1. Introduction
Bread wheat (
Triticum aestivum L.) is an important widely grown cereal crop grown with an annual production of 760 M tons [
1], occupying the biggest land area under cultivation of one crop in the 21st century with more than 219 million ha planted annually. Wheat end-use products play an essential role in human nutrition contributing to up to 20% of the daily intake of proteins and 21% of food calories. By 2050, the global population is expected to reach 9 billion people on Earth. The production and material to feed this number are not set up today, which will demand approximately a 70% increase in worldwide wheat production [
2,
3,
4]. Wheat grain morphology traits are all grain parameters related to the physical characteristics of the grain. The size, color, and shape of the grain, as well as the ratio of grain width by grain length, can vary depending on the genotype and environmental conditions [
5,
6] and, therefore, can affect the nutritional content, processing quality, and market value. Grain morphology has been linked to thousand kernel weight (TKW), as well as flour yield and end-use quality [
7]. In addition, grain length, width, and TKW influence directly test weight which determines flour quality grade and therefore impacts flour market price for millers [
6]. Wheat yield is determined by several complex agronomic traits, such as plant height, earliness, vernalization and photoperiod response, resistance to abiotic stresses, etc. However, grain characteristics also play a major role in increasing yield [
8], and are correlated with yield due to their stability and effect on grain weight [
9,
10,
11]. Grain size and weight are considered as main components influencing wheat yield. Longer and wider grains do not only directly influence yield but also positively impact seeding vigor and early development, which in turn promote and stabilize yield potential [
12]. However, it is important to consider the potential implications for seed cost. Larger seeds typically require more resources and may result in higher seed costs for farmers. All these morphological grain characteristics as well as TKW are traits highly influenced by the environment [
13]. Usually, large grains contain more starch and nutrients than small grains, which leads to higher yield and more flour. To optimize yield and nutritional quality, scientists must consider these factors related to grain morphology by selecting genotypes with desirable shape characteristics for each specific environment to maximize yield, quality, and profitability. In wheat production, yield is influenced by several components, including the number of grains per plant, grain size, and grain weight [
14]. For example, it could be advantageous to select genotypes with larger grains in an irrigated environment where water availability is more constant. These characteristics often correlate with higher yield potential and can contribute to better water and nutrient utilization, ultimately enhancing profitability. Conversely, in a rainfed environment where water availability fluctuates smaller genotypes may be more suitable due to their greater drought tolerance and higher water use efficiency, supporting yield stability and profitability. While cultivars with larger grains are not necessarily more drought tolerant, they may exhibit characteristics such as deeper root systems or greater water use efficiency, which can contribute to drought resilience [
15]. Therefore, when considering drought tolerance, it is essential to evaluate multiple characteristics beyond grain size. Several environmental factors affect wheat productivity and quality; sowing dates emerge as the key determinant factor. The optimal sowing date in regions like Morocco and Spain usually starts in the second half of November. However, delaying or advancing sowing dates may impact significantly wheat growth, yield, and quality [
16]. Early sowing dates may coincide with high-temperature stress during the initial developmental stages, which reduces germination rate and therefore reduces plant population and yield. Late sowing dates may coincide with terminal heat during the flowering and/or grain-filling period, which affect significantly the grain characteristics and therefore yield and quality [
17,
18,
19,
20]. Genotype by environment interaction (GEI) is a common phenomenon in all multi-environment trials. They represent a major challenge for breeders who want to develop and advance materials throughout generations that are more adapted to several environmental conditions. The statistical analysis of GEI is a widely used method to identify the optimum value of the targeted trait in several environments [
21]. Understanding the GEI effect on grain morphology is therefore an important consideration in bread wheat breeding programs, as it affects not only the agronomic parameters but also quality parameters [
22]. The objectives of this study were (1) to study the morphological parameters, including grain shape and size, and protein content of the Winter Wheat Association Genetics Initiative (WWAGI) panel under eleven environments (location × year × sowing date combinations) using multi-environment trial data and identify ideal genotypes with high performance and wide adaptability, (2) to disseminate the effect and importance of the climatic variables on each trait, (3) to identify molecular markers associated with morphological grain traits and protein content.
