Combining Ability of Maize Landraces for Yield and Basic Chemical Composition of Grain

Abstract

The launch of a successful quality-oriented breeding program requires both mining the residual diversity in grain quality parameters contained in the elite, high-yielding breeding material with good agronomic performance and introgression of new germplasm, such as local landraces, with a high level of targeted quality parameters per se. This study analyzed the combining abilities of 31 maize landraces and two divergent inbred lines–testers (ZPL217 and ZPL-255/75-5) regarding the yield and protein, starch, and lipid content, assessed by Near Infrared Reflectance (NIR) spectroscopy as a fast, non-destructive, and cost-effective method. The general combining ability (GCA) defines the average behavior of genotype in hybrid combination, resulting from additive gene action, so positive GCA values of landraces AN13 and AN197 for protein, AN632 for lipids, and AN594 for starch content indicate that they can be valuable sources of the mentioned properties in quality-oriented maize breeding programs. The obtained correlation between starch content and protein and yield (−0.948 **; 0.587 **) pointed out that an increase in the protein content during breeding will be accompanied by a decrease in the starch content and yield. The specific combining ability (SCA) of the testers used, suggests their possible application in establishing and improving quality breeding programs’ initial material.

1. Introduction

Maize (Zea mays L.) is a highly polymorphic species that encompasses a wide cultivation area. High adaptability and yield potential, along with its diverse applications, contribute to its status as a dominant crop in various countries [1]. Commercial maize breeding programs are often based on a narrow genetic foundation, utilizing elite inbred lines as parental components in various hybrids, aiming to achieve optimal agronomic performance in a short time frame. In the race for profit, the nutritional value of the grain is often overlooked, particularly regarding its use in human nutrition [2]. Compared with commercial hybrids, landraces are usually richer in protein and lipid content as well as in molecules with a role as antioxidants [3].

Recent trends underscore the importance of healthy diets enriched with nutritional values for both human and animal nutrition [4]. This trend directs plant breeding toward the development of varieties with improved grain quality. The importance of enhancing grain quality is recognized in the context of its utility in industrial processing, human nutrition, and livestock feeding. An international network (EVA maize) involving European genebanks, research institutes, and breeders is currently studying traditional European landraces to explore their variability and their possible exploitation as pre-breeding materials [5].

Various quantitative genetic approaches have been applied so far to genetically improve grain quality [6]. Successful genetic approaches include recurrent selection, incorporation of exotic germplasm, and the use of mutants [7,8,9]. Identifying important traits and the genes that define them is crucial for the rapid advancement of selection and the creation of hybrids rich in nutritional value. Research by Vaz Patto et al. [10] on local maize landraces (open-pollinated variety) in Portugal showed that higher quality populations (with increased protein content, lower amylose content, and lower viscosity) are utilized to produce flour for traditional dishes [11].

The standard quality maize grain contains approximately 90 g kg⁻¹ protein, 40 g kg⁻¹ oil, 730 g kg⁻¹ starch, and 140 g kg⁻¹ other components, primarily fiber [12]. The distribution of nutrients within the grain is such that oil is predominantly found in the germ, while starch and proteins are concentrated in the endosperm. Analyzing the composition of maize grain, Cook et al. [13] observed variations in protein content (ranging from 12.3% to 15.3%), oil (from 3.5% to 5.5%), and starch (from 62.3% to 69.6%), indicating a need for a more detailed understanding of the genetic differences among selected genotypes for these three traits.

Redaelli et al. [3] highlighted the importance of germplasm characterization in promoting the recovery and valorization of local biodiversity and also pointed out the important role of ex situ conservation in maintaining the allelic richness of crop species for future breeding programs. When selecting initial breeding material, special attention is paid to adaptation to local growing conditions, and a high level of desirable traits allows for more efficient breeding of superior lines [14]. An initial population with a high value for the selected trait enables fewer cycles of selection to achieve the desired trait level within that population. Genetic diversity estimates help to structure germplasm defining, for example, heterotic pools, and provide useful information to select opposite parental lines for new breeding populations [15]. The value of the initial population represents a crucial factor in the selection process, whose final product in the form of hybrids not only derives from inherent traits but also depends on the ability to combine with other populations, groups of populations, or lines. Opposite parents during hybridization contribute to the complex dynamics of genetic material and the final expression of phenotypic characteristics [16]. Furthermore, a good combinator is not necessarily the final step in selection, as besides good combining abilities, the genotype must also contain other desirable characteristics [17].

The concept of combining ability (CA) reflects a parent’s potential to produce superior offspring when crossed with another parent and is crucial for the development of high-yielding hybrids. Combining abilities fall into two categories: general combining ability (GCA) and specific combining ability (SCA) [18]. Information on combining ability and heterosis is imperative in a genetic improvement program to develop hybrids or synthetic varieties [19]. The GCA defines the average behavior of a line in its hybrid combinations, and SCA evaluates hybrid combinations with respect to the average behavior of the lines. GCA results from additive gene action, while SCA depends on dominance, predominate, and epistasis [20]. The development of high-yielding hybrids necessitates the careful selection of parents based on their combining ability and genetic structure [21]. A positive aspect of Line × Tester analysis [22,23] is the ability to assess inbred lines and populations crossed with good GCA testers, with or without the parents themselves included in the trial, thus providing the opportunity to test a greater volume at once. Although combining abilities have predominantly been used for GCA/SCA calculations for yield, such practices have also found their place in non-yield purposes, such as nutritive compounds [24,25,26,27].

Modern analytical methods for measuring the nutritional values of maize seeds bring significant changes by eliminating the need for large quantities of seeds and destructive approaches. Notably, Near Infrared Reflectance (NIR) spectroscopy stands out, as its technique allows for the measurement of starch, protein, and lipid content in whole maize grains without disrupting their structure [28]. Numerous studies highlight the significant predictive power of NIR spectroscopy, achieved through reflective mode and analysis of grain composition based on calibration curves [29,30].

This study aims to investigate the general (GCA) and specific combining abilities (SCA) of 31 maize landraces crossed with two testers, ZPL217 and ZPL-255/75-5, each representing different heterotic backgrounds while focusing on the basic chemical composition of grain, specifically protein, starch, and lipids, as well as yield. The results of these studies can be utilized for selecting initial material in the development of maize hybrids with enhanced basic chemical composition of the grain.

