Applications of Artificial Intelligence in Wheat Breeding for Sustainable Food Security

Abstract

In agriculture, especially in crop breeding, innovative approaches are required to address the urgent issues posed by climate change and global food security. Artificial intelligence (AI) is a revolutionary technology in wheat breeding that provides new approaches to improve the ability of crops to withstand and produce higher yields in response to changing climate circumstances. This review paper examines the incorporation of artificial intelligence (AI) into conventional wheat breeding methods, with a focus on the contribution of AI in tackling the intricacies of contemporary agriculture. This review aims to assess the influence of AI technologies on enhancing the efficiency, precision, and sustainability of wheat breeding projects. We conduct a thorough analysis of recent research to evaluate several applications of artificial intelligence, such as machine learning (ML), deep learning (DL), and genomic selection (GS). These technologies expedite the swift analysis and interpretation of extensive datasets, augmenting the process of selecting and breeding wheat varieties that are well-suited to a wide range of environmental circumstances. The findings from the examined research demonstrate notable progress in wheat breeding as a result of artificial intelligence. ML algorithms have enhanced the precision of predicting phenotypic traits, whereas genomic selection has reduced the duration of breeding cycles. Utilizing artificial intelligence, high-throughput phenotyping allows for meticulous examination of plant characteristics under different stress environments, facilitating the identification of robust varieties. Furthermore, AI-driven models have exhibited superior predicted accuracies for crop productivity and disease resistance in comparison to conventional methods. AI technologies play a crucial role in the modernization of wheat breeding, providing significant enhancements in crop performance and adaptability. This integration not only facilitates the growth of wheat cultivars that provide large yields and can withstand stressful conditions but also strengthens global food security in the context of climate change. Ongoing study and collaboration across several fields are crucial to improving and optimizing these AI applications, ultimately enhancing their influence on sustainable agriculture.

1. Introduction

Food security is not about the condition in which individuals have access to all the necessary resources to consume a nourishing diet and maintain good health. Having enough food is important for good health and nutrition. A shortage of food can potentially cause malnutrition [1]. Food insecurity often happens when people do not have enough food to eat. Sometimes, it is because of natural disasters like droughts, floods, conflicts, or earthquakes. But most of the time, it is because people have been poor for a long time, not just because of disasters. Access to food is influenced by factors like income, skills, and privileges. Scientists use two main approaches to study this [2]. The first approach examines the overall situation, linking climate change to household responses. The second approach investigates how communities and households adapt to climate change on a smaller scale. The broader approach utilizes complex models that integrate climate, crop, and economic data to forecast the impact of climate change on crop yields, prices, incomes, and trade. However, this approach does not fully account for people’s adaptations to climate change. Smaller-scale studies rely on surveys of households to understand how people and communities adjust to climate change. The International Food Policy Research Institute’s (IFPRI) International Model for Policy Analysis of Agricultural Commodities and Trade (IMPACT) establishes a connection between climate change scenarios and the results of food supply, market, and prices. It also examines the economic implications of factors that affect food accessibility, such as food availability and utilization. These implications encompass food energy consumption and the nutritional status of children.

Climate change is the alteration of weather patterns in a specific area. It is caused by natural variations in the climate system and has significant implications for society and the environment. Artificial intelligence (AI) applications are transforming plant breeding and genetics to tackle the challenges posed by climate change. AI techniques such as machine learning (ML) and big data analytics enable the analysis of vast amounts of genetic and phenotypic data, accelerating the breeding of climate-resilient crops. For example, researchers have been applying next-generation AI to integrate multi-omics data, enhancing the breeding process by linking complex biological mechanisms with plant responses to environmental changes [3]. Additionally, ML-assisted approaches in modern plant breeding programs utilize high-throughput omics technologies to develop new crop varieties that perform better under diverse climate conditions [4]. These technologies are critical for improving yield performance, enhancing pest and disease resistance, and increasing overall resilience to climate variations, thereby securing food production with global climate change. Figure 1 visually represents how AI-based technologies, critical in enhancing yield performance and pest resistance, are being integrated into modern wheat breeding.

Climate change can affect food systems in various ways, such as altering crop production or changing market dynamics. The impact on food security varies by region. Urbanization and globalization are also driving changes in food systems [5]. Climate change affects food security differently across countries. In regions with lower hunger rates, the impact is milder and expected to ameliorate. Failing to address climate change does not pose immediate severe risks for undernutrition and malnutrition, with there being potential for improvements. Nonetheless, disparities in food access are projected to increase due to climate change variations. Vulnerable communities face heightened susceptibility to extreme weather events, necessitating immediate adaptation measures to counteract rapid adjustment for a few decades. This urgency is tied to prior greenhouse gas emissions. A higher frequency of extreme weather events in the future presents increased risks [6]. Formally identifying impacts requires comparing observed changes with a defined reference point. In the case of food production, non-climate factors make establishing a reference point difficult. These factors lack comprehensive characterization and are challenging to quantify [7]. Wheat, rice, and soybeans have a 1.0% annual increase in yields, while maize has a 1.5% annual increase. To prevent a 30% increase in grain prices by 2050, a minimum acceptable rate of yield progress is 1.1% per year [8]. The increasing demand for food requires intensified agricultural practices that manage nutrients carefully. Food security requires productive and sustainable production systems. Combined nutrient management is recommended for maximizing nutrient usage. Implementation of research findings in practical guidelines is necessary. Ongoing research is needed to support sustainable food security [9].

Traditional breeding methods have been essential in developing climate-resilient wheat varieties by leveraging natural agrobiodiversity and farmers’ traditional knowledge. However, genomic breeding methods have significantly enhanced the process by utilizing genomics, marker-assisted selection, and high-throughput phenotyping platforms to identify stress-tolerant genotypes and novel QTLs and genes for crop improvement [10]. These methods have been particularly effective in developing stress-tolerant wheat cultivars that can withstand drought and heat stress, which is crucial for combating the threats posed by climate change. By integrating genomics with participatory varietal selection, breeding programs can capture both genetic variations and environmental adaptations, leading to the development of high-performing, drought-adapted wheat varieties tailored to local farmer needs [11]. The combination of traditional and genomic breeding approaches allows for the creation of high-yielding climate-resilient wheat varieties that can thrive under challenging environmental conditions, ensuring food security for the growing population.

Artificial intelligence (AI) is revolutionizing wheat breeding programs by enhancing crop resilience and yield in the context of climate change. Recent advancements in AI have enabled non-linear modeling approaches, capturing complex interactions in genome-wide association studies (GWAS) and genomic selection (GS) to improve genomics-assisted breeding (GAB) [12]. Additionally, AI techniques have facilitated the integration of speed breeding, significantly reducing the time required for conventional breeding and ensuring greater accuracy and efficiency in developing stress-tolerant wheat cultivars [13]. Furthermore, machine learning models leveraging genotype and phenotype data, along with data from unmanned aerial systems, have shown promising results in predicting yield under varying environmental conditions, ultimately speeding up the delivery of improved wheat varieties with superior performance and resilience to climate challenges [14]. These AI-driven approaches hold the potential to address food security concerns by producing climate-smart wheat varieties with enhanced productivity and resilience.

Machine learning algorithms play a crucial role in enhancing the prediction of phenotypic traits in wheat to adapt better to climate variations. By leveraging both genotype and phenotype data, machine learning models can significantly improve yield prediction accuracy, as shown in studies focusing on wheat crop yield prediction [15]. These models, such as deep learning neural networks, can effectively integrate genetic and phenotypic information, accounting for genotype–environment interactions and providing more accurate predictions even in unseen environments. Additionally, incorporating environmental factors and climate indices into statistical regression models can further enhance the timeliness and robustness of yield predictions, aiding in developing adaptation strategies to mitigate the negative effects of climate change on crop productivity [16]. Ultimately, these advancements in machine learning algorithms offer valuable tools for scientists and farmers to optimize agricultural practices and ensure food security in the face of changing climates.

