Effects of Extreme Combined Abiotic Stress on Yield and Quality of Maize Hybrids

Abstract

Maize is one of the top five field crops worldwide and is indispensable as animal feed, serves as a raw material in many industries, and is a staple for human food. However, its production is under increasing pressure mainly due to abiotic stress. Drought and excessive precipitation, air temperature fluctuations, and reduced soil fertility due to inadequate soil pH reactions are among the biggest challenges that must be overcome. Therefore, the goal of this study was to determine the effects of these combined stressful abiotic conditions on maize grain yield and quality and to determine the genetic-specific response of maize genotypes in such conditions. The experiment was set up in eastern Croatia according to the randomized complete block design in four replications. A total of 10 maize hybrids of different FAO maturity groups were evaluated across four diverse environments, each subjected to one or two abiotic stresses (extreme precipitation, drought, high air temperatures, and acidic soil). Analysis of variance revealed that all treatment effects were statistically significant, except for the effect of hybrids on grain yield. Depending on the effect of abiotic stress, the variations among environments were up to 51.4% for yield and up to 12.1%, 18.9%, and 0.81% for protein, oil, and starch content, respectively. Differences among hybrids were less pronounced for yield (7.9%), while for protein (13.5%), oil (17.3%), and starch content (1.5%) were similar. However, the largest variation was found for the interaction effect. In the conducted research, ENV2 recorded the highest grain yield, along with the highest oil and starch content, as well as the second-highest protein content, while the hybrid effect remained unclear. Generally, ENV4 had the greatest negative impact due to the combined effects of extreme abiotic stresses, including soil acidity, drought, and high air temperatures.

1. Introduction

Field crops are very often exposed to unfavorable or even extreme biotic and abiotic factors that adversely affect plant metabolism, growth, and development, as well as yield and quality. Primary food production in the world is, among other things, limited by soil fertility and climate, e.g., weather conditions [1,2,3]. For successful maize production, water and heat are of the utmost importance. Maize is a plant that requires an adequate water supply for physiological development and adequate yield. However, water demand differs across stages of development, being lowest during early development and peaking during the reproductive phase and grain filling. Drought stress, particularly when combined with high temperatures during flowering, pollination, and fertilization, delays silking, causes silk drying, and leads to pollen abortion, which may result in poor fertilization or incomplete grain development [4,5]. Heat stress during the late vegetative phase causes maize yield reduction of up to 30%, while during the flowering and lag phase, up to 50% [6]. When combined with drought, a drastically greater yield loss is possible. In general, insufficient precipitation in combination with high temperatures represents one of the main limiting factors for plant growth and development that are least manageable by agronomic practices and cause agricultural losses of more than 220 billion dollars [7]. Kim and Lee [8] state that climate change negatively affects a wide range of physiological and developmental processes: it reduces seed germination and seedling growth, reduces leaf expansion, stunts shoot and root growth, accelerates ripening, reduces plant height, disrupts photosynthesis, affects nutrient absorption, decreases pollination and fertilization, lowers grain yield, etc. Although Croatia, especially its eastern part, belongs to an optimal agroecological maize growing zone, significant year-to-year variations in the average maize yield are often caused primarily by drought and high temperatures in the summer months during maize flowering and fertilization stages or other agrotechnical practices [9,10]. Maize is the dominant field crop in Croatia. According to the Croatian Bureau of Statics, around 30% of arable land is occupied with this crop, with an annual production of around 2 million tons. In total cereal production, expressed in terms of quantity, maize accounts for 55% to 60% and contributes around 8% to 10% of the total agricultural output expressed in prices [11].

Besides the influence of climate, to achieve high yields and quality, maize requires adequate soil fertility, which is very often impaired by a lack or excess of certain mineral elements. Soil acidity is one of the most widespread global constraints to successful production. It is estimated that approximately 50% of all arable soils worldwide have some level of acidity [12]. In Croatia, about 32% of total agricultural land is acidic, mostly in the eastern region, where the majority of agricultural production takes place [13,14]. According to research, two soil types dominate: pseudogley and alluvial soil. Soil acidification is a very slow and long-term natural process of accumulation of H⁺ ions in both the liquid and solid phases of the soil. However, this process has been accelerated in recent decades due to anthropogenic activities. In general, low soil fertility in acidic soils is the result of the toxicity of Al, Mn, and Fe and the reduced availability of P, Ca, and Mg. Maize requires at least 16 macro and microelements for normal growth and development through the entire life cycle, which are mainly acquired from the soil [15]. Microelements, whose role is almost equal to macroelements, are essential, although they are present in very low concentrations in the soil and the plant. Under conditions of excessive soil acidity, the availability of most microelements may increase. However, their uptake and translocation within the plant depend on numerous factors [16]. According to the literature, there are two main approaches to mitigate the adverse effects of soil acidity. One proposes agrotechnical and/or land improvement measures, such as liming, phosphatization, and targeted fertilization [17,18,19]. The other underlines the genetic variability in maize, suggesting that differences in the sensitivity or tolerance of genotypes to specific soil conditions can be used for better plant adaptation rather than the soil constantly adapting to the plant [20,21].