2. Materials and Methods

2.1. Plant Material

Thirty-one local maize landraces were used in this study, being previously selected through a stratified process within pre-breeding activities as a suitable initial material for bordering Maize Research Institute Zemun Polje (MRIZP), Belgrade - Republic of Serbia commercial breeding germplasm pools [1]. Landraces are a part of the MRIZP Gene Bank collection. The selected landraces (

Table 1. The per se values of the yield and basic chemical composition of grain for exanimated landraces.

No.	Accession Number	Domestic Name	Country of Origin	Protein (%)	Starch (%)	Lipid (%)	Yield 1 (t ha−1)
1	1890	Domaći kukuruz	HR	11.606	64.703	3.537	4.27
2	2144	Domaći kukuruz	HR	11.102	66.627	3.399	3.41
3	467	Žuti tvrdunac	BA	12.075	63.818	3.895	3.68
4	1960	Domaći crveni	BA	10.775	66.983	3.386	4.84
5	13	Žuti jarik	ME	12.166	64.992	4.207	3.14
6	1895	Domaći kukuruz	HR	11.311	65.590	3.378	5.14
7	1267	Belo seme	ME	11.181	65.363	4.114	4.53
8	1276	Selarsko seme	ME	11.365	65.387	3.543	4.97
9	773	Žuti osmak	RS	10.007	69.449	3.818	4.23
10	1384	Domaći kukuruz	BA	10.841	67.390	3.315	3.99
11	594	Crvena pčenka	MK	11.199	68.275	2.642	3.04
12	871	Brzica skorovno seme	MK	11.822	64.903	3.629	3.55
13	846	Žuti polutvrdunac	BA	12.470	64.530	3.436	3.35
14	642	Žuti osmak	RS	11.703	64.711	3.643	4.29
15	1798	Žuti tvrdunac—vragati	RS	10.899	66.419	3.706	5.36
16	2033	Domaći kukuruz	HR	11.054	66.572	3.359	5.47
17	2006	Domaći tvrdunac	BA	12.068	63.660	3.621	5.48
18	2036	Bosanac	BA	11.162	65.386	3.622	5.98
19	1945	Domaći tvrdunac—žuti	BA	11.268	65.817	3.235	4.45
20	1346	Srednje belo seme	ME	9.4790	68.528	4.329	5.88
21	1665	Žuti zuban	MK	11.153	66.051	3.322	5.79
22	1509	Avguštana	SI	10.489	66.596	4.060	6.92
23	1534	Domaći beli	RS	11.564	65.751	3.647	5.44
24	2249	Muratovača	BA	10.434	68.076	3.271	7.74
25	2047	Domaći kukuruz	HR	11.244	65.594	3.586	4.69
26	1450	Domaći kukuruz	HR	10.588	67.285	3.516	5.73
27	632	Žuti zuban	RS	10.224	67.637	4.439	5.34
28	877	Čađo	BA	11.590	66.204	2.853	5.36
29	197	Brdska zobanka	SI	11.622	64.344	3.924	4.65
30	288	Žuti zuban	RS	11.494	65.198	3.468	5.23
31	1569	Zobačka	SI	10.602	66.993	4.117	5.92

HR—Croatia, MK—North Macedonia, RS—Serbia, ME—Montenegro, BA—Bosnia and Herzegovina, SI—Slovenia. ¹ The more details about the landraces can be found in the manuscripts of Popović et al. [1,31].

) were crossed in technical isolation with two MRIZP divergent testers, ZPL217 and ZPL-255/75-5, which are part of different heterotic groups (Iowa Dent and Lancaster). Those are two widely used commercial testers in commercial MRIZP breeding programs and are components of a large number of realized hybrids (

Table 2. Mean values of basic nutritional parameters of maize inbred testers.

Testers	Protein (%)	Starch (%)	Lipid (%)	Yield 1 (t ha−1)
ZPL217	12.29	65.07	2.68	3.20
ZPL-255/75-5	11.64	66.41	3.13	2.96

¹ The data for yield were already published in [31].

). A total of 62 test cross hybrids (maize landrace × tester) were developed for the study.

2.2. Field Experiment

This experiment was conducted during the 2022 and 2023 growing seasons, with two replications at four locations in Serbia: Zemun Polje, Pančevo, Sremska Mitrovica, and Bečej (

Table 3. Geographic coordinates and elevation of experimental locations.

Location	Latitude	Longitude	Altitude (m)
Zemun Polje	N44°51′	E20°18′	73
Pančevo	N44°88′	E20°77′	78
Sremska Mitrovica	N45°02′	E19°64′	88
Bečej	N45°69′	E19°91′	80

). A total of 62 test cross hybrids (maize landrace × tester) have been tested. The elementary plots were 5 m long, with an inter-row spacing of 75 cm, covering a total area of 7.5 m². Standard agronomic practices were followed at all locations, including soil preparation, fertilization, and pest management.

The experiment was set up using a partially balanced incomplete block design (PBIB) [32]. The locations used in the experiment belong to the Pannonian 3 (PAN3) zone as defined by European Environmental Stratification [33]. The last meter of each row was designated for chain pollination to produce seed needed for NIR analysis [34], while the final 6 m² were reserved for grain yield assessment. Seeding and harvesting were performed using Wintersteiger equipment, Ried im Innkreis (Innviertel region), Austria (Dual Dynamic Disc Seeder and Split Plot Harvester) at a density of 66,667 plants ha^–1, with yield adjusted to 14% moisture content. Main soil characteristics and standard cropping practices applied are presented in

Table 4. Main soil characteristics and cropping practices applied.

Soil Properties	Zemun Polje	Pančevo	Sremska Mitrovica	Bečej
pH in H2O	7.43	7.59	7.79	7.86
pH in KCl	6.96	7.02	7.08	7.08
Total N (%)	0.24	0.21	0.28	0.31
P2O5 (mg/100 g)	14.53	12.20	18.89	22.15
K2O (mg/100 g)	30.26	23.95	31.92	33.12
Organic matter (%)	3.89	3.68	4.12	4.43
Soil Type	Slightly calcareous chernozem	Silty chernozem	Calcareous chernozem	Humus-accumulative chernozem
Humus-accumulative horizon	Profile: Ah-AhC-C
Tillage	Deep tillage (at 30 cm in the autumn) and pre-sowing soil preparation (in the spring)
Fertilization	120 kg ha−1 of NPK (10:52:0) in the autumn 280–300 kg ha−1 of urea in the spring before sowing
Weed control	Pre-emergence application with terbuthylazine and S-metolachlor Post-emergence application with nicosulfuron and mesotrion

.