High-throughput phenotyping plays a crucial role in identifying wheat varieties resilient to extreme climate conditions by enabling precise and non-destructive evaluation of plant traits under stress [17]. Through daily imaging with RGB cameras and other sensors, high-throughput phenotyping allows for the monitoring of plant growth, yield-related traits, and responses to abiotic stresses like drought and heat. This method facilitates the identification of genotypes with stable harvest quality under stress conditions, as well as the discovery of beneficial QTLs that enhance yield and photosynthetic performance under drought [18]. Additionally, high-throughput phenotyping aids in understanding the genetic responses of wheat varieties to environmental covariates like temperature, contributing to the development of climate-resilient genotypes through informed breeding strategies.

Precision agriculture plays a crucial role in optimizing wheat production under the stress of changing climate patterns by addressing field-scale variability and efficiently matching resource inputs with crop needs. This approach involves the use of technologies like hyperspectral remote sensing, GIS, and GPS to enhance agricultural practices, improve crop yields, and reduce negative environmental impacts [19]. By synchronizing nitrogen fertilizer application with crop demand and managing spatial and temporal variability within fields, precision agriculture helps improve fertilizer use efficiency, reduce nitrogen losses, and ensure sustainable soil and crop management [20]. Additionally, the implementation of precision agriculture technologies has been shown to guarantee economic and agri-environmental efficiency, contributing to farm profitability and sustainability, especially in the face of climate change challenges.

2. Integrating AI and Plant Genetics to Enhance Food Security in Agriculture

Artificial intelligence (AI) is a technology that has human-like cognitive capabilities for acquiring knowledge, executing tasks, and making decisions. In agriculture, AI is being applied in various stages of the food system, including crop production, livestock production, postharvest management, food processing, and waste management [21]. Artificial intelligence empowers farmers and farm managers to apply accurate and focused farming techniques using field measurements that are relevant to the site’s agroclimatic conditions. It offers multiple benefits, such as improved sustainability, increased productivity, and enhanced decision-making for farmers and entrepreneurs. AI models are being used for prediction, weed control, resource management, and advanced crop care, among other functions, to achieve sustainable agriculture [22]. Figure 1 illustrates how AI models employ learning and vision technologies to enhance prediction accuracy and resource management in sustainable agriculture.

AI is utilized in crop breeding and improvement processes to enhance efficiency and accuracy. Advancements in high-throughput phenomics and genomics-assisted breeding (GAB) offer a great opportunity to generate crop cultivars that are tolerant to stress. Artificial intelligence (AI) enables the use of advanced computer capacity and new tools to process high-throughput data related to phenotyping, genotyping, and envirotyping. This facilitates the rapid discovery of genes and accelerates the progress of crop development programs [23]. In farming, AI helps by watching over the soil and crops, predicting what might happen next, and controlling robots on the farm. Refer to Figure 1 to see how AI-based learning and vision technologies are being utilized in farming to monitor soil and crop health and control robotic operations. This enables farmers to collect data for analysis and informed decision-making. Artificial intelligence, enhanced by deep learning, is an effective tool for efficiently handling vast and intricate datasets, such as the interactions between the microbiome, crop plants, and their surroundings, with the goal of enhancing agricultural systems. AI approaches are utilized in several domains, such as taxonomy, functional annotation, linking microbiome community with host features, genetic selection, field phenotyping, and illness forecasting [24]. As shown in Figure 1, AI-based learning and vision technologies play a crucial role in advancing wheat breeding and crop management through various techniques, including taxonomy and genomic selection. AI techniques can rapidly assess plant growth, identify beneficial genes, and optimize gene combinations for specific locations, thus accelerating breeding processes.

AI has various applications in wheat breeding research. As depicted in Figure 1, various AI-based learning and vision technologies are applied in wheat breeding to optimize both research and practical applications. One specific application is the use of AI-powered software systems to analyze wheat spikes and yield performance traits using low-cost drones and deep learning models [25]. AI technology can be used to analyze high-throughput phenotyping, genotyping, and envirotyping data, which aids in comprehending the intricate mechanisms involved in the production of agronomic traits. Envirotyping is a pivotal technique that delves into the non-genetic factors influencing the phenotypic adaptation of organisms. It involves characterizing environments, profiling environmental quality descriptors, and identifying environmental similarities to enhance genomic predictions and breeding strategies [26]. Additionally, artificial intelligence (AI) is utilized to create intelligent artificial climate chambers. These chambers are cost-effective and can be set up anywhere, allowing for crop cultivation and accurate data collection on the characteristics of the crop canopy. This technology enables breeders to research wheat cultivation and canopy characteristics with great precision [27]. Furthermore, AI tools such as geographical information systems (GIS) and fuzzy inference systems (FIS) are used to evaluate wheat’s suitability for cultivation, providing quantitative findings that can benefit decision-makers.

AI technologies offer several advantages over traditional methods in wheat breeding. Firstly, AI enables the capture of non-linear and epistatic interactions in genome-wide association studies (GWAS) and genomic selection (GS), which traditional linear models fail to capture. This allows for a more accurate understanding of complex traits and their interactions, leading to improved crop improvement [25]. Secondly, AI-powered software systems, such as CropQuant-Air, utilize deep learning models and image processing algorithms to automate large-scale phenotyping and phenotypic analysis of wheat spikes, providing a cost-effective and efficient approach. Additionally, AI models, such as the deep learning method used in DS1, have shown higher genomic prediction accuracy compared to conventional models, especially when predicting across different environments [28]. AI technologies offer the potential to revolutionize wheat breeding by enhancing accuracy, efficiency, and the ability to analyze complex traits and interactions.

The objective of studying the applications of artificial intelligence (AI) in wheat breeding and sustainable agriculture is to harness advanced technological solutions that enhance productivity, efficiency, and sustainability in agricultural practices. This study is crucial as it addresses several pressing challenges faced by the agriculture sector, including the increasing demand for food due to global population growth, environmental sustainability concerns, and adaptation to climate change. AI’s integration into wheat breeding programs can significantly accelerate the development of high-yielding, pest-resistant wheat varieties, which are critical for ensuring food security. Moreover, AI-driven tools can optimize resource use and reduce the environmental footprint of farming practices by facilitating precise farming techniques, thus promoting sustainability. As the challenges of climate change and resource scarcity intensify, exploring and adopting AI solutions in agriculture becomes not only beneficial but essential for adapting and thriving in changing global conditions. Research has underscored the transformative potential of AI in revolutionizing farming practices, enhancing crop management, and fostering sustainable agricultural outputs [29,30].

3. AI-Powered Advancements in Sustainable Wheat Breeding

In this warm weather, all climate forecasts indicate severe weather patterns [31]. Climate change impacts crop productivity differently. A rise in temperature by 1 °C globally reduces crop yield by 10–20% [32]. The effect may worsen by the century’s end, with a projected temperature increase of 2–4 °C. This will impact crop production [33]. Climate change permanently altered weather conditions and impacted global agriculture [34]. Extreme temperature changes at sensitive stages impact wheat yield, grain weight, and grain size significantly [35]. Wheat production was tested with high temperature and CO₂-enhanced concentrations. The findings indicated that at a temperature of 36 ± 2 °C during anthesis, there was a 13% decrease in production, and most grains were sterile [36].

Various studies have documented the occurrence of climate change, which has a direct impact on the prediction of crop yield. It has been observed that a mere increase of 1 degree in temperature leads to a significant decrease in the growth attributes of crops, ultimately affecting the yield. Furthermore, a comprehensive analysis of the change in temperature during the growing season has been documented. These studies have successfully predicted the crop model for global climate change over a span of 100 years, taking into consideration variations in temperature and rainfall and their consequent effects on wheat yield. This prediction is based on extensive data collected over the course of 100 years [37]. AI techniques such as machine learning (ML) and big data analytics enable the analysis of vast amounts of genetic and phenotypic data, accelerating the breeding of climate-resilient crops. For example, researchers have been applying next-generation AI to integrate multi-omics data, enhancing the breeding process by linking complex biological mechanisms with plant responses to environmental changes [3].