3. Discussion
Wheat is a staple crop of global importance for a large portion of the worldwide population, contributing to up to 20% of its daily intake. In recent years, wheat breeders and scientists have achieved slow yield progress, and the yield plateau has become an alarming threat to food security. Wheat quality is a complex concept, including grain morphology and nutritional composition. Grain morphological traits are some of the most important quality characteristics due to their significant impact on grain weight, flour yield, and market value [
27]. However, the negative relationship between protein content and grain yield components in wheats has limited the development of both traits [
28,
29]. The notable heritability of a wide range of grain quality traits in wheat renders it highly amenable to DNA marker selection application. This technique allows for the simultaneous selection of varieties with high yield, as assessed through conventional field testing, alongside those with superior grain quality. Furthermore, resilience is vital for adapting to variable weather conditions. Aligning with this goal, our study is dedicated to the effective identification of markers that denote resilience in wheat, specifically relating to grain morphology and protein content. In the current study, we used GEI to understand grain morphology and protein content changes and performance of the genotypes (advanced lines and varieties) in winter wheat multi-environment trials. In our population, genotypes showed a significant level of variation for morphological grain size, TKW, and protein content with moderate to high heritability. In agreement with many previous studies, there was a significant strong positive correlation between grain morphological traits, and negative when they were compared to protein content [
30,
31]. Elsewhere [
32], a significant negative correlation between TKW and protein content in a population of 194 of bread wheat was observed. Moreover, a strong positive correlation between grain size and weight in a population of 231 synthetic wheats has been reported [
27]. Gao et al. [
33] reported a strong correlation between grain area, perimeter, length, width, and TKW. Simmonds et al. [
34] revealed that a mutation in
TaGW2-A1 resulted in an increase in grain width and length, which in turn increased thousand kernel weight. The relationship between these traits, linking positively grain morphological characteristics and grain weight, and negatively with protein content, suggest that longer and wider grains have the capacity to store more starch and carbohydrates, which increases grain weight and results in the dilution effect for protein content. Kenzhebayeva et al. [
35] reported a low association between grain shape and protein content using wheat transformed by radiation. The highest mean of grain length, width, and TKW was recorded in LI17E, the environment with the lowest rainfall and the highest number of days, as compared to ANN17 and ANN18, but with good monthly distribution, suggesting that this environment could be considered as favorable cropping season for these genotypes in relation to rainfall. At the same time, the late planting in ANN18 and LI17 exhibited the highest value for protein content, while all other traits showed a significant decrease from the early planting to the late planting. This increase in protein can be attributed to the dilution effect under the low-stressed conditions in the early and medium planting dates due to an increase in carbohydrates in the grain as suggested by the higher TKW. According to the ANOVA table, genotype, environment, and GE interaction were highly significant in all trials. An exception was noted for grain color, which was not influenced by GE interaction. This suggests that the ranking and the selection for grain color remain stable across locations and environments. Similarly, Zhao et al. [
36] revealed no effect of genotype by environment on grain color using seventy-nine diverse spring wheat genotypes where the environment had the lowest impact on genotypic performance for color with only 17% of the total variation. The GEI significantly contributed to the variation in protein content, accounting for 60% and representing a tenfold greater effect than the genotype. Similarly, for length, width, and TKW, the GEI also captured the highest value of variation with 37%, 44%, and 45%, respectively, with moderate influence as compared to the environment and genotype effect. This pattern suggests that the genotype by environment interaction accounts for the highest source of variation for protein content and therefore drives protein content variability for our population. At the same time, the GEI followed by environment represents the main component of variation for width and TKW. The genetic influence is almost as significant as the environment and GEI when it comes to length. Similar research results about the comparative importance of environment, genotype, and GEI were reported [
6,
37,
38,
39]. The substantial interaction observed between the genotype and environment results in reduced grain morphological traits and protein content stability in our panel. Thus, it is advised for wheat breeders to increase the number of genotypes in a mega-environment to identify the stable genotypes with better grain shape and optimal quality.
The Mediterranean area, which includes Spain and Morocco, is renowned for having a very variable climate within and between the seasons. Therefore, more advanced statistical techniques than the conventional ANOVA are needed for a thorough assessment of GE interactions. Based on the GGE biplot analysis, a Mega-Environment (ME) can be determined when different genotypes are adapted to distinct groups of environments and the variance between groups is higher than within groups [
23,
24,
25,
37,
40]. In our study, the different environments tested were clustered into two MEs for TKW and three for grain length and width. This result is in agreement with the result reported by [
41] where two winter wheat MEs in Ontario were suggested after examining the ME analysis and test site for winter wheat in Canada. Similar conclusions were made by [
42] in relation to winter wheat tested at eight locations who found between three and four MEs for grain yield. However, for protein content, the targeted region was not better presented as MEs but rather as a single environment, indicating that the evaluation of genotypes for this specific trait must be performed over years and locations. Therefore, the selection of future varieties adapted to the Moroccan and Spanish environments must prioritize genotypes with wide adaptability rather than specific adaptations or high protein concentration. The results of this study indicate that there was one or more promising genotype(s) for each group of MEs as the winning genotype(s) except for protein which was different every year. Genotypes 208, 129, and 234 were consistently the winning genotypes for grain length, while 64, 117, and 243 were the promising genotypes for grain width and genotypes 64, 181, and 243 for TKW. The definition of a ME states that a sufficient number of locations is necessary to confirm that genotype responses are consistent across locations within a ME [
37]. However, in the case of protein, this study’s limitation to just two locations, each with three sowing dates, made it difficult to draw definitive conclusions. Consequently, for more conclusive evidence of these MEs, future studies are recommended to include a broader range of locations. A study [
32] reported an unstructured GEI for grain protein content in 194 F7 RILs of bread wheat. Altogether, these results suggested that the behavior of these traits can be complex and unpredictable across different environmental conditions and sowing dates. This further supports the need for extensive research across varied locations to decode the complexities of genotype and environment interactions.