Although the experiment was initially planned at four locations, the Sremska Mitrovica site from the 2023 growing season was excluded from the analysis due to severe storm damage. Consequently, seven external environments were considered for statistical analysis. The combined ANOVA included the effects of the environment, as well as the interactions of GCA and SCA with the environment.

2.3. Chemical Composition Analysis

The analysis of the basic chemical composition of the grain was conducted using the Infraneo spectrometer from Chopin Technologies, KPM Analytics (Manufacturing headquarters for Chopin Technologies), Paris, France. This spectrometer employs near infrared radiation in the range of 710 to 1100 nm, specifically designed for transmission analysis of solid products. With a high monochromator resolution of 0.1 nm, it effectively measures absorption through greater thicknesses, making it suitable for heterogeneous samples such as maize grains. This method provides rapid determination of protein, starch, and lipid content in maize grains. According to Paulsen et al. [35], NIR spectroscopy provides rapid analysis and measures starch content, where the content indicates the amount of starch present in the grain. Calibration of NIR spectroscopy is performed using a calibration set consisting of samples with known chemical composition. Reference values of the calibration set generated by standard laboratory methods, such as the Kjeldahl method for proteins, the polarimetric method for starch, and Soxhlet extraction for lipids, were used to develop calibration models by applying the Partial Least Squares Regression (PLSR) method, enabling the prediction of composition based on spectral data. The bias of the calibration curve is expressed as the difference between the arithmetic means of the laboratory and NIR measurements. The slope of the linear regression and the bias calibration curve were tested by t-test. Calibration is conducted by correlating sample spectra with laboratory-determined values, after which the model is validated using an independent set of samples [29]. Validation includes statistical parameters such as the coefficient of determination (R²), root mean square error of calibration (RMSEC), root mean square error of prediction (RMSEP), and bias, which determine the accuracy and precision of the model.

2.4. Statistical Analysis

General and specific combining abilities (GCA/SCA) were calculated using the AGD-R software (Analysis of Genetic Designs in R for Windows) Version 5.1 (3 August 2022) [36]. The “line by tester analysis” design was applied. The effects of maize landraces (GCA for landraces), testers (GCA for testers), and their interaction (SCA) were treated as fixed, whereas the effects of replications, environment, and their interactions with GCA and SCA were considered random.

The applied statistical model used to perform the Line × Tester analysis within the multi-environment RCBD framework is as follows:

$$Y_{ijk}=\mu +E_d+{REP}_k\left(E_d\right)+l_i+t_j+l_i\times t_j+E_d\times l_i+E_d\times t_j+E_d\times l_i\times t_j+e_{ijk}$$

where

$Y_{ijk}$ is the observed value;

$\mu$ is the general mean;

$E_d$ is the environmental effect $\left(d=1,2,\dots ,s\right)$;

${REP}_k\left(E_d\right)$ is the effect of replicate $k$ nested in environment $d\left(k=1,2,\dots ,r\right)$;

$l_i$ is the line effect $\left(i=1,2,\dots ,m\right)$;

$t_j$ is the tester effect $\left(j=1,2,\dots ,f\right)$;

$e_{ijk}$ is the residual.

3. Results

The analysis of general combining ability (GCA) and specific combining ability (SCA) revealed promising maize landraces for specific nutritional quality traits.

3.1. General Combining Ability (GCA)

For protein content, two GCA combiners, Accession Number (AN) AN13 and AN197, were significantly positive at the 0.05 probability level, with values of 0.54 * (9.95%) and 0.47 * (9.88%), respectively. For lipids, AN632 was the only highly significant positive GCA combiner among the 31 landraces, with a value of 0.28 ** (4.38%). The only significant positive GCA value for starch was discovered for the landrace AN594, with a value of 1.25 * and a starch content of 70.39%. Landrace AN13 exhibited the lowest GCA value of −1.27 ** (67.87%), indicating antagonistic effects and suggesting potential breeding directions for these traits. For grain yield, AN1267 was identified as the best GCA combiner, with a value of 0.87 * and an average yield of 8.20 t ha^–1 (

Table 5. Mean performance, GCA values, and rank of maize landraces for protein, starch, lipid content, and grain yield.

Accession Number	Protein			Starch			Lipid			Grain Yield
	Line Mean	GCA	Rank	Line Mean	GCA	Rank	Line Mean	GCA	Rank	Line Mean	GCA	Rank
AN 1895	9.38	−0.02	15	69.26	0.12	15	4.03	−0.08	24	7.43	0.09	12
AN 1267	9.27	−0.14	21	69.15	0.01	19	4.31	0.20	2	8.20	0.87 *	1
AN1276	9.49	0.08	10	68.70	−0.44	25	4.22	0.11	5	7.66	0.32	8
AN 773	9.25	−0.16	25	69.54	0.39	7	4.10	0.00	16	7.25	−0.09	21
AN1384	9.27	−0.14	23	69.55	0.41	6	3.92	−0.19	30	7.35	0.01	14
AN 594	9.05	−0.36	31	70.39	1.25 *	1	3.92	−0.19	29	7.53	0.19	10
AN 871	9.28	−0.13	20	69.48	0.34	8	4.13	0.03	12	7.32	−0.02	15
AN 846	9.52	0.12	9	68.64	−0.50	26	4.13	0.02	13	6.99	−0.35	27
AN 642	9.61	0.20	6	68.84	−0.30	24	4.05	−0.05	22	7.29	−0.05	19
AN 1798	9.60	0.19	7	68.88	−0.26	22	4.03	−0.08	23	7.71	0.37	7
AN 2033	9.31	−0.10	18	69.36	0.22	13	4.02	−0.08	25	7.14	−0.20	24
AN 2006	9.62	0.21	5	68.63	−0.51	27	4.19	0.09	8	6.95	−0.39	28
AN 2036	9.42	0.02	12	69.08	−0.06	20	4.17	0.06	10	7.94	0.60	3
AN 1945	9.26	−0.14	24	69.66	0.52	3	3.98	−0.13	28	7.40	0.06	13
AN 1346	9.20	−0.21	27	69.45	0.30	10	4.27	0.16	3	7.95	0.62	2
AN 1665	9.32	−0.09	17	69.47	0.33	9	4.07	−0.03	18	7.64	0.30	9
AN 1890	9.24	−0.17	26	69.43	0.29	11	4.09	−0.02	17	6.08	−1.26 **	31
AN 1509	9.11	−0.29	30	69.57	0.43	5	4.23	0.13	4	7.77	0.43	5
AN 1534	9.78	0.38	3	68.32	−0.82	28	4.06	−0.04	21	7.01	−0.33	26
AN 2249	9.15	−0.26	28	69.61	0.47	4	4.01	−0.10	26	7.30	−0.04	16
AN 2047	9.59	0.18	8	68.87	−0.27	23	4.07	−0.04	19	7.25	−0.08	20
AN 1450	9.41	0.01	14	69.19	0.05	17	4.14	0.03	11	7.17	−0.16	22
AN 632	9.27	−0.14	22	69.16	0.02	18	4.38	0.28 *	1	7.08	−0.26	25
AN 877	9.14	−0.27	29	70.09	0.95	2	3.88	−0.23	31	7.73	0.39	6
AN 197	9.88	0.47 *	2	68.08	−1.07	30	4.11	0.01	15	7.82	0.48	4
AN 288	9.42	0.02	13	69.07	−0.07	21	4.12	0.01	14	7.29	−0.05	18
AN 2144	9.43	0.03	11	69.23	0.09	16	4.00	−0.11	27	7.48	0.14	11
AN 1569	9.34	−0.07	16	69.27	0.13	14	4.21	0.10	6	7.30	−0.04	17
AN 467	9.76	0.36	4	68.15	−0.99	29	4.19	0.09	9	6.76	−0.58	29
AN 1960	9.28	−0.13	19	69.38	0.24	12	4.06	−0.04	20	7.14	−0.20	23
AN 13	9.95	0.54 *	1	67.87	−1.27 *	31	4.20	0.09	7	6.56	−0.78	30
Grand Mean	9.41			69.14			4.11			7.34
Standard Error	0.22			0.55			0.11			0.42