Production is diminished, and spikes become susceptible to disease stress during periods of high-temperature spikes. When the temperature exceeds 32 °C during the anthesis period, the size of the grain becomes shorter, and the duration of grain filling in the spikes is also decreased, thereby impacting the overall yield of wheat [38]. When it comes to growing wheat in places where rain is the main source of water, changes in how much it rains really affect things. If there is less rain, it directly makes it harder to grow wheat, and that means less wheat is produced. For every degree the temperature goes up, the amount of wheat produced can drop by about 5–7 percent [39].

3.1. AI Application and Morphological Traits for Sustainable Wheat Yield

Root architecture plays a crucial role in nutrient and water uptake in wheat. Studies have shown that different root traits are associated with improved water-limited situations that affect the intake of resources [40]. Cultivars released after the Green Revolution period exhibited increased root length, depth, and steep angle frequency, which are important for resource uptake in water-limited environments [41]. Additionally, a direct relationship exists between the depth of plant roots and features linked to crop yield when water availability is limited [42]. Genetic studies have identified quantitative trait loci (QTLs) for root traits in wheat, including a major QTL for total root length (TRL) that showed pleiotropic effects on various root traits [43]. Furthermore, a pot experiment comparing wheat varieties with different root sizes found that water and nitrogen uptake were more strongly influenced by resource availability than root size [44]. These findings suggest that root architecture, including traits such as root length, depth, and branching, can significantly impact nutrient and water uptake in wheat, making them important targets for breeding programs in drought-prone regions.

Plant height is an important agronomic trait that affects wheat yield. The reduced-height (RHT) dwarfing alleles (Rht1b) have a positive effect on the lodging and harvest index, but they also reduce coleoptile length, biomass production, and yield potential in some environments [45]. Hyperspectral Imaging for Crop Analysis: Automated machine learning (AutoML) technologies integrated with hyperspectral imaging systems are used to assess crop physiological characteristics and predict yield. Several genes have also been identified that control plant height in wheat. QTL analysis has also identified genetic regions associated with plant height and yield component traits in wheat. LiDAR (Light Detection and Ranging) generates detailed 3D models of wheat fields for structural analysis, aiding in accurate plant height. Overall, plant height plays a significant role in wheat yield under different environmental conditions, and understanding the genetic factors controlling plant height can help in improving grain yield in wheat [46].

Canopy structure has a significant impact on light interception and photosynthesis in wheat. Optimizing canopy light distribution (CLD) has been shown to improve light utility and yield without changing other inputs [47]. Studies have found that canopy structure affects the distribution of light and nitrogen (N) within the canopy, with N-efficient cultivars having lower light attenuation coefficients and higher plant compactness [48]. Precise architectural characteristics are essential for accurately simulating the photosynthetic production of a canopy. Models that lack architectural aspects may deviate significantly from the actual photosynthetic production [49]. The genetic architecture of flag leaf traits, such as area, length, width, and angle, also influences canopy photosynthesis and yield in wheat [50]. Additionally, nonfoliar tissues can contribute to canopy photosynthesis, with their contribution varying at different growth stages [51]. Overall, understanding and optimizing canopy structure is crucial for improving light interception and photosynthesis in wheat.

Genetic factors influencing tillering capacity in wheat include the genes TERMINAL FLOWER 1 (TFL1) [52], tiller number1 (TN1) [53], deacetylase HST1-like (TaHST1L) [54], and tiller inhibition 5 (TIN5) [55]. TFL1 homologs, such as TaTFL1-5s, are implicated in tiller regulation through auxin and cytokinin signaling [56]. TN1 promotes tiller bud outgrowth by inhibiting abscisic acid (ABA) biosynthesis and signaling. TaHST1L controls the tiller angle by mediating auxin signal transduction and regulating endogenous auxin levels. TIN5, located on chromosome Tu7, is involved in tiller formation and regulation patterns. These genetic factors play important roles in determining shoot architecture, tiller number, and grain yield in wheat. Automated ground robots operate autonomously in wheat fields to collect data on soil properties, plant health, and microclimate conditions, reducing labor needs.

Breeding strategies aimed at optimizing morphological traits for sustainable wheat production have been explored in several studies. One study analyzed a genetically diverse wheat population and found that multi-trait selection indices can efficiently optimize the correlations between characteristics and combine desired features, such as yield and weed competitive ability [57]. Another study focused on molecular breeding techniques and highlighted their success in improving wheat cultivars with multiple resistance traits against pests and diseases [58]. Additionally, a study on tetraploid wheat landraces suggested a methodology that integrates molecular genotyping with the analysis of morpho-agronomic features in order to assess genetic diversity and select parent lines for breeding programs [59]. Furthermore, a study evaluated different wheat genotypes under heat and drought stress conditions and identified potential sources of heat and combined heat-drought tolerance for breeding climate-resilient wheat cultivars [60]. These studies demonstrate the importance of breeding strategies in optimizing morphological traits for sustainable wheat production.

3.2. AI Application and Physiological Traits for Sustainable Wheat Yield

Photosynthetic efficiency in wheat is influenced by various mechanisms. Optimizing canopy light distribution (CLD) improves light utility and yield by altering the abundance of proteins associated with photosynthetic electron transport, redox state, and carbon–nitrogen assimilation. Nitrogen and water deficiencies both lead to reduced photosynthesis and grain yield, with water deficiency having a larger impact on photosynthetic proteins [61]. Ethylene plays a role in regulating photosynthetic efficiency under heat stress by modulating proline biosynthesis and the antioxidant system. Drought stress affects chlorophyll content and chlorophyll fluorescence parameters, and breeding has increased photosynthetic efficiency in newer wheat cultivars’ exogenously sourced nitric oxide (NO) and salicylic acid (SA) protect against heat stress by enhancing sulfur assimilation and reducing oxidative stress [62]. These findings provide insights into the molecular mechanisms underlying photosynthetic efficiency in wheat and can inform strategies for improving crop adaptation and stress tolerance.

Wheat varieties differ in their water use efficiency, and this trait is influenced by various factors. One study found that in a pot experiment, water and nitrogen uptake were more strongly associated with resource uptake availability than root size [44]. Another study identified loci associated with water use efficiency that were also linked to other traits such as grain carbon isotope discrimination, chlorophyll content, and plant height [63]. Additionally, differences in root traits between wheat cultivars were found to be robust to drought treatment and soil types, indicating the potential for genetic analysis of key root traits. Another study found that wheat varieties with higher yield potentials had different water use patterns and root morphology, which contributed to increased water use efficiency. Finally, a study on water-saving wheat breeding highlighted the importance of a new index called the water-saving index (WSI) in evaluating water-saving performance and identified specific markers that could be used in breeding for water-saving wheat varieties [64].