In addition to agroecological factors such as climate and soil that influence the growth and development of maize, the role of genotype is also very important because different genotypes are characterized by the varying ability to absorb available elements in the soil, even under the same agroecological conditions. Selecting a suitable maize genotype for a specific growing region can lead to higher grain yields, better quality as well as increased uptake and accumulation of microelements in the grain [22,23].

Based on the hypothesis, abiotic stress can be harmful to maize plants, reducing both yield and grain quality, while the combination of two or more abiotic stresses can be unpredictable and may cause severe damage to the plant. Therefore, the aim of this research was to evaluate the effects of stressful abiotic conditions, extreme precipitation, drought, high air temperature, and soil acidity on maize grain yield and quality, as well as to determine the genotype-specific responses of ten maize hybrids under such stressful agroecological conditions.

2. Materials and Methods

2.1. Plant Material and Field Experiment Description

Ten commercial maize hybrids (HYB) of standard dent kernel type and different FAO groups used in the study were developed by the Agricultural Institute Osijek (

Table 1. List of hybrids in research and FAO groups.

Label	Name	FAO Group
HYB1	Drava 404	420
HYB2	OS 430	440
HYB3	OSSK 444	450
HYB4	OS 499	490
HYB5	OSSK 515	520
HYB6	OS 5717	520
HYB7	OSSK 552	580
HYB8	OSSK 596	590
HYB9	OSSK 602	620
HYB10	OSSK 617	610

). They are adapted for growing in diverse environments of the southern part of the Pannonian plain, mostly for grain production, and are the most common choice of maize producers. Drava 404 is the earliest hybrid used for both grain and silage production. The hybrids OS 430 and OSSK 444 are suitable for low-input production, whereas OS 499 is tolerant to water deficit. The hybrids OSSK 515, OS 5717, and OS 522 have high-yield stability across the region, while OSSK 596 and OSSK 602 are more tolerant to European corn borer and Western corn rootworms. OSSK 617 is a stay-green hybrid suitable for grain and silage production.

The research was conducted under field conditions during 2016 and 2017 in four environments affected by one or two abiotic stresses in eastern Croatia (Osijek-Baranja County). Environment 1 (ENV1) is characterized by a neutral soil pH reaction and excessive rainfall during the maize vegetation period (Figure 1,

Table 2. Basic chemical properties of all environments.

Environment	pH	pH	P2O5	K2O	Organic Matter	CaCO3	HA 1
	(H2O)	(KCl)	mg 100 g−1		%	%
ENV1	7.39	6.52	28.6	39.8	1.97	2.11
ENV2	6.84	5.62	31.9	34.8	1.84	1.47
ENV3	5.80	4.20	28.6	35.1	2.02		6.81
ENV4	5.07	4.06	24.2	37.8	1.70		6.06

¹ HA—hydrolytic acidity.

and Table S1). Environment 2 (ENV2) also had a neutral soil pH reaction but a pronounced dry period and higher air temperatures during the growing season (Figure 2, Table 2 and Table S1). Environment 3 (ENV3) and 4 (ENV4) were characterized by double abiotic stress caused by low soil pH along with either excessive rainfall (Figure 3, Table 2 and Table S2) or drought (Figure 4, Table 2 and Table S2) during the growing season. The location of the field trials for ENV1 and ENV2 was Osijek (45°32′12.8′′ N 18°44′19.0′′ E) and Podgorač for ENV3 and ENV4 (45°29′02.0′′ N 18°13′34.4′′ E). The locations belong to the optimal maize growing area with a very small difference in altitude from 94 m (ENV 1 and ENV2) to 125 m (ENV3 and ENV4).