* p < 0.05; ** p < 0.01.

).

Tester ZPL217 exhibited a positive GCA for protein 0.03 (9.44%) and starch content 0.04 (69.18%). In contrast, tester ZPL-255/75-5 was identified as a positive GCA combiner for lipids 0.19 (4.30%) but exhibited negative GCA values for starch and proteins. The tester L217, belonging to the Iowa Dent heterotic pool, demonstrated a positive GCA for grain yield (0.35) (7.69 t ha⁻¹) compared to tester ZPL-255/75-5, which has a Lancaster heterotic background (

Table 6. General combining ability (GCA) values, mean performance, and basic chemical composition and grain yield for testers.

Source	Basic Chemical Composition of Grain						Grain Yield (t ha−1)
	Protein (%)		Starch (%)		Lipid (%)
Tester	1 T	2 T	1 T	2 T	2 T	1 T	1 T	2 T
Tester Mean	9.44	9.38	69.18	69.10	4.30	3.91	7.69	6.99
Grand Mean	9.41		69.14		4.11		7.34
GCA Value	0.03	−0.03	0.04	−0.04	0.19	−0.19	0.35	−0.35
Standard Error	0.03		0.04		0.19		0.35

p = 0.3559 > 0.05; 1 T—ZPL217; 2 T—ZPL-255/75-5.

).

Highly significant effects on protein, starch, lipid, and yield were observed for the factors of location, genotype, and the interaction between the environment and tester. Specifically, the effect of the tester was highly significant only for lipid and yield, whereas no significant effect was found for protein and starch content. Furthermore, it was noted that all factors had a highly significant effect on lipid content, while the tester factor was statistically non-significant for starch. The interaction Site × Line × Tester had a significant effect on yield variation but no effect on the variation in protein, starch, or lipid content. Other interactions, such as Site × Genotypes and Site × Line, showed significant effects on all traits, but specific interactions were not significant for all traits.

The relative importance of additive (GCA) and non-additive (SCA) gene action indicated a greater contribution of GCA for all investigated traits. The proportion of variance explained by GCA for protein, starch, and lipid content was consistently high, ranging from 82.24% to 84.27%, suggesting that these traits are primarily governed by additive genetic effects. In contrast, the proportion of GCA for grain yield was lower (64.89%), with a relatively higher contribution of SCA (35.11%), indicating a greater influence of non-additive genetic effects on this trait. This suggests that grain yield is more affected by dominance and epistatic interactions, making hybrid performance less predictable based solely on parental GCA values (

Table 7. Line × Tester analysis across environments for grain chemical composition and grain yield.

Source	df	Sum Sq
		Protein	Starch	Lipid	Grain Yield
ENVIRONMENT (E)	6	197.12 **	383.04 **	35.07 **	2504.39 **
REP (E)	7	37.39 **	86.35 **	4.05 **	5.21 **
GENOTYPE (G)	61	52.27 **	311.01 **	45.82 **	350.13 **
LINE (L)	30	43.28 **	261.12 **	11.25 **	155.71 **
TESTER (T)	1	0.70	1.14	31.91**	108.21 **
L × T	30	8.29	48.75 **	2.66 **	86.22 **
E × G	366	95.66 *	408.10 **	18.59 **	615.30
E × L	180	41.28	206.76 **	10.98 **	206.44
E × T	6	17.30 **	31.23 **	2.47 **	65.60 **
E × L × T	180	37.09	170.11	5.14	343.26*
Residuals	427	91.52	357.07	15.38	631.17
% GCA SS		83.93	84.27	82.24	64.89
% SCA SS		16.07	15.73	17.76	35.11

* p < 0.05; ** p < 0.01.

).

3.2. Specific Combining Ability (SCA)

SCA values indicated that the test hybrid AN877 × ZPL-255/75-5 was the most significantly positive combination for protein content, with an SCA value of 0.19 * and a protein content of 9.30% (

Table 8. SCA values for protein, starch, lipids, and grain yield in Line × Tester combinations.