Efficient nutrient uptake in wheat is influenced by several physiological processes. Phosphorus (P) nutrition optimization has been shown to improve photosynthetic efficiency and nutrient uptake, leading to increased growth and performance in wheat [65]. Protein phosphorylation-based regulation of specific ion transporters or channel proteins plays a crucial role in mediating the uptake of ammonium, nitrate, phosphorus, and potassium ions by wheat roots [66]. Enhancing nitrogen use efficiency (NUE) in wheat involves factors such as transporter proteins, kinases, transcription factors, and micro RNAs, which participate in nitrogen uptake, assimilation, and remobilization processes. The HAK/KUP/KT gene family, including TaHAK13, acts as potassium transporters and contributes to potassium uptake and utilization in wheat [67]. The association of wheat roots with arbuscular mycorrhizal fungi (AMF) has been found to affect NUE by influencing N uptake efficiency, N remobilization efficiency, and overall grain N concentration [68]. Calcium plays a crucial role in enhancing physiological traits and stress tolerance in wheat plants. Exogenous calcium application has been shown to promote root development, increase photosynthetic pigments, induce the accumulation of osmolytes like proline and soluble sugars, and enhance antioxidant enzyme activities, such as superoxide dismutase, peroxidase, and catalase, which scavenge reactive oxygen species (ROS) [69]. Additionally, calcium acts as a universal second messenger, regulating plant responses to various stress factors, including drought and heat stress, by triggering primary physiological actions and activating stress tolerance genes [70]. The protective role of exogenous calcium in conferring better tolerance against induced stress conditions highlights its potential to improve wheat’s ability to withstand environmental stresses and ultimately enhance crop productivity. Magnesium transporters play a crucial role in facilitating magnesium uptake in wheat. Research has identified multiple magnesium transporter genes (MGT) in wheat, such as TaMGT1A, TaMGT1B, and TaMGT1D, which are involved in maintaining magnesium ion homeostasis [71]. These transporters are essential for magnesium mobilization from vacuoles to meet cytoplasmic demands, especially under a limited external magnesium supply [72]. Magnesium is vital for photosynthesis, with magnesium application stabilizing the net photosynthetic rate by maintaining Rubisco activation, thus mitigating yield losses caused by high-temperature stress during wheat grain filling. Additionally, magnesium deficiency affects wheat yield by reducing individual seed weight rather than seed number per spike, highlighting its role in regulating assimilate partitioning and protecting roots from oxidative stress and toxicity [73].

Environmental stresses such as drought and heat have a significant impact on physiological traits relevant to wheat breeding. These stresses can lead to reductions in biomass, grain yield, and thousand kernel weight [74]. However, they can also have positive effects on gluten parameters, resulting in a positive correlation between spike traits and gluten strength [75]. Under heat stress, certain wheat genotypes show remarkable stability of yield-related traits, while others exhibit increased spike fertility index and proline content. The response to combined stresses is more complex, with a significant shift in plant response and increased expression of stress-related genes. Abiotic stress also affects biochemical contents, physiological parameters, and plant–water relations, leading to reductions in agronomic traits. Overall, understanding the impact of environmental stresses on physiological traits is crucial for identifying stress-tolerant genotypes and developing breeding strategies for climate resilience in wheat.

Physiological traits play a crucial role in wheat breeding programs focused on sustainability. Understanding the impact of stress conditions, such as cold stress, drought stress, and heat stress on physiological and yield traits can help identify genotypes with tolerance to these stresses. For example, in the case of drought stress, genotypes that accumulate more potassium and proline in their shoots exhibit better tissue hydration and physiological functioning [76]. Similarly, under heat stress, changes in carbohydrate and antioxidant metabolism, as well as transpiration efficiency, are important for maintaining the balance between water-saving strategies and biomass production. These physiological traits can be used as surrogates to evaluate genetic variation and select heat-tolerant genotypes for breeding programs. By incorporating these traits into breeding efforts, it is possible to develop wheat cultivars that are better adapted to stress conditions, ensuring sustainable wheat production in the face of climate change.

3.3. AI Application and Biochemical Traits for Sustainable Yield

Wheat varieties have different nutritional profiles, particularly in terms of protein content. The protein content of wheat seeds is important for bread-making quality and grain nutritional value [77]. The seed-storage proteins in wheat, gliadins, and glutenins play a major role in determining bread-making quality. The protein content of wheat seeds can vary among genotypes and seasons. Wholegrain flour generally has a higher protein content compared to refined wheat flour. The protein content of different wheat varieties can be influenced by genetic factors as well as environmental factors. The availability of molecular markers can assist in breeding strategies to improve protein content and nutritional quality in wheat.

Wheat plants defend against diseases through various mechanisms, including the activation of basal defense mechanisms such as PAMP-triggered immunity (PTI) and the phenylpropanoid pathway [78]. These defense mechanisms involve the upregulation of genes encoding pathogenesis-related proteins (PR-proteins) and the accumulation of metabolites such as hydroxycinnamic acid amides and pipecolic acid. Additionally, wheat plants can exhibit nonhost resistance, which involves the recognition of pathogen effectors by disease-resistance genes and the suppression of innate immune responses. In response to stem rust infection, wheat plants undergo biochemical changes in the expression of metabolites, including flavonoids, carboxylic acids, and alkaloids, as part of their defense response [79]. Reactive oxygen species (ROS) and hypersensitive response (HR)-mediated cell death also play critical roles in wheat immunity to pathogens, with intraROS accumulation and localized cell death responses observed in infected wheat cells carrying nucleotide-binding leucine-rich repeat R genes.

Antioxidants play a crucial role in stress tolerance and resilience in wheat. They help maintain homeostasis by balancing the oxidative metabolism and antioxidant systems in response to environmental changes [80]. Exogenously applied antioxidants have been shown to improve plant growth and enhance stress tolerance in wheat. Different antioxidants, both chemical and natural, have been found to have similar effects on plant stress tolerance, photosynthesis, native antioxidant systems, and phytohormones. The enzymatic antioxidant system, including enzymes such as peroxidase, superoxide dismutase, catalase, and glutathione peroxidase, plays a significant role in scavenging reactive oxygen species (ROS) and maintaining redox homeostasis in wheat plants under stress conditions [81]. Additionally, the modulation of hormonal levels, such as abscisic acid (ABA) and gibberellic acid (GA3), has been observed in response to stress, contributing to the antioxidant defense system in wheat. Overall, antioxidants help protect wheat plants from oxidative stress and enhance their ability to tolerate and recover from various stressors.

Biochemical markers and pathways associated with desirable traits for sustainable wheat production have been identified in several studies. Genetic breeding techniques, such as transgenic breeding, molecular marker-assisted breeding, and gene editing, have been successfully used to improve wheat cultivars with multiple resistance traits, combating fungal diseases and herbivorous insects [58]. Drought tolerance is another important trait for sustainable wheat production, and studies have identified genes and proteins associated with drought tolerance in wheat. Additionally, a genome-wide association study (GWAS) identified markers linked to yield and drought tolerance attributes in wheat, providing insights into the genetic basis of these traits. Furthermore, a study on Indian bread wheat varieties identified a set of phenolic compounds that affect processing and nutrition quality, which can be improved through molecular breeding and functional genomics tools [82]. These findings highlight the potential for using biochemical markers and pathways to enhance desirable traits for sustainable wheat production.

Biochemical traits in wheat breeding can be targeted to enhance sustainability by considering non-targeted traits that influence the environmental performance of the crop. Wheat type and environment have significant effects on grain yield and quality traits, as well as the biofortification potential of the crop [83]. Breeding strategies should take into account the influence of non-targeted traits on the environmental performance and ecological sustainability of the crop. Additionally, marker-assisted selection and genomic selection can be used to enhance mineral uptake, translocation, and bioavailability, thereby increasing grain zinc and iron concentrations. By combining different layers of the canopy in physiological modeling, the understanding of wheat physiology can be improved, leading to better selection for radiation use efficiency and yield [84]. Overall, targeting biochemical traits in wheat breeding can contribute to the development of varieties with higher micronutrient levels, improved environmental performance, and increased yield sustainability.

3.4. AI-Application for Identifying Optimal Plant Breeding Hybrids, Parents, and Traits

AI algorithms are revolutionizing plant breeding by analyzing genetic and phenotypic data to identify optimal hybrids resilient to climate change impacts. The integration of AI, machine learning (ML), and deep learning techniques enables the efficient evaluation of crop cultivars for accelerated genotype-to-phenotype connections [85]. These technologies facilitate the differentiation of genotypes under various growth conditions, enhancing the selection of beneficial traits crucial for adapting to changing climates. AI’s computational power aids in processing big data from omics approaches, improving the understanding of complex biological mechanisms driving agronomic traits and gene functions [86]. By leveraging AI tools like support vector machines, neural networks, and genetic algorithms, breeders can analyze vast datasets to expedite the development of resilient plant varieties.