The experiment was set up according to a randomized complete block design (RCBD) with four replications and three treatments (environment, hybrid, and their interaction). The basic plot of each hybrid was 14.0 m² and consisted of two rows 10.0 m long with a row spacing of 20 cm for hybrids HYB1 to HYB7 (planned set 71,428 plants per hectare) and 22 cm for hybrids HYB8 to HYB10 (planned set 64,935 plants per hectare). The replicates were separated by a 1 m path, and a protective belt was sown around the entire experiment to avoid the possible effect of edge plants and reduce the possible negative impact of wind. The total area of each replicate was 140 m², and the entire experiment per environment was 560 m². At both locations, standard agricultural techniques for maize were implemented, and the sowing dates were at the end of April in all four environments. Maize was manually harvested at the full ripening phase so that two rows of each hybrid were collected from each plot. The cob and grain mass were measured using a digital electronic scale (Kern CH 25 K50 and Kern MH 10 K10, Frankfurt am Main, Germany). Grain moisture at harvest was measured with a digital moisture device (Wile 55, Farmcomp Agroelectronics, Tuusula, Finland). Grain yield per hectare was calculated based on ear weight per plot, ear fraction, and grain moisture content and expressed on the basis of the realized plant density in t ha⁻¹ at 14% moisture. Determination of protein, oil, and starch content in maize grain samples was carried out in the laboratory of the Agricultural Institute Osijek using the Infratec 1241 Grain Analyzer (Foss, Hillerød, Denmark), a device that works on the principle of NIT technology (near-infrared transmission), i.e., near-infrared (570–1050 nm) transmission.

2.2. Soil and Weather Analysis

For research purposes, soil samples were taken immediately after harvesting the previous pre-crop with a soil probe. At each location, an average sample of approximately 1 kg was taken, consisting of 20 individual samples. After being transported to the laboratory, the soil samples were cleaned of organic residues and other impurities and dried in a thin layer at room temperature. The air-dried soil samples were then ground using a special soil mill, sieved through a 2 mm sieve, and homogenized, after which they were prepared for analysis according to the standard prescribed procedure [24]. Furthermore, the basic chemical properties of the soil were determined—soil pH reaction, organic matter, concentration of available phosphorus and potassium, hydrolytic acidity, and CaCO₃ content. Current acidity (pH_H2O) was determined by electrochemical measurement in a soil suspension with distilled water and substitutional or exchangeable acidity (pH_KCl) in a soil suspension with a 1 M KCl solution [25]. Hydrolytic acidity, which represents the total potential acidity of the soil, was determined by the Na-acetate extraction method, according to Kappen, and the carbonate content in the soil by the volumetric method [26]. The organic matter content was determined by the bichromate method [27]. The content of available phosphorus and potassium in the soil was determined by the AL method, according to Egnér, Riehm, and Domingo [28].

The weather analysis in this study is based on decadal and total monthly precipitation (mm) and mean decadal and monthly air temperatures (°C) for the maize growing season compared to the multi-year average values for the last 30-year period. Data were obtained from meteorological stations of the State Hydrometeorological Institute of the Republic of Croatia. Since the maize growing season takes place in the warm part of the year, the values shown refer to the period from April to October.

2.3. Environment Description

In eastern Croatia, or rather in the research area, conditions are suitable for growing maize, and the main cause of yield variations from year to year is the lack of precipitation, which most often occurs in July and August when maize needs for water are high. In the study, two excessive and two drought conditions in the context of precipitation and air temperature were observed. Two excessive environments are characterized by higher precipitation than the multi-year average and very similar air temperatures. The total amount of precipitation at ENV1 during the vegetation period was 744 mm, or 60% more than the multi-year average, with all months except July having higher amounts of precipitation (Figure 1). The precipitation values at ENV3 also follow the trend of the specific conditions of the vegetation year. Thus, a total of 806 mm of precipitation fell, which is as much as 62% more than the average (Figure 3). The average daily air temperatures in both environments were at the level of the multi-year average.

Conversely, stressed environments were warmer and had less precipitation than the multi-year average. At ENV2, the total amount of precipitation during the maize growing season (April–October) was only 275 mm or 39% lower than the observed multi-year average (Figure 2). In the same period, the ENV4 area received 41% less precipitation than average or only 307 mm (Figure 4). Stressful conditions are even more pronounced when compared to an excessive environment. Thus, at ENV2 and ENV4, the amount of precipitation was 170% and 163% less compared to ENV1 and ENV3, respectively. A pronounced lack of precipitation in both environments was recorded during August, September, and June for the ENV2 location. At the same time, the average daily air temperature at both environments was higher during the growing season compared to the multi-year observed average. At ENV2, it was higher in total by 1.0 °C, and at ENV4 by a total of 1.2 °C. Such conditions are not conducive to normal maize development and the achievement of high yields and good quality.

In all environments, soil chemical analyses showed that these soils are rich in accessible phosphorus and potassium but with different pH values. The soil in ENV1 was neutral and very slightly acidic in ENV2. In contrast, the soil of ENV3 and ENV4 experimental plots was characterized by a very acidic pH reaction, and on average, the pH in KCl was only 4.13. The average hydrolytic acidity was 6.44.