Rank	SCA Protein				SCA Starch				SCA Lipid				SCA Grain Yield
	Line	T	L × T (%)	SCA	Line	T	L × T (%)	SCA	Line	T	L × T (%)	SCA	Line	T	L × T (t ha−1)	SCA
1	AN 877	2 T	9.30	0.19 *	AN 877	1 T	70.73	0.60 *	AN 1890	1 T	4.04	0.14 *	AN 13	2 T	7.05	0.84 **
2	AN2036	1 T	9.64	0.18	AN 2036	2 T	69.56	0.52 *	AN 13	1 T	4.11	0.11 *	AN 1569	2 T	7.53	0.58
3	AN 288	2 T	9.57	0.17	AN 632	2 T	69.57	0.44	AN 846	1 T	4.04	0.10	AN 2033	1 T	8.01	0.52
4	AN 2249	2 T	9.29	0.17	AN 1945	2 T	70.03	0.41	AN 2249	2 T	4.30	0.10	AN 594	1 T	8.34	0.46
5	AN 1945	1 T	9.44	0.15	AN 1534	2 T	68.58	0.29	AN 877	2 T	4.15	0.08	AN 1665	2 T	7.73	0.45
6	AN 197	2 T	9.97	0.12	AN 1384	1 T	69.86	0.27	AN 2033	2 T	4.29	0.07	AN 1450	1 T	7.94	0.41
7	AN 13	1 T	10.10	0.12	AN 288	1 T	69.37	0.26	AN 467	1 T	4.07	0.07	AN 1276	1 T	8.41	0.40
8	AN 1450	1 T	9.55	0.11	AN 197	1 T	68.36	0.25	AN 632	1 T	4.26	0.07	AN 642	1 T	7.98	0.34
9	AN 1895	2 T	9.47	0.11	AN 2006	1 T	68.91	0.23	AN 1384	2 T	4.17	0.06	AN 467	2 T	6.74	0.33
10	AN 1384	2 T	9.35	0.11	AN 1895	1 T	69.53	0.23	AN 1895	2 T	4.27	0.05	AN 1945	2 T	7.37	0.32
11	AN 1346	1 T	9.33	0.10	AN 642	1 T	69.10	0.22	AN 642	2 T	4.29	0.05	AN 2047	2 T	7.20	0.30
12	AN 1569	1 T	9.47	0.10	AN 1267	2 T	69.31	0.19	AN 1534	1 T	3.92	0.05	AN 2144	1 T	8.13	0.30
13	AN 1267	1 T	9.40	0.10	AN 13	2 T	68.02	0.18	AN 1960	2 T	4.30	0.05	AN 1798	2 T	7.63	0.27
14	AN 632	1 T	9.39	0.09	AN 1569	2 T	69.40	0.17	AN 2047	2 T	4.30	0.04	AN 197	1 T	8.44	0.26
15	AN 871	2 T	9.33	0.08	AN 2047	2 T	68.99	0.16	AN 1346	2 T	4.50	0.04	AN 1267	1 T	8.80	0.24
16	AN 1534	1 T	9.87	0.06	AN 871	1 T	69.67	0.15	AN 288	1 T	3.96	0.04	AN 1890	1 T	6.67	0.24
17	AN 2006	2 T	9.65	0.06	AN 773	1 T	69.72	0.15	AN 1509	1 T	4.07	0.03	AN 632	1 T	7.65	0.22
18	AN 2047	1 T	9.67	0.06	AN 2249	1 T	69.79	0.14	AN 1945	2 T	4.20	0.03	AN 1895	2 T	7.29	0.21
19	AN 846	2 T	9.55	0.06	AN 1346	2 T	69.54	0.13	AN 1798	2 T	4.25	0.03	AN 1509	2 T	7.60	0.18
20	AN 594	2 T	9.07	0.05	AN 1798	2 T	68.97	0.13	AN 594	2 T	4.14	0.03	AN 877	1 T	8.25	0.17
21	AN 642	2 T	9.64	0.05	AN 1450	2 T	69.27	0.12	AN 2036	2 T	4.38	0.02	AN 846	2 T	6.80	0.16
22	AN 1798	1 T	9.67	0.05	AN 2033	1 T	69.51	0.11	AN 871	1 T	3.96	0.02	AN 1346	1 T	8.45	0.14
23	AN 1665	2 T	9.34	0.05	AN 846	1 T	68.79	0.11	AN 773	1 T	3.93	0.02	AN 1384	2 T	7.14	0.14
24	AN 1960	1 T	9.35	0.04	AN 2144	2 T	69.29	0.10	AN 1450	1 T	3.96	0.01	AN 773	2 T	7.00	0.10
25	AN 1276	1 T	9.55	0.04	AN 594	1 T	70.53	0.10	AN 2006	1 T	4.01	0.01	AN 2249	1 T	7.74	0.08
26	AN 1890	2 T	9.25	0.04	AN 1665	1 T	69.58	0.07	AN 1569	2 T	4.41	0.01	AN 1534	1 T	7.45	0.08
27	AN 2144	1 T	9.49	0.03	AN 1890	2 T	69.44	0.05	AN 197	2 T	4.31	0.00	AN 1960	1 T	7.56	0.07
28	AN 1509	1 T	9.16	0.02	AN 1276	2 T	68.67	0.01	AN 1267	1 T	4.12	0.00	AN 871	2 T	7.01	0.04
29	AN 467	2 T	9.76	0.02	AN 1960	2 T	69.35	0.01	AN 2144	2 T	4.19	0.00	AN 288	2 T	6.97	0.04
30	AN 773	1 T	9.29	0.01	AN 1509	2 T	69.54	0.00	AN 1276	1 T	4.03	0.00	AN 2006	1 T	7.34	0.03
31	AN 2033	2 T	9.29	0.01	AN 467	2 T	68.12	0.00	AN 1665	1 T	3.88	0.00	AN 2036	1 T	8.30	0.01
32	AN 2033	1 T	9.33	−0.01	AN 467	1 T	69.60	0.00	AN 1665	2 T	4.26	0.00	AN 2036	2 T	7.58	−0.01
33	AN 773	2 T	9.21	−0.01	AN 1509	1 T	68.19	0.00	AN 1276	2 T	4.41	0.00	AN 2006	2 T	6.57	−0.03
34	AN 467	1 T	9.77	−0.02	AN 1960	1 T	69.41	−0.01	AN 2144	1 T	3.81	0.00	AN 288	1 T	7.61	−0.04
35	AN 1509	2 T	9.06	−0.02	AN 1276	1 T	68.72	−0.01	AN 1267	2 T	4.49	0.00	AN 871	1 T	7.63	−0.04
36	AN 2144	2 T	9.37	−0.03	AN 1890	1 T	69.42	−0.05	AN 197	1 T	3.92	0.00	AN 1960	2 T	6.72	−0.07
37	AN 1890	1 T	9.23	−0.04	AN 1665	2 T	69.36	−0.07	AN 1569	1 T	4.01	−0.01	AN 1534	2 T	6.58	−0.08
38	AN 1276	2 T	9.42	−0.04	AN 594	2 T	70.25	−0.10	AN 2006	2 T	4.38	−0.01	AN 2249	2 T	6.87	−0.08
39	AN 1960	2 T	9.21	−0.04	AN 2144	1 T	69.16	−0.10	AN 1450	2 T	4.32	−0.01	AN 773	1 T	7.50	−0.10
40	AN 1665	1 T	9.30	−0.05	AN 846	2 T	68.50	−0.11	AN 773	2 T	4.28	−0.02	AN 1384	1 T	7.56	−0.14
41	AN 1798	2 T	9.52	−0.05	AN 2033	2 T	69.21	−0.11	AN 871	2 T	4.31	−0.02	AN 1346	2 T	7.46	−0.14
42	AN 642	1 T	9.59	−0.05	AN 1450	1 T	69.11	−0.12	AN 2036	1 T	3.95	−0.02	AN 846	1 T	7.18	−0.16
43	AN 594	1 T	9.02	−0.05	AN 1798	1 T	68.79	−0.13	AN 594	1 T	3.70	−0.03	AN 877	2 T	7.20	−0.17
44	AN 846	1 T	9.49	−0.06	AN 1346	1 T	69.35	−0.13	AN 1798	1 T	3.81	−0.03	AN 1509	1 T	7.94	−0.18
45	AN 2047	2 T	9.50	−0.06	AN 2249	2 T	69.43	−0.14	AN 1945	1 T	3.76	−0.03	AN 1895	1 T	7.57	−0.21
46	AN 2006	1 T	9.58	−0.06	AN 773	2 T	69.35	−0.15	AN 1509	2 T	4.39	−0.03	AN 632	2 T	6.51	−0.22
47	AN 1534	2 T	9.69	−0.06	AN 871	2 T	69.29	−0.15	AN 288	2 T	4.27	−0.04	AN 1890	2 T	5.49	−0.24
48	AN 871	1 T	9.23	−0.08	AN 2047	1 T	68.75	−0.16	AN 1346	1 T	4.04	−0.04	AN 1267	2 T	7.61	−0.24
49	AN 632	2 T	9.15	−0.09	AN 1569	1 T	69.13	−0.17	AN 2047	1 T	3.83	−0.04	AN 197	2 T	7.21	−0.26
50	AN 1267	2 T	9.14	−0.10	AN 13	1 T	67.72	−0.18	AN 1960	1 T	3.82	−0.05	AN 1798	1 T	7.79	−0.27
51	AN 1569	2 T	9.21	−0.10	AN 1267	1 T	68.99	−0.19	AN 1534	2 T	4.20	−0.05	AN 2144	2 T	6.83	−0.30
52	AN 1346	2 T	9.07	−0.10	AN 642	2 T	68.58	−0.22	AN 642	1 T	3.81	−0.05	AN 2047	1 T	7.31	−0.30
53	AN 1384	1 T	9.19	−0.11	AN 1895	2 T	68.99	−0.23	AN 1895	1 T	3.78	−0.05	AN 1945	1 T	7.44	−0.32
54	AN 1895	1 T	9.30	−0.11	AN 2006	2 T	68.36	−0.23	AN 1384	1 T	3.66	−0.06	AN 467	1 T	6.78	−0.33
55	AN 1450	2 T	9.27	−0.11	AN 197	2 T	67.79	−0.25	AN 632	2 T	4.51	−0.07	AN 642	2 T	6.59	−0.34
56	AN 13	2 T	9.80	−0.12	AN 288	2 T	68.77	−0.26	AN 467	2 T	4.31	−0.07	AN 1276	2 T	6.90	−0.40
57	AN 197	1 T	9.79	−0.12	AN 1384	2 T	69.24	−0.27	AN 2033	1 T	3.76	−0.07	AN 1450	2 T	6.41	−0.41
58	AN 1945	2 T	9.08	−0.15	AN 1534	1 T	68.07	−0.29	AN 877	1 T	3.61	−0.08	AN 1665	1 T	7.54	−0.45
59	AN 2249	1 T	9.00	−0.17	AN 1945	1 T	69.28	−0.41	AN 2249	1 T	3.72	−0.10	AN 594	2 T	6.72	−0.46
60	AN 288	1 T	9.28	−0.17	AN 632	1 T	68.75	−0.44	AN 846	2 T	4.22	−0.10	AN 2033	2 T	6.26	−0.52
61	AN2036	2 T	9.21	−0.18	AN 2036	1 T	68.59	−0.52 *	AN 13	2 T	4.28	−0.11 *	AN 1569	1 T	7.07	−0.58
62	AN 877	1 T	8.98	−0.19 *	AN 877	2 T	69.46	−0.60 *	AN 1890	2 T	4.14	−0.14 *	AN 13	1 T	6.07	−0.84 **
Grand Mean			9.41		Grand Mean		69.14		Grand Mean		4.11		Grand Mean		7.34
St. Error			0.10		St. Error		0.24		St. Error		0.06		St. Error		0.32