To select the most promising parent plants for breeding climate-resilient and high-yielding wheat varieties, various AI techniques are employed in modern breeding programs. These techniques include machine learning (ML) methods, such as neural networks, genetic algorithms, fuzzy logic, support vector machines, and K-nearest neighbor algorithms, which have been extensively utilized in agriculture for tasks like disease forecasting, climate prediction, and weed detection [87]. AI, coupled with high-throughput phenomics and genomics-assisted breeding (GAB), enables the identification of desirable traits and the acceleration of crop improvement by capturing non-linear interactions and epistatic relationships in genome-wide association studies (GWAS) and genomic selection (GS) [13]. By integrating AI with omics tools, breeders can rapidly identify genes; enhance accuracy in phenotyping, genotyping, and envirotyping data; and ultimately expedite crop-improvement programs to develop climate-resilient wheat varieties with high yields.

Artificial intelligence (AI) plays a crucial role in enhancing wheat breeding strategies by facilitating the identification and prediction of desirable traits like drought tolerance and disease resistance. AI techniques, including deep learning (DL) neural networks and machine learning algorithms, have been utilized to improve genomic prediction accuracy for unobserved phenotypes in wheat breeding [23,88].

4. Applications of Artificial Intelligence in Wheat Breeding Program

Artificial intelligence (AI) has been applied to address challenges in wheat breeding by developing software systems that utilize deep learning models and image processing algorithms to analyze wheat spikes and phenotypic traits [25]. These systems enable the detection of wheat spikes and the quantification of traits such as spike number per m² (SNpM2) using low-cost drones and canopy images. AI techniques have also been used to improve the accuracy of genomic prediction (GP) models in plant breeding research by incorporating deep learning neural networks that capture non-linear interactions and genotype–environment interactions [13]. Additionally, machine learning models have been developed to predict wheat yield by fusing genetic variants with multiple data sources collected by unmanned aerial systems, resulting in improved performance and prediction accuracy.

AI has been used to optimize various processes and aspects of wheat breeding. Machine and deep learning algorithms have been applied to complex traits in plants, such as grain yield and grain protein content, to improve prediction accuracies [89]. These AI models have been compared to traditional genomic best linear unbiased predictor (GBLUP) and Bayesian models and have shown superior performance in predicting traits. In addition, AI-powered software systems have been developed to enable the detection of wheat spikes and phenotypic analysis using wheat canopy images acquired by low-cost drones. These systems combine deep learning models and image processing algorithms to accurately quantify key yield components and assess wheat’s yield performance [89]. AI has also been used to model and forecast in vitro shoot regeneration outcomes of wheat, allowing for the development and optimization of genotype-independent regeneration protocols.

AI-driven wheat nutrient management has the potential to provide cost-effective and thorough nutrient assessment and decision-making solutions, improving agricultural productivity. Inoculation with plant growth-promoting rhizobacteria (PGPR) such as A. brasilense can enhance wheat biomass, grain yield, and nutrient uptake, leading to increased profitability [90]. Digital agriculture models that employ AI, such as neural networks and random forests, can detect crop problems early and reduce costs, with random forests providing explainable recommendations. AI technologies in agriculture, including machine learning models and autonomous machines, have the potential to improve crop management, plant phenotyping, and agronomic advice but require a thorough risk assessment to mitigate unintended socio-ecological consequences and security concerns. AI, if implemented ethically, can enhance sustainability in food systems by conserving resources, providing land-related services, and mitigating climate change, but issues such as carbon footprint and inequalities need to be addressed [91].

Wheat Nutrient Management While Breeding Using AI Technology

Artificial intelligence (AI) is utilized in several ways to optimize wheat nutrient management. Machine learning techniques are employed to identify factors that can improve nitrogen recommendations, such as soil nitrogen content, nitrogen fertilizer, and nitrogen deposition [92]. The variable fertilization rate (VFR) technique, combined with machine learning, allows for adapting fertilizer doses to crop needs, reducing nutrient losses. AI algorithms, such as particle swarm optimization (PSO) and back-propagation neural networks (BPNNs), are used to optimize nutrient solution control systems, adjusting control parameters for better system performance. Machine learning algorithms, including stochastic gradient boosting, are trained using various data sources, such as fertilization, pedoclimatic, and remote sensing data, to predict wheat yield [93].

AI-driven approaches have been employed to improve the shelf life of wheat products. One study developed CropQuant-Air, an AI-powered software system that uses deep learning models and image processing algorithms to detect wheat spikes and analyze phenotypic traits using canopy images acquired by drones [94]. Another study proposed a system to detect wheat rust disease and classify its infection types using deep learning classifiers, achieving high accuracy in identifying rust severity levels. These AI-driven approaches enable the accurate and timely detection of diseases and provide valuable tools for breeders, researchers, growers, and farmers. Figure 2 below details the iterative process of AI integration in wheat breeding, which is essential for the accurate and timely detection of diseases as discussed to assess crop yield performance and employ preventive measures [25].

5. Advanced AI Applications in Modern Wheat Breeding for Enhanced Phenotyping and Genetic Analysis

Artificial intelligence (AI) plays a crucial role in smart germplasm selection and CRISPR technology. AI technology is utilized in high-throughput phenotyping and gene functional analysis, enabling the rapid identification of genes driving agronomic trait formations [23].

Table 1. AI technology applications in wheat breeding. This table summarizes diverse tools like UAVs for aerial surveillance and machine learning for trait analysis in modern wheat breeding efforts.

