of kernels/row were assigned for 103 and L104 under timely sowing and L103 under late sowing. Inbred lines L101 and L103 demonstrated the greatest positive GCA effect for grain yield under both sowing dates.

Table 4 General combining ability (GCA) of seven maize inbred lines for agronomic traits across sowing dates.
Full size table

Specific combining ability (SCA) estimates

SCA analysis indicated variation among hybrids, with specific combinations showing advantageous effects for yield and related traits. The hybrids showed diverse variations in SCAs for all traits. The crosses L101 × L102, L105 × L106 and L105 × L107 under timely sowing, and L101 × L107 and L103 × L104 under both timely and late sowing expressed significant negative effects for plant height (Table 5). The data showed that three hybrids, L101 × L107, L103 × L107, and L105 × L106, significantly negatively affected ear height. Likewise, crosses L101 × L105, L103 × L107, and L106 × L107 had significantly negative SCA effects for ear position. Otherwise, the uppermost significant and positive SCA effects for ear length were recorded by L101 × L104, L102 × L103, L103 × L106, and L104 × L107 under timely sowing, and L101 × L103, L101 × L105, L102 × L105, and L106 × L107 under both sowing dates. The highest positive effects for ear diameter were observed in timely sowing for hybrids L101 × L103, L101 × L105, L102 × L104, L102 × L106, and L106 × L107, and under late sowing for hybrids L101 × L105, L102 × L104, L103 × L104, L104 × L107, and L106 × L107. Meanwhile, the hybrids L101 × L105, L102 × L104, L103 × L105, L104 × L107, and L106 × L107 displayed the highest positive effects for number of rows/ear under both sowing dates. Likewise, the advantageous SCA effect for number of kernel/row was noticed in the hybrids L101 × L103, L101 × L105, L102 × L104, L102 × L106, L103 × L107, L104 × L107, and L106 × L107 under both sowing dates. Regarding grain yield, six and five crosses possessed highly significant positive effects under timely and late sowing conditions. The highest desirable effects were observed for crosses L101 × L105, L102 × L106, L104 × L107, and L106 × L107 under both conditions. Noticeably, there were no hybrids that simultaneously displayed favorable SCA effects for all the traits under study. However, specific hybrids exhibited beneficial effects for grain yield and advantageous SCA effects for one or more of its associated traits. The hybrid L101 × L105, L104 × L107 and L106 × L107 exhibited favorable SCA effects and were good specific combiners for most yield-attributing traits.

Table 5 Specific combining ability (SCA) effects for twenty-one F1 maize hybrids on agronomic traits under different sowing conditions.
Full size table

Discussion

Late sowing of maize results in significant yield losses exacerbated by the unpredictability of weather conditions. The adverse effects of late sowing underscore the need to develop resilient and high-yielding genotypes to ensure sustainable maize production under the current climate challenges10,11. In the present study, delayed sowing significantly decreased ear length, plant height, ear diameter, number of rows per ear, number of kernels per row, and grain yield compared to timely sowing conditions. Late sowing shortens development phases, leading to reduced photosynthetic efficiency and disruptions in the source-sink relationship coupled with decreased number and activity of growing grains8,11. This, in turn, can negatively impact yield-contributing traits and grain yield. The adverse effects of late sowing on maize growth and yield traits also were demonstrated by Caviglia et al.21, Bonelli, et al.9, and Zhiipao, et al.7. The observed significant variations among the evaluated hybrids indicate sufficient variability among the tested hybrids, which allows the possibility of selecting good hybrids under timely and late sowing dates. In this context, Omar et al.11, Han, et al.22, and Choudhury et al.23 elucidated high genetic variability for different agronomic traits under different sowing dates. The significant interaction detected between hybrids and sowing dates for grain yield and other assessed traits indicates there were dissimilar responses among the hybrids to sowing dates. This presents opportunities to identify hybrids with broad adaptation to different sowing dates. Similar findings have been reported by other authors24.