* p < 0.05; ** p < 0.01; T—tester; 1 T—tester ZPL217; 2 T—tester ZPL-255/75-5.

). This result was obtained by crossing two parents with negative GCA values, where AN877 had a GCA of −0.27 (9.14%) (Table 5), while the tester line ZPL-255/75-5 had a GCA of −0.03 (9.36%) (Table 6).

For starch, the best SCA combinations were AN877 × ZPL217 and AN2036 × ZPL-255/75-5, with SCA values of 0.60 ** and 0.52 *, and starch contents of 70.73% and 69.56%, respectively (Table 8). The first of these combinations involved two positive GCA combiners, with AN877 being the second-best GCA combiner with a value of 0.95 (70.09%), though just below the threshold for significance. The second combination, AN2036 × ZPL-255/75-5 (Table 8), was a cross between two negative GCA combiners, with values of −0.06 (69.08%) (Table 5) and −0.04 (69.10%) (Table 6), respectively.

Regarding lipids, AN1890 × ZPL-217 and AN13 × ZPL217 exhibited the most significant SCA values of 0.14 ** (4.04%) and 0.11 * (4.11%), respectively (Table 8). These results were obtained by crossing one negative (AN1890; −0.02) and one positive (AN13; 0.09) GCA combiner with the negative ZPL217 tester −0.19 (3.91%) (Table 6).

While AN1890 × ZPL217 stood out as a top SCA combination, the remaining top ten hybrids were largely derived from crosses between opposite GCA combinations, pointing to a clear direction for maize breeding aimed at improving lipid content. Interestingly, the highest lipid content hybrids in the top ten were all obtained by crossing positive GCA combiners with the tester ZPL217.

For grain yield, the most significant and the only SCA combination was AN13 × ZPL-255/75-5, with an SCA value of 0.84 ** (7.05 t ha⁻¹) (Table 8), resulting from a cross between two negative GCA combiners, AN13 −0.78 (6.56 t ha⁻¹) (Table 5) and ZPL-255/75-5 −0.35 (6.99 t ha⁻¹) (Table 6). This combination produced an average grain yield of 7.34 t ha^–1. The highest-yielding hybrid, AN1267 × ZPL217, resulted from a cross between two positive GCA combiners, and while its SCA value was just 0.24, the average yield reached 8.80 t ha^–1 (Table 8).