Sr. No.	AI Technology/ Application	Programming Languages Used	Role in Wheat Breeding	Effects On
1	Machine Learning (ML)	Python, R	Enhances precision in predicting phenotypic traits and speeds up genomic selection processes.	Genomic Selection, Phenotypic Prediction
2	Deep Learning (DL)	Python, TensorFlow, PyTorch	Analyzes complex data for better prediction of genetic traits and forecasting yield outcomes.	Genomic Analysis, Yield Forecasting
3	Genomic Selection (GS)	Python, R	Accelerates the analysis of genomic data to select wheat varieties with optimal traits faster.	Genomics, Trait Selection
4	High-throughput Phenotyping	Python, MATLAB	Facilitates the rapid measurement of plant traits under varying environmental conditions.	Phenotypic Selection
5	LiDAR	C++, Python	Generates 3D models for structural analysis of wheat fields, aiding in accurate phenotypic assessments.	Structural Phenotyping
6	Hyperspectral Imaging Systems	Python, MATLAB	Assesses plant health and nutrient content, guiding precision interventions in wheat cultivation.	Nutrient Management, Plant Health
7	Unmanned Aerial Vehicles (UAVs)	Python, C++	Captures aerial images for monitoring crop health and growth stages, which is crucial for managing wheat breeding.	Crop Monitoring, Growth Stage Assessment
8	Automated Ground Robots	Python, C++	Collects detailed data on soil properties and plant health, reducing labor and enhancing data accuracy.	Soil Analysis, Plant Health
9	Neural Networks	Python, TensorFlow, PyTorch	Identifies patterns and anomalies in growth and stress responses, which is crucial for early intervention.	Pattern Recognition, Stress Response
10	Computer Vision Systems	Python, OpenCV	Processes images to detect disease and pests, supporting timely decisions in wheat breeding.	Disease Detection, Pest Management
11	Quantitative Trait Loci (QTL) Mapping	Python, R	Maps loci associated with important agronomic traits, aiding in the identification and selection of beneficial traits.	Trait Mapping, Genetic Analysis
12	Genotyping by Target Sequencing (GBTS)	Python, R	Enables rapid screening of genetic variants, improving the selection accuracy in breeding programs.	Genetic Screening, Variant Analysis
13	Image Processing Algorithms	Python, MATLAB, OpenCV	Analyzes imagery from various sources to assess crop traits and health, aiding in phenotypic selection.	Image Analysis, Phenotypic Assessment
14	Genomics-assisted Breeding (GAB)	Python, R	Utilizes genomic information to enhance the efficiency and accuracy of breeding decisions.	Genomic Information Utilization
15	CRISPR Technology	Python, R	Allows for precise genetic editing to develop wheat varieties with desired agronomic traits.	Genetic Editing, Trait Development
16	Particle Swarm Optimization (PSO)	Python, Java	Optimizes parameters in breeding simulations to achieve optimal outcomes in trait selection.	Simulation Optimization, Trait Selection
17	Back-propagation Neural Networks (BPNN)	Python, TensorFlow, PyTorch	Enhances prediction and classification accuracy in phenotypic and genomic data analysis.	Data Analysis, Phenotypic Classification
18	Random Forests	Python, R	Used for decision-making processes in selection based on complex datasets from wheat fields.	Decision-Making, Data Analysis
19	Gradient Boosting Machine (GBM)	Python, R	Improves prediction models for traits and yield based on historical data and simulations.	Trait Prediction, Yield Simulation
20	Support Vector Regression (SVR)	Python, R	Applies advanced regression techniques to predict wheat yields under varying conditions.	Yield Prediction, Regression Analysis
21	Fuzzy Inference Systems (FIS)	MATLAB, Python	Used for analyzing environmental data and making decisions in precision agriculture.	Environmental Analysis, Decision-Making
22	Geographical Information Systems (GIS)	Python, JavaScript	Manages and analyzes geographical data for land suitability and crop management in wheat breeding.	Land Suitability, Crop Management
23	Artificial Neural Networks (ANNs)	Python, TensorFlow, PyTorch	Simulates complex brain-like processing to enhance pattern recognition and decision-making in breeding.	Pattern Recognition, Decision Support
24	Molecular Simulations	Python, C++, Fortran	Simulates molecular interactions to understand and manipulate genetic factors in wheat.	Molecular Interaction Analysis
25	Genomic Prediction (GP)	Python, R	Predicts the performance of genotypes in different environments, enhancing selection accuracy.	Genotypic Performance, Environmental Adaptation
26	Genotype–Environment Interactions	Python, R	Studies the interaction effects to select genotypes that perform well across diverse conditions.	Environmental Adaptation, Genotype Selection
27	Big Data Analytics	Python, R	Analysis of large-scale genetic and phenotypic data.	Data-driven Decision-Making
28	Speed Breeding	Python, R	Accelerates breeding cycles.	Reduces Time to Develop New Cultivars
29	AutoML (Automated Machine Learning)	Python, TensorFlow, PyTorch	Automates ML model selection and optimization.	Enhances Model Accuracy and Efficiency
30	Next-Generation AI for Multi-Omics Data Integration	Python, R	Integrates diverse biological data for comprehensive analysis.	Improves Understanding of Complex Traits
31	High-Throughput Omics Technologies	Python, R	Analyzes omics data at high speed and accuracy.	Enhances Genomic Selection Processes
32	Unmanned Aerial Systems for Phenotyping	Python, C++	Captures aerial images for crop monitoring.	Improves Phenotypic Assessment
33	Federated Learning	Python, TensorFlow	Distributed learning across multiple data sources.	Enhances Model Robustness and Privacy

summarizes the applications of AI technologies in wheat breeding, highlighting their roles in phenotyping and gene analysis, as mentioned. Additionally, AI brings computational power and new tools for future wheat breeding, increasing the accuracy of high-throughput crop phenotyping, genotyping, and envirotyping data. Refer to Table 1 below for a comprehensive summary of how AI technologies are applied across different stages of wheat breeding for improved accuracy and efficiency. Deep learning (DL) models, a subset of AI, have been adopted for genomic selection (GS), revolutionizing the development of artificial intelligence systems [95]. DL models have surpassed human abilities in tasks such as classifying images and solving computer vision problems. Furthermore, AI techniques have opened doors to non-linear modeling approaches in crop breeding, capturing non-linear and epistatic interactions in genome-wide association studies (GWAS) and genomic selection (GS) [13], as detailed applications mentioned in Table 1. AI also plays a role in predicting the genotype of CRISPR-Cas9 gene editing products, enabling precise correction of disease-associated mutations.

AI revolutionizes molecular lab work, including DNA extraction, spectrophotometry, and PCR. See Table 1 for a detailed listing of AI applications in molecular lab work crucial for modern wheat breeding. Molecular simulations enhanced by AI can deliver new solutions and possibilities by introducing concepts and techniques from modern AI, such as differentiable programming and high-throughput simulations [96]. AI can also be integrated with microfluidic paper-based analytical devices (µPADs) to enhance the efficiency of PCR tests by forecasting the final output and trend of qPCR, as well as providing real-time quantitative analysis. An AI nanopore can predict the signature transmission functions of nucleotides, improving the accuracy and efficiency of solid-state nanopore-based DNA sequencing. Additionally, ML algorithms can be used to develop an AI autoverification system for laboratory testing, reducing the number of invalid reports and enhancing efficiency in the biochemistry laboratory [97].

6. Innovative Techniques for Sustainable Wheat Breeding

Innovative techniques and methodologies being employed to enhance sustainability in wheat breeding include the integration of genomics with participatory varietal selection to capture farmers’ traditional knowledge and inform breeding decisions. Additionally, recent progress in sequencing and high-throughput genomics technologies has enabled the identification of functional alleles, haplotypes, and candidate genes for stress tolerance in wheat, allowing for the development of marker-assisted breeding and genome editing [10]. The use of omics tools and large volumes of data from crop breeding programs can be integrated to predict wheat performance under different climate change scenarios, leading to the design and delivery of future wheat ideotypes for yield improvement. Molecular breeding techniques such as marker-assisted selection, quantitative trait loci mapping, genome-wide association studies, and the CRISPR/Cas-9 system have also been successful in developing disease resistance in wheat [58].

AI technology plays a crucial role in marker development and genotyping, offering new opportunities for sustainable wheat breeding. AI tools and strategies provide computational power and enable the rapid identification of genes, accelerating crop improvement programs [23]. Advances in genotyping by target sequencing (GBTS) technology have led to the development of high-throughput, convenient, reliable, and cost-efficient genotyping platforms, allowing breeders to quickly screen germplasm resources and parental breeding materials for excellent allelic variants. Furthermore, AI can be utilized in genome-based breeding to achieve progressive genetic gains in wheat. It enables the use of analytical approaches such as environmental genome-wide association studies, haplotype-based analyses, and genome-based prediction of heterosis patterns, leading to the development of high-yielding, climate-resilient wheat cultivars with high nutritional quality [27].

AI plays a crucial role in greenhouse experiments and field trials for sustainable wheat breeding. Figure 3 shows the complex interconnections between AI-driven technologies used in greenhouse experiments and field trials. It enables high-throughput phenotyping and gene functional analysis, allowing researchers to evaluate important agronomic traits in larger-sized germplasm at reduced time intervals in the early growth stages. AI-powered software systems, such as CropQuant-Air, utilize deep learning models and image processing algorithms to detect wheat spikes and analyze their phenotypic traits using low-cost drones, providing a cost-effective approach for assessing crop yield performance [98]. Additionally, AI-based models integrated with IoT technologies can automate greenhouse environment-related activities, optimizing resource utilization and reducing energy consumption. Figure 3 maps out how AI-driven technologies, including IoT, are interconnected to optimize resource utilization in greenhouse environments while maintaining the desired climate for maximizing plant production [25].