Molecular markers are valuable tools for assessing the genetic diversity present among maize inbred lines25. In the present work, SSR markers were employed to demonstrate the level of genetic diversity among parental lines. These markers provide valuable information for maize breeders in selecting high-performing hybrids with desired traits. The molecular diversity analysis in this study utilized thirteen polymorphic markers to assess genetic variation among the utilized seven parental inbred lines. While this number of markers may seem limited, they were chosen for their ability to reveal a broad range of genetic differences within the studied inbred lines. The primary goal of this analysis was to establish initial genetic diversity and complement phenotypic evaluations, providing insights into the genetic variability. Several previous studies have utilized a comparable number of molecular markers to investigate genetic diversity such as Menkir et al.26; Wang et al.27; Kapoor et al.28; Vasumathy and Alagu29; Chen et al.30; Gelotar et al.31. We acknowledge that a larger set of markers would allow for finer genetic resolution and a more comprehensive understanding of genome-wide variation. However, this analysis focused on supporting the selection of inbred lines with potential diversity rather than detailed genome mapping. The findings revealed a range of 2 to 5 alleles per locus, with an average of 3.31. This result concurred with the results proved by Menkir et al.26, and Kamara et al.18, which reported averages of 3.0 and 4.2 alleles/locus, in the same order. Otherwise, it was lower than the depicted averages of 6.8 and 7.7 by Pejic et al.32, and Mathiang et al.2, respectively. Differences in allele counts per locus across several studies may be attributed to factors such as sample size and the specific SSR markers utilized. The Polymorphic Information Content (PIC) value of 0.50 indicates the ability of the SSR loci to discern genetic disparities among the maize inbred lines, a figure comparable to or exceeding previous maize studies employing SSR markers33. This suggests that the SSR markers used in our study were informative and effective in capturing genetic diversity among the parental maize inbred lines. The umc1014, umc2332, and phi233376 loci possessed superior PIC values > 0.63. This affirms the effectiveness of these three loci in distinguishing diverse maize inbred lines in future studies. The high average genetic diversity value of 0.76 observed through the SSR analysis in this study indicates a significant amount of genetic variation among the parental inbreds. SSR markers successfully divided the parental inbred lines into two distinct clusters. Crossbreeding genetically diverse parents from different clusters has the potential to generate hybrids with complementary traits and increased heterosis18,34. Insights gleaned from cluster analysis may streamline field evaluations by reducing the number of crosses needing assessment, enabling breeders to concentrate efforts on a selected subset of crosses with greater potential for yielding desirable hybrid combinations and optimizing breeding resource allocation.

Progress in the maize breeding program depends heavily on the ability of parental lines to transmit favorable alleles to their progeny effectively. Thus, selecting parental inbred lines with favorable alleles and ensuring their transmission to the offspring is critical to the breeding process. Breeders favored positive SCA effects for most traits, except for plant height, ear height, and ear position, where negative values were preferred. The evaluated inbreds’ GCA effects exhibited positive to negative variation for all studied traits under timely and late sown, indicating the significant impact of sowing dates on the observed GCA effects, which aligned with the results of Turkey et al.35. Line L106 was determined to be a valuable combiner for decreasing plant and ear heights, which is crucial for improving tolerance to lodging under timely and late sowing dates. The favorable GCA effects demonstrated by L101 and L10 for grain yield and some of its related traits. This indicates that these inbreds possess valuable alleles that can be transmitted to their offspring to develop high-yielding maize hybrids under timely and late sowing dates36. The identification of desirable hybrids is accomplished through the use of SCA estimates. In this study, many evaluated hybrids demonstrated favorable effects of SCA for at least one trait under timely and late sown conditions. The cross combinations, L101 × L105, L104 × L107, and L106 × L107, were determined to be the superior specific combiner for breeding high yielding maize hybrids compared to the others. Subsequently, these hybrids can be employed efficiently in maize breeding programs to boost yield potential and enhance overall productivity under timely and late sowing conditions. It is worth mentioning that these hybrids were produced through crosses involving parents exhibiting either good × good or good × poor general combining abilities (GCA). This phenomenon might stem from one parent demonstrating robust combining abilities with favorable additive effects while the other parent, despite having a poor GCA, contributes to epistatic effects19,37.

Understanding the inheritance pattern of agronomic traits is crucial for breeding programs. In the present study, both additive and non-additive gene actions were detected to be proportionately significant in the inheritance of the studied agronomic traits as shown by significant GCA and SCA effects. The GCA/SCA ratio was lower than one for nearly all measured traits except for ear position. This alludes to non-additive gene action playing a substantial role in regulating the inheritance of evaluated traits. Therefore, the crossing method is highly efficient in improving these characteristics and exploiting the heterosis effect. This corroborated the findings of Turkey et al.35, Ahmed38, Makumbi et al.39, Kamara et al.40, Al-Kahtani et al.15, and Badu-Apraku et al.41 elucidated that non-additive gene action plays a decisive role in the inheritance of yield traits in maize under different sowing dates. It, however, contradicted Hefny19 findings that the additive gene action plays a crucial role in determining the inheritance of yield traits in maize under timely and late sown conditions.