The highest grain-yielding combinations included the top eight hybrids: AN1267 × ZPL217 (8.80 t ha⁻¹), AN1346 × ZPL217 (8.45 t ha⁻¹), AN197 × ZPL217 (8.44 t ha⁻¹), AN1276 × ZPL217 (8.41 t ha⁻¹), AN594 × ZPL217 (8.34 t ha⁻¹), AN2036 × ZPL217 (8.30 t ha⁻¹), AN877 × ZPL217 (8.25 t ha⁻¹), and AN2144 × ZPL217 (8.13 t ha⁻¹). These hybrids consistently involved two positive GCA combiners, with ZPL217 as the common tester across all combinations. These findings are consistent with previous studies, such as those by Singh et al. [22], who demonstrated that hybrids derived from parents with high GCA effects exhibit superior grain yield. Similarly, Muluneh et al. [37] reported that the combination of two positive GCA parents significantly enhanced grain yield performance across diverse environments. Our results, which highlight the role of ZPL217 as a common tester contributing to higher yields, further support the hypothesis that hybrids formed by two positive GCA combiners, particularly when paired with well-adapted testers, achieve significantly higher grain yields.

For protein content, the top eight hybrids with the highest protein percentages were as follows: AN13 × ZPL217 (10.10%), AN13 × ZPL-255/75-5 (9.80%), AN197 × ZPL-255/75-5 (9.97%), AN197 × ZPL217 (9.79%), AN1534 × ZPL217 (9.87%), AN1534 × ZPL-255/75-5 (9.69%), AN467 × ZPL217 (9.77%), AN467 × ZPL-255/75-5 (9.76%), and AN2047 × ZPL217. Each of these combinations contained either two positive GCA combiners or one negative and one positive GCA combiner, emphasizing the critical role of positive GCA combiners in enhancing protein content.

This trend was also observed among the top 24 hybrids, which predominantly consisted of crosses between positive and negative GCA combiners. Hybrids involving two negative GCA combiners, such as AN1895 × ZPL-255/75-5, were ranked from 25th place onward, further underscoring the importance of selecting positive GCA combiners for improving protein content in maize breeding.

The relationships among basic chemical composition traits (proteins, starch, and lipids) and their interdependence with a key agronomic trait (yield) were examined to provide insights into potential trade-offs and synergies in maize breeding programs (

Table 9. Correlation matrix for yield and basic chemical composition.

	Proteins	Starch	Lipids	Yield
PROTEINS	1	−0.948 **	0.130	−0.718 **
STARCH	−0.948 **	1	−0.368 *	0.587 **
LIPIDS	0.130	−0.368 *	1	0.024
YIELD	−0.718 **	0.587 **	0.024	1

Note: Significance levels: ** (p < 0.01); * (p < 0.05). Determinant = 0.014.

).

A very strong negative correlation was observed between protein and starch content (r = −0.948; p < 0.01) as well as between protein content and grain yield (r = −0.718; p < 0.01). Conversely, a moderate positive correlation was identified between starch content and grain yield (r = 0.587; p < 0.01). Correlations between other traits, including proteins and lipids, as well as lipids and grain yield, were not statistically significant.

4. Discussion

In this study, we investigated the combining ability of 31 local maize landraces, focusing on the basic chemical composition of the kernels, including protein, starch, and lipid content, as well as yield. Improving the nutritional value of maize kernels is of great importance. Research by Jiang et al. [38] confirmed the equivalence of results obtained by Near Infrared Reflectance (NIR) spectroscopy and traditional wet chemical methods for protein, starch, and lipid content. The NIR spectroscopy for the analysis of nutritional components in whole maize kernels has been confirmed as a rapid and efficient method, consistent with the findings of previous studies [29,30]. Unfortunately, quality parameters are often negatively correlated with yield. Therefore, breeding efforts must find a balance between improving nutritional value and minimizing yield loss [12]. General (GCA) and specific (SCA) combining abilities play a crucial role in selecting breeding material and developing hybrids with desirable characteristics [39]. Prai-anun et al. [27] state that understanding the GCA and SCA effects between inbred lines is critical to optimizing heterosis in hybrid-based biofortification breeding, both for yield-related traits and carotenoids. It is essential to identify germplasm with high-yielding potential that simultaneously exhibits high levels of protein, starch, and lipids. Moreover, these landraces must possess good combining abilities for the aforementioned traits.

Combining ability analysis can estimate the relative importance and modes of gene action involved in the commercial hybrids to desired traits [26]. Dodiya and Joshi [40] emphasized the dominant role of non-additive genetic components in the inheritance of oil and protein content. However, the results of our study demonstrate that additive effects had a more significant impact, particularly in the AN13 population for protein content. This discrepancy may be attributed to the different genotypes used in the studies, as well as the specific agroecological conditions in the PAN3 zone.

Identification of local landraces, such as AN13 and AN197, highlights the importance of ex situ genetic resources for increasing protein content in hybrid combinations. Genebanks store representative germplasm samples of crop species and provide access to these materials for their evaluation and integration into modern breeding programs [3], but a deep characterization is required to be able to exploit available agrobiodiversity. AN13, as a positive GCA combiner for protein, has the potential to improve the nutritional profile of hybrids. Through inheritance mechanisms and gene expression, this landrace contributes to increased protein synthesis in the offspring, achieving the desired effect of higher protein content in the final product. Simultaneously, the identified landraces, AN13 and AN197, as positive GCA combiners for protein content, show a significant impact on decreasing starch content. Huang et al. [41] reported QTL on chromosome 9 (THP9) that encodes asparagine synthetase 4 enzyme, for high protein in teosinte. They also state that the introgression of this QTL into modern maize inbreds and hybrids will enhance the accumulation of free amino acids, especially asparagine, without affecting yield. Future research could therefore be directed toward the identification of QTLs responsible for increased protein synthesis in selected landraces and their introgression into elite breeding germplasm.

A strong negative correlation between protein and starch content (−0.948 **) was observed, indicating that increases in protein content occur at the expense of starch and vice versa. This finding underscores the complex interplay between these two components and aligns with the physiological and biochemical constraints governing their accumulation in maize grain. The negative correlation between protein content and grain yield (−0.718 **) suggests that selecting for higher protein content could adversely affect yield, posing a challenge for breeding programs aiming to optimize both traits. These results are consistent with previous findings by Semenčenko et al. [42] and Zhang et al. [43], which similarly reported a trade-off between protein content and grain yield. While protein content shows a trade-off with yield, starch content displays a complementary relationship.