There was no significant difference among the environments in terms of the supply of plant-available phosphorus and potassium to the soil. On average, for ENV3 and ENV4, 26.4 mg 100 g⁻¹ of P₂O₅ and 36.4 mg 100 g⁻¹ of K₂O were determined, and 30.2 mg 100 g⁻¹ of P₂O₅ and 37.3 mg 100 g⁻¹ of K₂O were determined at ENV1 and ENV2 (Table 2). All tested environments had very low levels of organic matter. According to the World Reference Base (WRB) classification, the soil type at ENV1 and ENV2 was identified as eutric cambisol, while ENV3 and ENV4 were characterized as stagnosol or pseudogley.

2.4. Description of Statistical Analysis

Statistical processing of data on the investigated properties was carried out by individual and combined analysis of variance (ANOVA). The data were analyzed using General Linear Models (GLM). In our model, hybrids and environments were treated as fixed effects, while replications within environments were treated as random effects. When the F-test indicated a statistically significant effect (p ≤ 0.05), we performed pairwise comparisons using Student’s t-test to identify specific differences between treatments. The obtained results were processed in the computer program SAS Software 9.1.3.

3. Results

3.1. Statistical Analysis

Based on the analysis of variance and the F-test, a significant influence of environment and interaction with the hybrids on maize grain yield was observed. Compared to the average yield in the Republic of Croatia, in this research, a significantly higher yield was achieved (5.70 t ha⁻¹ and 8.61 t ha⁻¹, respectively). However, no statistical significance was observed for the hybrids even though a larger number were tested (

Table 3. Analysis of variance of main tested properties.

	Mean	Coeff. Var.	R-Square	F Value	Pr > F	LSD0.05
	Yield
ENV	8.61	9.97	0.86	324.7	<0.0001	0.379
HYB	8.61	27.2	0.01	0.15	0.998	-
Interaction	8.61	8.59	0.92	35.9	<0.0001	1.036
	Protein
ENV	7.88	6.93	0.32	24.5	<0.0001	0.241
HYB	7.88	6.98	0.33	8.46	<0.0001	0.384
Interaction	7.88	4.56	0.77	10.5	<0.0001	0.503
	Oil
ENV	3.45	7.97	0.48	48.5	<0.0001	0.121
HYB	3.45	9.26	0.33	8.18	<0.0001	0.223
Interaction	3.45	4.01	0.90	27.3	<0.0001	0.194
	Starch
ENV	73.0	0.84	0.14	8.57	<0.0001	0.272
HYB	73.0	0.79	0.28	6.42	<0.0001	0.402
Interaction	73.0	0.64	0.62	5.12	<0.0001	0.650

Coeff. Var.—coefficient of variation; LSD—last significant difference.

). In the context of quality traits, ANOVA indicated statistically significant effects for all treatments (environment, hybrid, and interaction, respectively). The greatest influence on all traits was the environment, followed by the interaction of the environment and hybrids for most traits except starch content, while the hybrids had the weakest influence on the tested traits, except starch content in maize grains.

3.2. Environment Effect

Grain yield and quality are the most important traits for the feed, food, and process industries, which are heavily influenced by environmental factors. In this study, maize hybrids were grown in four completely different environments, including excessive rainfall, drought, high air temperature, and acidic soil in various combinations. The highest yield was achieved on ENV2, while almost half the yield (51.4%) was achieved on ENV4 (11.3 t ha⁻¹ and 5.5 t ha⁻¹). In general, under double stress conditions (ENV3 and ENV4), average yields were the lowest (Figure 5a).

The results regarding quality parameters were less consistent than those for grain yield (Figure 5b–d). For grain protein content, the best environment turned out to be the one with the most pronounced stress conditions (ENV4, 8.3%). For starch content, ENV2 and ENV3 (73.3% and 73.2%) were shown to be the best options, while ENV2 stood out for oil content (3.81%). Determined variations between environments for protein and oil content were 12.1% and 18.9%, respectively. The least pronounced differences, although statistically significant, were achieved between environments for grain starch content and varied from 73.3% (ENV2) to 72.7% (ENV4) or only 0.81%.

In the conducted research, ENV2 achieved the highest grain yield, both oil and starch content, as well as the second-best protein content.

3.3. Hybrid Effect

In contrast to the environment, the hybrid effect was less pronounced, and the differences between them were smaller. Unfortunately, analysis of variance did not show statistical significance for grain yield, although variations were identified. The most productive hybrids across all four environments were HYB4 and HYB5 (8.92 t ha⁻¹ and 8.86 t ha⁻¹), while HYB3 and HYB7 achieved the lowest yields of 8.33 t ha⁻¹ and 8.20 t ha⁻¹ (Figure 6). A total of 720 kg ha⁻¹ of grain is the difference between the highest- and lowest-yielding maize hybrids, representing only 7.9%.