Materials and methods

Plant materials and molecular characterization

This study used seven distinct local and exotic maize inbred lines acquired from various sources. The genotypes used in this study were obtained from the Agricultural Research Center, Egypt, and the International Maize and Wheat Improvement Center (CIMMYT). The source and pedigree information for these inbreds are detailed in Supplementary Table S1. These inbred lines complied with national, international, and institutional legislation and guidelines. The utilized inbred lines were chosen for their diverse genetic backgrounds, representing a wide range of genetic variability. These lines also exhibit contrasting agronomic traits, including differences in yield potential and phenological responses to varying sowing dates. This genetic and agronomic diversity provides a proper basis for evaluating their adaptability and genetic potential under the distinct environmental conditions associated with timely and late sowing. Ten seeds from each inbred line were grown in Petri dishes within germination incubators set at 25 °C with a 12-hour photoperiod and 12-hour darkness cycle. The DNA was extracted from young fresh leaves two weeks post-planting using the CTAB method, as referenced by Doyle42. A NanoDrop spectrophotometer assessed the DNA quality and quantity. Then, the DNA was amplified using 25 SSR primer pairs, detailed in Table S2 (sourced from the Maize-GDB database), through polymerase chain reaction (PCR) in a ten µL mixture. This mixture contained one µL of 20 ng/µL genomic DNA, 1 unit of Taq DNA polymerase, two mM-MgCl2, 0.2 mM-dNTPs, and 0.5 µM of each primer. PCR conditions were set as follows: initial denaturation at 94 °C for 2-min, followed by 35 cycles of denaturation at 94 °C for 30-s, annealing at 55 °C for 30-s, and extension at 72 °C for 30-s, with a final elongation step at 72 °C for 3-min. The amplification products were analyzed on 1.5% agarose gels, and the presence or absence of bands for each SSR marker in a binary data matrix was recorded. The dendrogram was generated using UPGMA within MVSP version 3.1. The allele number, gene diversity, and polymorphic information content (PIC) were calculated for all markers with PowerMarker version 3.25 and to determine genetic distances using the PAST program.

Generation of F1 hybrids and field evaluation

Seeds of seven inbred lines were sown in the summer of 2019 on three different dates (15th, 22nd, and 29th May) to synchronize flowering times and produce sufficient F1 hybrid seeds. The inbreds were sown in rows 6 m long and 0.70 m apart. Half diallel mating design was used to create 21 possible F1 hybrids. The developed F1 hybrids and the commercial check hybrid SC-10 were evaluated in field trials during the 2020 and 2021 growing seasons under both timely (20th May) and late sowing (31st July) conditions at the Kafrelsheikh University Experimental Farm, Egypt. The site is characterized by hot, dry weather with no rainfall during the maize season, as depicted in Fig. 9. Soil characteristics for both seasons are presented in Table S3. A Randomized Complete Block Design in three replicates was used. Each plot consisted of three rows, each with 5 m in length and spaced 0.65 m apart. Within each row, seeds were sown in hills at 0.25-meter intervals, resulting in an average density of 61,540 plants per hectare. Fertilizers were added at 116 kg K2O, 76 kg P2O5, and 290 kg N per hectare.

Fig. 9
figure 9

Daily minimum and maximum temperatures, alongside solar radiation levels and relative humidity across two growing seasons.

Full size image

Data collection

Plant height was recorded from the soil to the top of the first tassel branch and ear height from the soil to the base of the uppermost ear, both in centimeters. Ear position was calculated using the formula: Ear position = (Ear height/Plant height) × 100. At harvest, ten ears were randomly collected from each plot to determine the number of kernels per row, ear length, ear diameter, and number of rows per ear. The grain yield was estimated from the harvested plot, and the weight of the shelled grain (adjusted to 15.5% grain moisture content) was converted to tons per hectare.

Statistical analysis

The data were analyzed using ANOVA with R statistical software version 4.2.2, applying the least significant difference (LSD) test at P < 0.05 to assess mean differences. Combining ability analysis followed Griffing’s method 4 model 116. Spearman coefficient correlation matrix was built with the GGally package and generated a heatmap with the pheatmap package. Cubic Cluster Criterion and the fuzzy C-means algorithm were used to assess data clustering, with the Ward method eventually selected for its superior agglomerative coefficient43. Principle component analysis was conducted using the prcomp function, and biplots were produced with fviz_pca_biplot in the factoextra package.

Conclusions

SSR markers revealed substantial genetic diversity among the assessed parental inbred lines. The findings from the molecular diversity analysis hold promise for maize breeding initiatives, facilitating the identification of distinct inbreds suitable for crossbreeding. The inbreds L101 and L103 emerged as superior combiners for enhancing ear characteristics and grain yield, while line 106 exhibited potential as a strong combiner for developing short-stature genotypes. The hybrid combinations such as L101 × L105, L104 × L107, and L106 × L107 demonstrated favorable specific combining ability effects, particularly in grain yield and associated traits across varying sowing conditions. Non-additive effects were predominant for all traits studied except ear position. The observed positive associations between certain traits such as ear length, plant height, number of kernels per row, and number of rows per ear with grain yield provide preliminary insights into their potential as indicators for indirect selection. However, further study validation involving more extensive genotypes is necessary to confirm their reliability for indirect selection in breeding programs.