Climatic variables and weather conditions including temperature and rainfall during the heading date had probably the most significant influence on grain yield and quality. It is also responsible for controlling the adaptation of wheat to a broad range of environments [
43,
44,
45]. Many key climatic variables influencing wheat growth development under Mediterranean climates in Morocco and Spain varied between the eleven environments used in this study. It has been shown that minimum and maximum temperatures during growth stages, rainfall, terminal heat stresses, and late frost are key climatic factors affecting wheat adaptation in Mediterranean environments [
46]. For this scope, the impact of the climatic variables was assessed between and within each environment. Across environments, a high minimum temperature during the grain setting stage was identified as the key climatic variable negatively impacting grain width, while, for grain length, a significant negative relationship was detected with minimum, maximum, and average temperatures during the grain setting and filling stages. Moreover, our findings highlighted a greater influence of GEI on grain width compared to grain length. This could be attributed to the association of grain width with the plant’s ability to effectively fill the grain, indicating a higher sensitivity of grain width to environmental variations, which may play a crucial role in shaping the lateral dimensions of developing grains. While grain length exhibits a relatively less sensitive response to environmental conditions, reflecting a more constitutive and stable trait, the relationship between climatic variables and TKW was not significant. However, using the factorial regression analysis, a negative effect of the average temperature on grain perimeter before heading raises the possibility of a relationship between lower temperatures and constrained grain spatial expansion. This indicated that heading is preceded by a temperature-sensitive period, during which the grain’s structural development is influenced by lower temperatures, potentially changing the grain’s overall size and shape. Furthermore, rainfall appeared as a key factor, negatively influencing grain area, color, width, and TKW during the grain setting stage. This may result in decreased grain expansion due to different sowing dates, which restricts roots from receiving enough key nutrients necessary for proper grain setting interfering with vital metabolic processes that facilitate grain expansion. Several authors suggested that the stem-elongation stage to anthesis and post-anthesis including the grain setting stage are the most sensitive times [
47,
48,
49]. However, ref. [
50] reported that drought after flowering resulted in a 5.2% decrease in TKW and 20% reduction in grain number using a bread wheat cultivar, while for the grain filling period, a dominant role of rainfall and average temperature negatively impacted grain color and grain length, respectively, indicating that high average temperature resulted in accelerating some metabolic processes, which might impact grain elongation. Conversely, a positive influence of day length on grain circularity, TKW, and width highlights the critical role of day length in producing rounded and heavier grain. The grain filling period has been subject to many studies to determine the environmental factors influencing grain shape and weight in wheat [
16,
17,
51,
52]. This is in agreement with the result reported by [
44,
53,
54] where they showed that different genotypes tended to stop grain development early and accelerate physiological maturity when temperatures increased during the grain filling period resulting in diminished final grain weight.