Further, based on analysis of variance, HYB3 and HYB2 had the statistically highest protein content (8.69% and 8.46%), while seven hybrids had significantly lower values with variations of 13.6%. It is interesting to highlight HYB4, which stands out with the highest yield but the lowest protein content, whereas HYB3 recorded the highest protein content and almost the lowest grain yield.

Oil content also showed statistically significant variation but with a regular distribution among hybrids from the earliest to the latest. Two early-maturing hybrids statistically achieved higher values compared to hybrids of the later one, and the variation ranged from 3.81% (HYB2) to 3.15% (HYB9) or 17.3%.

In contrast to protein and oil, the starch content was less pronounced and ranged from 73.6% (HYB6) to 72.5% (HYB 2) or only 1.5%, which indicates a more stable distribution among hybrids. Hybrids with the highest starch content had the lowest protein values, which was expected given the negative correlation.

3.4. Interaction Effect

Although the environment had the greatest influence on the tested traits according to the analysis of variance (Table 3), its interaction with hybrid showed the greatest differences in variation. Also, variations have been identified both between hybrids and environments. In ENV1, the yield variation was from 10.84 t ha⁻¹ (HYB8) to 8.59 t ha⁻¹ (HYB7) or 20.8%. In ENV2, the yield variation was from 11.92 t ha⁻¹ (HYB1) to 10.74 t ha⁻¹ (HYB10) or 9.9%. Further, ENV3 achieved a variation of 23.6% (8.97 t ha⁻¹, HYB9 and 6.85 t ha⁻¹, HYB6), while ENV4 had the highest value of 27.8% or from 6.59 t ha⁻¹ (HYB5) to 4.76 t ha⁻¹ (HYB7). Across all four environments and ten maize hybrid yields, variation was from 11.92 t ha⁻¹ (HYB1, ENV2) to only 4.75 t ha⁻¹ (HYB7, ENV4) or 60.1% (

Table 4. Interaction effect on mean grain yield values (t ha⁻¹) of ten maize hybrids and four environments according to analysis of variance.

Environment/	ENV1	ENV2	ENV3	ENV4
Hybrid	t ha−1	t ha−1	t ha−1	t ha−1
HYB1	9.70 efg	11.92 a	8.10 ijk	5.32 prq
HYB2	8.85 ghi	11.23 ab	8.30 hijk	6.18 nop
HYB3	8.84 ghi	10.83 bcd	7.98 ijk	5.68 oprq
HYB4	9.96 cdef	11.26 ab	8.85 ghi	5.59 oprq
HYB5	9.80 defg	11.60 ab	7.43 klm	6.59 mno
HYB6	9.97 cdef	11.64 ab	6.85 lmn	5.81 opq
HYB7	8.59 hij	11.15 ab	8.30 hijk	4.76 r
HYB8	10.84 bc	11.39 ab	7.69 jkl	4.96 rq
HYB9	9.22 fgh	11.03 ab	8.97 fghi	5.40 prq
HYB10	9.95 cdef	10.74 bcde	8.35 hijk	4.78 rq

Means marked with the same letter are not significantly different according to LSD test at the p ≤ 0.05 level.

). The stress-tolerant hybrids in this study were in the following order: HYB4–HYB5–HYB1–HYB8–HYB9–HYB2–HYB6–HYB10–HYB3–HYB7.

Variation in values was recorded for grain quality properties. The highest protein content in grain was determined at ENV4 (8.32%) and the lowest in ENV3 (7.31%), while environments 1 and 2 had somewhat more stable protein contents (

Table 5. Interaction effect on mean grain protein content (%) of ten maize hybrids and four environments according to analysis of variance.

Environment/	ENV1	ENV2	ENV3	ENV4
Hybrid	%	%	%	%
HYB1	8.00 fghij	7.78 hijkl	7.08 mnop	8.15 efgh
HYB2	8.38 bcdef	8.75 bc	8.20 defgh	8.53 bcde
HYB3	8.20 defgh	8.85 b	8.30 cdefg	9.40 a
HYB4	7.48 jklmn	7.60 ijkl	7.03 nopq	8.08 efghi
HYB5	7.60 ijkl	8.08 efghi	7.30 klmno	7.95 fghij
HYB6	7.53 jklmn	7.50 jklmn	6.80 pq	8.53 bcde
HYB7	8.18 defgh	7.90 fghij	7.38 klmno	8.68 bcd
HYB8	7.75 hijk	7.98 fghij	6.55 q	7.75 hijkl
HYB9	7.83 ghijk	7.93 fghij	7.58 ijkl	7.93 fghij
HYB10	7.60 ijkl	7.98 fghij	6.88 opq	8.25 cdefgh

Means marked with the same letter are not significantly different according to LSD test at the p ≤ 0.05 level.