Innovative techniques such as marker-assisted selection (MAS), quantitative trait loci (QTL) mapping, genome-wide association studies (GWAS), and the CRISPR/Cas-9 system have significantly contributed to the overall sustainability and efficiency of wheat breeding programs. These techniques have allowed breeders to develop wheat cultivars with improved disease resistance, stress tolerance, and agronomic traits. As depicted in Figure 2, the flowchart of AI integration highlights how innovative techniques are applied to develop cultivars with improved traits, leading to increased crop productivity and reduced yield losses [99]. Molecular markers, such as SCAR, RAPD, SSR, and SNP, have been used to identify and select desirable traits in wheat, enabling breeders to develop cultivars with broad-spectrum disease resistance [58]. The use of MAS, QTL mapping, and GWAS has facilitated the identification and introgression of beneficial genes into wheat breeding programs, resulting in the development of high-yielding and stress-tolerant cultivars. Additionally, the CRISPR/Cas-9 system has provided a powerful tool for precise gene editing, allowing breeders to modify specific genes associated with disease resistance and stress tolerance in wheat. These innovative techniques have enhanced the efficiency and sustainability of wheat breeding programs by accelerating the development of improved cultivars with desirable traits.

AI-Powered Marker Discovery in Wheat Breeding

AI technology facilitates the identification and utilization of genetic markers in wheat breeding programs by providing computational power and new tools for analysis. Refer to Figure 3 for a visual representation of how different AI technologies interact to facilitate genetic marker identification in wheat breeding. It helps in solving the challenges of high-throughput phenotyping and gene functional analysis, as well as linking genotype to phenotype [23]. Marker-assisted breeding, facilitated by AI, can overcome the drawbacks of anomalous linkage disequilibrium and improve the physical alignment of markers, leading to the detection of selective sweeps for important agronomic traits. AI algorithms, such as machine learning and deep learning, can optimize multitrait models for predicting grain yield and grain protein content in wheat, resulting in increased prediction accuracy. Additionally, AI-based intelligent artificial climate chambers provide a low-cost and site-independent facility for studying wheat cultivation and phenotype acquisition, enabling breeders to develop excellent germplasm varieties [27]. DL neural networks, a type of AI, have been developed to increase the genomic prediction accuracy of unobserved phenotypes while accounting for genotype–environment interaction.

Different types of genetic markers commonly used in wheat breeding include sequence tagged site (STS), simple sequence repeat (SSR), genotyping by sequencing (GBS), single-nucleotide polymorphism (SNP) arrays, exome capture, Kompetitive Allele-Specific PCR (KASP), cleaved amplified polymorphic sequence (CAPS), semi-thermal asymmetric reverse PCR (STARP), and genotyping by target sequencing (GBTS) [100]. These markers are applied in marker-assisted selection (MAS) in wheat breeding programs. As illustrated in Figure 3, see how AI-driven technologies interact in the application of markers in MAS. MAS is a time-saving, cost-effective, and goal-oriented approach that uses molecular markers to select plants with desired traits. MAS can be used to identify and select plants with genes associated with valuable traits such as disease resistance, drought tolerance, and grain quality [101].

AI has been successfully integrated into wheat breeding in various case studies around the world. One notable case study is the work by Chen et al. [25], who developed CropQuant-Air, an AI-powered software system that uses deep learning models and image processing algorithms to detect wheat spikes and analyze phenotypic traits using low-cost drones. Togninalli et al. [15] proposed a machine learning model that combines genotype and phenotype data collected by unmanned aerial systems to predict wheat yield, achieving significant improvements in prediction accuracy compared to conventional methods. Montesinos-López et al. compared a novel deep learning method with conventional genomic prediction models and found that the deep learning approach provided better accuracy in certain scenarios, highlighting its potential for integrating genomics and phenomics in wheat breeding [88].

AI has shown great potential in addressing challenges and improving outcomes in wheat breeding. This study proposed a machine learning model that combines genotype and phenotype data collected by unmanned aerial systems to predict yield with improved accuracy [15]. A different approach used deep learning neural networks to increase the accuracy of genomic prediction by linking genomics with phenomics, specifically imaging. Additionally, a study developed a neural network capable of analyzing cell phone camera images to phenotype Fusarium-damaged kernels, providing an automated and objective method for assessing resistance to Fusarium head blight. These case studies demonstrate how AI can enable large-scale phenotyping, enhance yield prediction, improve breeding programs, and provide efficient and objective phenotyping methods, ultimately contributing to the development of improved wheat varieties.

7. Artificial Intelligence in Multi-Year Evaluation and Enhanced Data Analysis

AI-enabled multi-year evaluation in wheat breeding provides valuable insights for improving grain yield and predicting crop suitability. The use of deep learning neural networks in genomic prediction models enhances the accuracy of unobserved phenotypes, accounting for genotype–environment interactions and linking genomics with phenomics. These models have been shown to outperform conventional models such as GBLUP, gradient boosting machine (GBM), and support vector regression (SVR) in terms of prediction accuracy [27]. Additionally, multi-trait models have been found to be more effective than single-trait models in predicting grain yield, especially when there is a high genetic correlation between traits. AI tools such as fuzzy inference systems (FIS) and geographical information systems (GIS) have also been used to evaluate wheat suitability for cultivation, providing quantitative findings that can aid decision-makers in improving food security.

AI enhances data analysis in wheat breeding programs by leveraging machine learning and deep learning algorithms to improve prediction accuracies and enable more informed decision-making. These algorithms utilize genotype and phenotype data, as well as data collected by unmanned aerial systems, to develop models that can predict yield and other performance traits of wheat varieties. By fusing genetic variants with multiple data sources, such as spectral and texture features, AI models can provide more accurate predictions of grain yield and protein content. These predictions help breeders and researchers assess wheat’s performance under complex field conditions, enabling them to select and develop improved varieties more efficiently. AI-based models also offer the advantage of interpretability, shedding light on the importance given to each input during prediction. Overall, AI-powered data analysis in wheat breeding programs promises to accelerate variety development and enhance global food security [28].

8. Challenges, Limitations, and Future Prospects

The major challenges and limitations faced in the integration of AI in wheat breeding include the complexity of phenotypic analysis of big image data, conducting large-scale phenotyping in an automated manner [25], the need for validated and usable ways to integrate and compare large, multi-dimensional datasets, the understanding of complex mechanisms behind genes driving agronomic-trait formations [23], and the high construction and maintenance costs of existing phenotyping platforms. Additionally, statistical and software challenges persist in AI-based models, and there is a need to link genotype to phenotype more effectively. These challenges hinder the optimal application of high-throughput phenotyping, genomics, and enviromics in wheat breeding. However, recent advances in AI techniques, such as deep learning models and image processing algorithms, provide promising solutions for automating phenotypic analysis and improving crop breeding efficiency.

AI has the potential to revolutionize sustainable wheat breeding by enhancing the efficiency and precision of crop improvement programs. Refer to Figure 2 for a visual explanation of how AI technologies are revolutionizing wheat breeding through enhanced efficiency and precision. The integration of AI with high-throughput phenomics and genomics-assisted breeding (GAB) can expedite the development of high-yielding stress-tolerant crop cultivars. As shown in Figure 2, the integration process of AI technologies into phenomics and genomics is critical for developing advanced wheat cultivars. AI techniques, such as deep learning models and image processing algorithms, can enable the analysis of large-scale phenotyping data and the detection of key yield components in wheat. By capturing non-linear and epistatic interactions in genome-wide association studies (GWAS) and genomic selection (GS), AI-based models can provide a more accurate understanding of complex traits. Figure 2 illustrates the AI integration flowchart, showing its role in capturing complex genetic interactions in GWAS and GS and facilitating rapid gene identification [25]. Furthermore, the combination of AI with speed breeding techniques can significantly reduce the time required for conventional breeding, ensuring greater efficiency and accuracy in crop cultivar development. Overall, the future prospects of utilizing AI in sustainable wheat breeding are promising, offering opportunities for faster and more effective crop improvement.