Kumar et al. [44] also found that both additive and non-additive gene effects were significant for oil content, with crosses exhibiting high SCA effects showing elevated heterotic values. The following research [27] supports the idea that both genetic components play essential roles in maize breeding programs focused on the basic chemical composition of grain and yield.

High starch content is particularly important for both the starch industry and food production. In this context, the local landrace AN594 stands out as a positive GCA combiner for starch content. This landrace can be valuable for developing breeding material aimed at optimizing maize for industrial purposes, including starch and bioethanol production [42]. Also, intra-population recurrent selection is inevitable, using trait indices, to increase targeted traits due to the observed larger additive genetic variance for starch, oil, and protein content in maize grain [45,46]. Lipids are primarily localized in the germ, while starch is found in the endosperm, which aligns with the literature data [47]. The local landrace AN632 stands out as a positive GCA combiner for lipids, offering the potential for breeding with increased lipid content. The local landrace AN1267 was identified as the best GCA combiner for yield, highlighting the importance of this genotype in achieving high yields.

A positive correlation between starch content and yield (0.587 **) was also identified, suggesting the potential for developing breeding material that combines high yield with high starch content. Furthermore, a negative correlation between starch and lipid content (−0.368) was observed, indicating that varieties with higher starch content typically have lower lipid levels. The observed correlations between chemical composition traits and grain yield provide important insights for maize breeding strategies. The strong negative correlation between protein and starch content (r = −0.948, p < 0.01) suggests a potential trade-off between these two components, meaning that selection for higher protein content may reduce starch content, which could negatively impact yield. Likewise, the negative correlation between protein content and yield (r = −0.718, p < 0.01) indicates that increasing protein content might result in a decrease in yield, highlighting the need for careful balancing in breeding programs. On the other hand, the positive correlation between starch content and yield (r = 0.587, p < 0.01) indicates that selecting higher starch content could contribute to increased grain yield. These findings emphasize the importance of considering trade-offs and synergies between traits when making selection decisions in maize breeding programs.

Our findings demonstrated that the environment, genotype, and their interaction with the tester significantly affect the amount of protein, oil, and starch, although the tester’s influence is only very significant for the amount of oil. It supports the findings of Ben Mariem et al. [48], who contended that, depending on certain circumstances and genotypes, environmental interactions can have a substantial impact on protein content. In line with our findings, the same authors noted that genotype interactions and environmental factors are important in determining yield potential, particularly in a variety of environmental settings.

Carena [49] reported that parents with high GCA values always produce hybrids with high SCA estimates. Conversely, Ivy and Hawlader [50] found that good general combining parents do not always exhibit high SCA effects in their hybrids. These differing viewpoints underline the complexity of hybrid performance and the importance of both GCA and SCA effects in selection strategies. Our study concurs with the findings of Joshi et al. [51] and Meseka et al. [52], who similarly identified superior parental lines and hybrids based on GCA and SCA effects. Specifically, we observed that GCA effects were more prominent for protein content, highlighting the importance of additive genetic components in the inheritance of this trait.

The difference between the grain yield results of the best SCA combination (AN13 × ZPL-255/75-5) and the combination between the best GCA combiners (AN1267 × ZPL217) can be explained by the fact that SCA indicates a specific interaction between parental lines in a particular cross, while GCA represents the general ability of parents to contribute to desired traits in all combinations. AN13 × ZPL-255/75-5 is the best specific combinator for yield because this combination produces an exceptionally good result in this cross, whereas AN1267 × ZPL217, as a general combinator, demonstrates consistently high yield across different combinations due to its positive GCA values. This is a common situation in plant breeding, where GCA and SCA can lead to different conclusions about the best parents in the best hybrids. Arellano-Vázquez et al. [20] stated that the performance of the maize hybrids was mainly determined by the genetic effects of dominance, epistasis, or interaction when the effects of SCA predominate, and the best way to investigate them is through hybridization.

The significance of both additive and non-additive effects in controlling key agronomic traits, such as protein, starch, and lipid content, as well as yield potential, underscores the necessity for a balanced breeding approach. This approach strategically combines additive and non-additive effects to optimize both quality and yield, addressing the trade-offs between these traits. It emphasizes the improvement of qualitative traits while not affecting grain yield as the ultimate trait. Both additive and non-additive genetic effects also played significant roles in the expression of quality [27] and yield-related traits [26] in maize.

The results of this study provide valuable guidelines for breeding programs aimed at developing genotypes with improved grain quality within the agroecological zone PAN3. The landraces AN13, AN197, AN632, and AN594, with positive GCA values for protein, lipids, and starch content, can be recommended as good sources of these traits for maize breeding programs for temperate climate regions. Information about the SCA of the testers used can be good guidelines for the mentioned landraces’ agronomic properties improvement while simultaneously preserving their heterotic pattern.

5. Conclusions

To further improve maize breeding strategies, a combination of conventional breeding techniques and marker-assisted selection (MAS) is recommended. The identification of QTLs associated with protein, starch, and lipid content could enable a more precise selection of superior genotypes, accelerating the development of hybrids with enhanced nutritional profiles. Additionally, breeding efforts should prioritize the integration of landraces with high GCA values, such as AN13, AN197, and AN632, into hybrid development programs to maximize both yield and grain quality.

The positive correlation between starch content and yield further underscores the potential for optimizing both yield and nutritional traits through careful selection of parent lines. The inbred lines ZPL217 and ZPL-255/75-5 showed distinct advantages, with ZPL217 contributing to higher protein and starch contents and ZPL-255/75-5 excelling in yield and lipid content. The inclusion of these maize lines in breeding programs aimed at improving grain quality would be highly beneficial.

NIR spectroscopy is a fast, non-destructive, and cost-effective tool for the evaluation of the basic chemical composition of maize grains. Its use in the early generations of maize line development to choose genotypes with better grain quality can significantly speed up and streamline the process of breeding maize for quality, making it a valuable method for modern breeding programs. The findings from this study contribute to the ongoing effort to enhance maize’s nutritional value, which is crucial not only for human consumption but also for animal feed and industrial use.

Future research should explore the genetic basis of these traits to deepen our understanding of how maize landraces interact with testers at the molecular level, paving the way for more efficient breeding strategies that align with both high yields and enhanced grain quality.