). Observed by environment, the deviation values between hybrids ranged from 10.7% (ENV1), 15.3% (ENV2), 17.6% (ENV4), and 21.2% (ENV3). However, in the entire experiment, the highest value was achieved by HYB3 on ENV4 (9.40%), while the statistically lowest content was achieved by HYB8 on ENV3 (6.55%) or 30.3% less. The hybrids were ranked from highest to lowest protein content as follows: HYB3–HYB2–HYB6–HYB9–HYB1–HYB5–HYB10–HYB7–HYB4–HYB8.

Similar indicators of variation have been found also for oil content (

Table 6. Interaction effect on mean grain oil content (%) of ten maize hybrids and four environments according to analysis of variance.

Environment/	ENV1	ENV2	ENV3	ENV4
Hybrid	%	%	%	%
HYB1	3.85 de	4.23 ab	3.48 hijk	3.43 ijkl
HYB2	3.85 de	4.33 a	3.65 fgh	3.40 jklmn
HYB3	3.33 klmnop	3.55 ghij	3.05 qrs	2.90 st
HYB4	3.60 ghi	3.83 def	3.43 ijklm	2.98 st
HYB5	3.73 efg	4.08 bc	3.38 jklmno	3.45 ijkl
HYB6	3.40 jkmn	3.60 ghi	3.27 lmnop	2.82 t
HYB7	3.83 def	3.92 cd	3.20 opqr	3.03 rs
HYB8	3.33 klmnop	3.55 ghij	3.35 klmnop	3.23 nopq
HYB9	3.18 pqr	3.38 jklmno	3.25 mnop	2.80 t
HYB10	3.35 klmnop	3.60 ghi	3.35 klmnop	2.90 st

Means marked with the same letter are not significantly different according to LSD test at the p ≤ 0.05 level.

). The average variation in all environments was 18.6% and ranged between hybrids from 17.4% for ENV1, 21.9% for ENV2, 16.4% for ENV3, and 18.8% for ENV4 (Table 5). In addition, the lowest oil content was achieved by HYB9 in ENV4 (2.80%) and the highest by HYB2 in ENV 2 (4.33%), which makes a difference of 35.2%. The schedule of oil content in the study was in the following order from highest to lowest: HYB2–HYB1–HYB5–HYB6–HYB4–HYB8–HYB10–HYB7–HYB3–HYB9.

In contrast to all other examined properties, the variation in starch content was minimal and ranged between hybrids from only 1.5% (ENV1), 1.8% (ENV3), 2.0% (ENV2), and 2.8% (ENV4). Even the variation between all environments and hybrids was minimal at only 3.2% in comparison to the other tested traits (

Table 7. Interaction effect on mean grain starch content (%) of ten maize hybrids and four environments according to analysis of variance.

Environment/	ENV1	ENV2	ENV3	ENV4
Hybrid	%	%	%	%
HYB1	72.6 klmnop	73.2 bcdefghij	73.6 abcde	73.2 bcdefghij
HYB2	72.3 mnop	72.5 lmnop	72.6 lmnop	72.8 ghijklmno
HYB3	72.9 fghijklmn	73.2 bcdefghij	73.2 bcdefghij	71.6 q
HYB4	73.0 efghijklm	73.9 ab	73.8 abc	73.7 abcd
HYB5	72.7 ijklmnop	72.9 fghijklmn	73.1 defghijkl	73.3 bcdefghij
HYB6	73.4 abcdefgh	74.0 a	73.9 ab	73.4 abcdefgh
HYB7	72.5 lmnop	73.3 abcdefghi	73.1 defghijkl	72.3 mnop
HYB8	73.1 defghijkl	73.3 abcdefghi	73.3 abcdefghi	72.2 op
HYB9	72.7 ijklmnop	73.4 abcdefg	72.3 mnop	72.2 op
HYB10	72.8 ghijklmno	73.5 abcdef	73.3 abcdefghi	72.8 ghijklmno

Means marked with the same letter are not significantly different according to LSD test at the p ≤ 0.05 level.

). The highest values were achieved by HYB6 in ENV2 (74.0%) and the lowest by HYB3 in ENV4 (71.6%). Although this trait is very important in the context of domestic animal nutrition because it provides energy and has a positive effect on body weight gain, this research shows that the effect of environment and hybrids is relatively weak. The highest starch content in maize kernels was found in hybrids in the following order: HYB4–HYB7–HYB1–HYB10–HYB5–HYB8–HYB6–HYB3–HYB9–HYB2.

In the conducted research, based on mean values, some statistically significant correlations between grain yield and protein, oil, and starch content were determined (

Table 8. Correlations between investigated properties.