Integrating AI in wheat breeding presents ethical considerations and potential social impacts that need to be addressed. Ethical challenges in digital agriculture, including fairness, transparency, accountability, sustainability, privacy, and robustness, may arise when using AI in farming [102]. Social impacts encompass power asymmetry between farmers and agricultural technology providers (ATPs), affecting farmers negatively and highlighting the importance of integrating AI assurance methods for more farmer control over data decision-making. To address these issues, it is crucial to engage a wide range of stakeholders in setting trajectories for AI in agriculture, incorporate stakeholder views into responsive practices like standards and codes of conduct, and increase public debate while creating laws and regulations to regulate AI use in wheat breeding [103].

Interdisciplinary collaborations between plant scientists, AI researchers, and agronomists can be fostered to drive innovation in sustainable wheat breeding by adopting an approach that integrates AI technologies with “omics” tools for high-throughput phenotyping and genotyping, enabling rapid gene identification and accelerating crop improvement programs [103]. This collaboration should also focus on developing high-throughput genomic tools, such as SNP arrays and molecular marker maps, to analyze genetic diversity, identify genomic regions associated with agronomic traits, and introduce novel genetic diversity into elite varieties. Additionally, the involvement of multiple actors from various fields of life, including scientists, policymakers, educators, and industry, is crucial for taking a systemic outlook to change and restructuring social meanings, consumer behavior, and business models in the agri-food sector to drive sustainability transitions [104]. By embracing an interdisciplinary approach, these stakeholders can work together to create robust, economically valuable, and socially desirable AI-powered systems in agriculture, leading to greater acceptance and trust among farmers when utilizing them.

To support the adoption of AI technologies in wheat breeding, regulatory and policy frameworks must be established to address privacy concerns, ensure safe and responsible use, and promote innovation. The existing regulatory gaps and the complexity of technology necessitate a comprehensive approach to AI regulation [105]. It is crucial to minimize risks to public safety, preserve human rights, and enable an innovative environment through new frameworks and institutional arrangements for governing AI technologies. A pragmatic approach to providing a technology assurance regulatory framework has been proposed, emphasizing the need for voluntary participation and regulation only in critical areas to avoid stifling innovation while ensuring necessary assurances [106]. By implementing suitable regulatory frameworks that balance privacy protection, safety, and innovation, the adoption of AI technologies in wheat breeding can be effectively supported.

AI technologies can be instrumental in ensuring long-term sustainability and resilience in wheat breeding programs by leveraging big data analytics and advanced computational power. By integrating AI with “omics” tools, rapid gene identification can be achieved, accelerating crop improvement programs [107]. Additionally, the use of AI in agriculture can help address challenges such as unpredictable climate change and food insecurity, leading to sustainable farming practices. Through the application of AI in high-throughput phenotyping and genotyping, the accuracy of crop trait evaluations can be significantly enhanced, facilitating the identification of superior genotypes for breeding programs [108]. Furthermore, the combination of human intelligence with machine-based algorithmic intelligence, known as hybrid intelligence, can support advanced human–system integration in agricultural practices, contributing to resilience and sustainability in wheat breeding.

To increase public and farmer engagement and acceptance of AI-driven innovations in wheat breeding, it is crucial to focus on transparency, accountability, and education. Transparency in AI model development and decision-making processes, as suggested, can help build trust among farmers and the public. Assigning clear responsibility and accountability to AI decisions, as proposed, can also enhance acceptance. Moreover, educating farmers and the public about the benefits and risks of AI in agriculture, as highlighted, can lead to better understanding and increased engagement. Additionally, incorporating interdisciplinary collaboration in the development of AI solutions, as recommended, can ensure robust, economically valuable, and socially desirable innovations, fostering greater acceptance and trust among farmers [109].

9. Conclusions

This review paper examines the integration of artificial intelligence (AI) into wheat breeding, showcasing its revolutionary impact on the breeding of climate-resilient, high-yielding wheat cultivars. The function of AI is crucial in capturing complex genetic interactions and accelerating the correlation between phenotype and genotype through the use of advanced analytical instruments and models. This integration addresses important challenges presented by climate change and the increasing worldwide need for food. Artificial intelligence (AI) technologies, including machine learning, deep learning, and high-throughput phenotyping, enable fast and accurate analysis of large genetic and environmental datasets, improving the breeding process. Integrating AI into several aspects of wheat production, including genetic analysis and nutritional management, greatly enhances efficiency, precision, and sustainability. These methods facilitate the creation of wheat varieties that possess enhanced stress tolerance, resistance to diseases, and potential for higher yields. AI integration enhances wheat breeding processes by offering novel tools and methodologies for efficient analysis of high-throughput phenotyping and genotyping data. The integration of AI technology and “omics” techniques enables rapid identification of genes and the advancement of agricultural types. Table 1 provides a detailed summary of AI applications in the integration with ‘omics’ techniques, facilitating rapid gene identification. AI approaches facilitate the identification and analysis of non-linear and epistatic interactions in genome-wide association studies (GWAS) and genomic selection (GS), hence improving the precision and effectiveness of crop enhancement. Refer to Table 1 to understand how AI is used to improve the precision and effectiveness of crop enhancement through the analysis of non-linear and epistatic interactions. Furthermore, artificial intelligence (AI)-based models have the ability to go beyond the limitations of linear models when it comes to comprehending the complex relationships within large sets of data. This, in turn, enhances our understanding of the fundamental processes driving the production of agronomic traits. Utilizing AI alongside high-throughput phenomics and genomics-assisted breeding (GAB) has the potential to completely transform agricultural breeding methods. This can result in the creation of stress-tolerant crop varieties with improved efficiency and accuracy. By combining AI technology with environmental monitoring sensors and regulating equipment, intelligent artificial climate chambers offer a cost-effective and location-independent solution for cultivating crops and collecting phenotypic data. This, in turn, speeds up the wheat breeding processes. AI is used in several ways to improve agricultural operations, including precision farming and crop management. This helps to make agriculture more sustainable by reducing environmental impacts and making better use of resources. Utilizing AI tools strategically aids in predicting and minimizing the impact of unfavorable weather conditions, hence ensuring food security. The widespread use of AI in agriculture is motivated by its capacity to transform crop breeding and management, offering solutions to urgent global concerns. As the challenges of climate change increase, the role of AI becomes essential in improving agricultural output and sustainability. Adopting AI in wheat breeding is not only beneficial but also necessary to safeguard global food supplies in the face of climate change and population growth. The crucial importance of AI in sustainable agriculture is shown by its capacity to boost genetic comprehension and enhance crop resilience and output. Advanced data analytics and machine learning approaches can now be integrated with the help of artificial intelligence (AI). This significantly speeds up the breeding process by facilitating rapid assessment of large genomic and phenotypic datasets. With the use of AI-powered models, breeders can choose the most resilient and competitive genotypes by accurately predicting how plants will respond to different environmental factors. The use of AI in genomics-assisted breeding (GAB) has proven to be more predictive than conventional techniques, ensuring that newly developed wheat cultivars would be resilient to the effects of climate change.

Furthermore, precise, environmentally friendly evaluation of plant characteristics under stress is made possible by AI-driven high-throughput phenotyping tools, which makes it easier to identify genotypes that perform effectively in a range of environments. These innovations improve our capacity for monitoring and assessing crop growth in real time, offering vital information that directs the creation of wheat cultivars that are both highly productive and stress-tolerant. We can maximize the utilization of resources and enhance crop management techniques by integrating artificial intelligence (AI) with automated data collecting and environmental monitoring. This will ultimately strengthen the sustainability and resilience of the world’s wheat production systems.