Given that GEI often impacts genetic analysis used to identify the genomic area for quantitative parameters across many environments, it is an essential criterion to assess the significance and dependability of certain loci. In our study, GWASs for grain shape, TKW, and protein content were conducted for the WWAGI population in each environment independently. Overall, 603 significant MTAs with LOD score of ≥3.0 across eleven environments were identified on all chromosomes. However, the most important MTAs are those identified for grain length, width, TKW, and protein content (275 MTAs) due to their immediate effect on improving grain yield and quality. In the current analysis, seventy-eight MTAs were associated with at least two traits each, with eleven markers associated in different environments. Numerous studies have reported mapping for grain weight and shape in wheat, but there are few that explain the variation and stability of grain weight, size, and shape across several environments [
27,
55,
56]. The most important QTLs for grain length, TKW, and width were mapped on chromosomes 6D, 5D, and 2D [
56]. However, in this study, significant SNPs on all other chromosomes, except 1A, 3D, 5A, 6A, 7B, and 7D for length, 1A, 2A, 3A, 3D, 4B, 5D, and 7D for TKW, and 3D, 4B, 5D for width, were identified. Chromosomes 2B and 5D exhibited the highest number of significant SNPs, each with seven SNPs for grain length. Regarding TKW, the highest number was identified on chromosome 6D and 1D with six SNPs each, while for grain width five significant SNPs were detected on chromosome 5B. Interestingly, we found nine chromosomes (1B, 1D, 2B, 2D, 3B, 4A, 5B, 6B, 6D) that showed an overlapping of significant SNPs with the three traits at the same time indicating these regions as stable. This is in agreement with the result reported [
27] indicating a co-linearity of MTAs for different grain morphology traits on chromosomes 1A, 2B, 3A, 3D, and 5B with a complete region on chromosome 2B (51–69.9 cM) exhibiting 31 significant MTAs. Similarly, [
57] reported a QTL on chromosome 1D for the grain length, width, and weight, while none of the other QTLs detected for grain length and width overlapped with grain weight. Furthermore, previous research on wheat RIL populations by [
58] indicated an association between chromosomes 4D and 7D and grain width. In addition, QTLs for grain weight on chromosomes 1D, 2D, 5D, and 7D were reported by [
32,
55]. In our study, 130 significant MTAs affecting grain size and shape for more than two traits and/or environment have been identified across all chromosomes, and many of them were found within the same regions, suggesting that they may have novel allelic variability for grain size and shape and underlying the importance for meta-QTL identification for grain size, shape, and weight. Several QTLs and meta-QTLs were previously reported for grain weight, width, and length on different chromosomes. Zhang et al. [
59] reported a meta-QTL for grain weight on chromosome 2D. Campbell et al. [
60] highlighted QTLs on chromosomes 1A, 2A, 2B, 2D, and 3D, and chromosome 6D; [
61] revealed a group of QTLs for grain length, width, and weight. Wheat grain color has been subject to many studies identifying three genes located on the long arm of homoeologous 3 (3AL, 3BL, and 3DL chromosomes) controlled by the R-1 (red) genes as the genes governing wheat grain color [
62]. Similarly, our study identified 37 significant MTAs on chromosomes 3A and 3D, with the majority (34 MTAs) identified on 3D.
The genetic dissection conducted in this study highlights the relationship between specific (SNPs) and climatic variables, illuminating their impacts on key morphological grain traits. The highlighted relationships play a crucial role in different developmental stages, ranging from the early vegetative to the late grain-filling period. Notably, the increase in grain width due to the negative effect of allele T at marker AX-95224161 when the average maximum temperature is low highlights the sensitivity of this developmental stage to high temperatures. This result is consistent with previous research that shows how temperature negatively affects grain size during grain filling [
63,
64]. When subjecting spring wheat to both day and night warming by changes in sowing date and additional infrared heating, [
65] observed a 3% reduction in grain weight for every degree Celsius of post-anthesis mean temperature increase. Furthermore, allele A at marker AX-95141980 demonstrates a strong negative effect on TKW under long days, suggesting that photoperiod sensitivity plays a major role in determining grain weight and highlighting the significance of longer day lengths in affecting the length of the reproductive phase. Some studies [
66,
67] highlighted the relationship between shorter daylength and the increased number of spikelets per spike independently of the temperature, which may result in increased grain number. However, a negative association has been generally reported between TKW and grain number in wheat [
68,
69]. The genetic effect of allele G at marker AX-94459169 exhibits a positive impact on TKW and grain width under drought conditions. This marker could then be used, after validation, to increase grain width and therefore TKW and potentially flour yield under drought. A high negative correlation between yield and heading date under irrigated and rainfed environments was reported by [
70].
This study dissects the intricate relationship between genetic factors and environmental conditions in wheat modern and advanced lines, particularly focusing on grain quality traits such as grain morphology and protein content. With the significant heritability of these traits, the application of DNA markers emerges as a promising strategy, enabling the simultaneous selection of varieties that achieve both high yield and superior grain quality. The research underscores the importance of resilience, supporting the development of wheat varieties capable of adapting to variable weather conditions given the challenge of enhancing both protein content and yield simultaneously. The substantial role of genotype–environment interaction in trait variability highlights the dynamic relationship between genotypes and their environmental contexts, further emphasizing the need for a deep understanding of Mega Environments in selecting broadly adaptable wheat varieties. Insights from genome-wide association studies (GWASs) offer valuable markers associated with key traits, providing targets for breeding programs. Additionally, the study’s exploration of how climatic variables like temperature and rainfall affect grain traits offers novel perspectives for breeding wheat varieties tailored to specific environmental challenges. Altogether, these findings mark a significant step forward to optimizing wheat quality and yield, ensuring the crop’s resilience and adaptability across diverse growing conditions.