	Yield	Protein	Oil	Starch
Yield	1
Protein	−0.16 ns	1
Oil	0.72 **	−0.04 ns	1
Starch	0.32 *	−0.53 **	0.10 ns	1

ns—not significant; *—significant at the p ≥ 0.05; **—significant at the p ≥ 0.01.

). Maize grain yield had significant correlations with all properties, with the exception of protein content, while a negative correlation was confirmed between protein and starch content. The strongest positive correlation of yield was with the content of oil in the grain (r = 0.721).

4. Discussion

The main goal of maize breeding and production is to achieve high and stable yields with the best possible grain quality. However, usually, they are limited by two major factors. These are the genotype, or the total inheritance of an individual, and the environment, which provides more or less favorable conditions for the expression of a particular genotype [29,30,31]. Also, under intense stress conditions, interactions between environment and genotypes are more pronounced and complex because genotypes show different levels of phenotypic expression in different environmental conditions, and the variations in values were extremely large [32,33,34]. In our study, analysis of variance determined the statistical significance of all treatments (environment, genotype, and their interaction) except hybrids for grain yield. Further, very large variations between environments, hybrids, and interactions were found.

According to the Croatian Bureau of Statistics [35], maize yields in the Republic of Croatia have varied greatly over the years. In the period from 2005 to 2023, the average yield was 5.7 t ha⁻¹ but with extremely large variation from 4.0 t ha⁻¹ (2010) to 9.0 t ha⁻¹ (2019). The biggest reasons for the variation are environmental factors, especially precipitation and temperature regimes, having a significant impact. In this study, a relatively high average yield (8.61 t ha⁻¹) was achieved despite negative environmental influences such as drought, high air temperature, soil acidity, or even excessive rainfall, which may be related to the good genetic material of the tested hybrids. The best yield was achieved on ENV2, which is characterized by neutral soil pH and severe drought during August and September (5 mm and 17 mm, respectively). At the same time, air temperature was higher for 1.8 °C and 3.6 °C. However, in the second and third decade of July, when maize is in the most sensitive phase of development, a small amount of precipitation fell (61 mm), which was sufficient for flowering and grain fertilization. Similar observations of the influence of weather conditions have been made by others [36,37,38,39]. The authors state that the application of conventional agricultural techniques, good fertilization and liming on acidic soils, and the sowing of drought-tolerant hybrids at optimal times or the application of new sowing technologies can mitigate negative impacts on maize yields. Also, conservation agriculture is one of the solutions for the future due to a whole range of positive advantages [40,41].

Maize’s water requirements depend primarily on the stage of development, but many other factors, such as weather conditions, soil fertility, photosynthesis intensity, hybrid properties, etc., can also affect water uptake and consumption. Maize needs about 64 mm of water in May, 106 mm in June, 121 mm in July, 124 mm in August, and 63 mm of water in September, while hybrids of later vegetation require a small amount of water in October as well [42]. Although maize requires large amounts of water, primarily due to its habit, in our study, environments with extreme rainfall (Figure 1 and Figure 3) were not the most productive. Thus, ENV1 and ENV3 had lower yields. For example, during May, ENV1 received twice as much precipitation as the observed multi-year average (121 mm and 62 mm, respectively), while in ENV3, the amount was even higher (190 mm and 70 mm), which is not conducive to the normal development of the maize plant. So, although maize requires large amounts of water, in extreme cases and in the early developmental stages, it can have the opposite effect, as observed in our research. Based on data from 1981 to 2016, Li et al. [43] estimated a yield reduction of up to 34% due to excessive rainfall, mainly in cooler areas and with poorly drained soils, while Markovic et al. [44] report a 7.6% yield reduction under the same conditions. In this research, yield was lower by 15.2% and 28.4% in extreme wet conditions compared to the best yield conditions (ENV2). In contrast to all other environments, ENV 4 achieved statistically the lowest maize yield (5.51 t ha⁻¹) due to the negative effect of double abiotic stress, reduced fertility, i.e., pronounced soil acidity as well as extreme drought, especially during April, August, and September (22 mm, 15 mm, and 17 mm, respectively) and higher air temperatures in all months. In such challenging conditions, only two hybrids had yields slightly more than 6 t ha and three hybrids less than 5 t ha (Table 4), which is an extremely low yield for maize growing conditions in eastern Croatia. Therefore, the maize response to different abiotic stress or their combination depends on the interaction of stresses [45]. For example, in our study, maize growing on acid soil in combination with extreme precipitation achieved better yield (8.08 t ha⁻¹) than growing on the same soil and drought conditions (5.51 t ha⁻¹). Depending on the abiotic outdoor environment, maize hybrids react reversibly or irreversibly, and the process is always complex and dynamic [46]. Over the last 40 years, maize breeding has increased grain yield, and one possible reason is that green, photosynthetically active leaves remain on the stem longer, which increases and prolongs the accumulation of nutrients in the grain [47]. Furthermore, plant architecture between hybrids, such as physiological metabolisms, photosynthesis characteristics, and stomatal morphologies of leaves, have a significant impact on maize yield under abiotic stress conditions [48]. In conducted research, hybrids showed high yield and great yield variability, although no statistical significance was determined between them. The highest-yielding hybrids (HYB4 and HYB5) were not always the best in the tested environments (Figure 6), but they showed a certain stability. Also, they showed good results in extremely rainy and dry environments, as well as in acidic soil. Such hybrids should be strongly considered when planning to grow across different environments, especially under abiotic stresses. In any case, the difference between the highest and lowest-yielding hybrid was 720 kg ha. Although the seed owner (Agricultural Institute Osijek) emphasizes the advantage of HYB5, HYB6, and HYB7 as a high-yield, stable hybrid across the environments, in our research, only HYB5 has shown relatively high value across all four stress environments. In addition, we cannot underline the advantage of the early and late maturity groups of maize hybrids in the context of yield because, from each group, some hybrids were excellent thought environments. The above indicates that sowing maize of different vegetation maturity groups and different specific characteristics is the best option for achieving high yields because there is no ideal hybrid that is totally resistant to abiotic stress. In terms of the best hybrids for wet or dry areas, it is not possible to draw a clear line. However, HYB4 and HYB10 may be a good choice for wetter areas, while for drier growing areas, we can only recommend HYB5.

In the context of maize farming, yield is always the dominant trait, while quality is given less importance. However, despite its relatively low protein and oil content, maize is consumed in larger quantities by livestock [49]. The highest protein percentage was recorded in ENV4, which is characterized by double abiotic stress. This is in line with other research, which shows that under any form of stress, plants accumulate a greater amount of protein [50,51]. In both environments where extreme rainfall was recorded, protein content was lower, especially in ENV3, which was exposed to excessive rainfall and an acid soil condition. In the context of tested hybrids, only two (HYB3 and HYB4) show statistically higher values (8.46% and 8.69%). The lowest protein content in the grain was achieved by HYB8 (7.51%), although no statistically significant difference was found with the other six hybrids. This indicates that a larger number of hybrids in the study achieved slightly lower values. Although Nehe et al. [52] claim that environment has the greatest influence on protein content, in our case, the influence of environment and genotype, i.e., hybrid, was almost equal.

Oil content in our research was under the influence of environment, hybrid, and interactions. Higher values were determined in environments with neutral soil pH, regardless of precipitation amounts and temperature (ENV1 and ENV2), which indicates that maize cultivation for the specific purpose of oil production should be on soils that have an almost neutral pH reaction. As with the protein content, the hybrids of earlier vegetation (HYB1 and HYB2) showed statistically higher values of oil content compared to all other hybrids.

In contrast to all other properties, starch content proved to be the most stable, as the variation among environments was only 0.81%, and among hybrids, 1.5%. Even in the interaction of environment and hybrids, the value was only 3.2%. Therefore, no matter under what abiotic stress conditions the tested hybrids are grown, they will show very uniform values. Very similar results were obtained by Wang et al. [53]. Based on the analysis of five maize hybrids and four environments, the deviation between environments was 0.92%, 0.56%, and 0.50%. In any case, the conclusions of other authors are contradictory. Some claim that abiotic stress (e.g., drought) significantly reduces starch content and increases amino acids [54], while others state that drought increases starch content [55]. The abiotic stress factors assessed in this study represent the greatest threat to maize production today—thus, this research provides useful insights into better understanding the outcomes of genotype–environment interactions in some modern maize hybrids. The results could be used in breeding programs to enhance overall productivity, especially in the context of climate change.

5. Conclusions

By growing maize hybrids under various abiotic stressful environments, the analysis of variance confirmed statistically significant effects for all treatments except one. Yield variation was most pronounced in the interaction between environment and hybrid (60.1%), followed by the environment alone (51.4%), while the smallest variation was observed among hybrids (7.9%). The most detrimental effect on yield was the combination of abiotic stress induced by low soil pH value associated with drought and high temperatures, followed by low soil pH reaction and extreme precipitation. In contrast, environments with adequate soil fertility and neutral pH values achieved the best results, regardless of weather conditions.

In terms of quality, the highest values were obtained from the interaction, whereas the environmental and hybrid treatments showed very similar values. Among all tested properties, starch content varied minimally across treatments, indicating a certain stability. Understanding the severity of abiotic stress and the specific responses of individual maize genotypes, it is possible to mitigate the negative impact of agroecological factors and improve both the yield and quality of maize grain.