Accurate disease classification of tomato leaf diseases is crucial for reducing yield losses and supporting sustainable agricultural practices. This study proposes a decision-level integration approach for the comprehensive assessment of phytosanitary status and soil and environmental conditions in tomato crops, combining EfficientNetB0-based disease classification with Random Forest-based analysis of edaphic and environmental variables. The proposed framework incorporates a deep learning model based on EfficientNetB0, fine-tuned to classify eleven tomato leaf disease classes using 38,500 annotated images, alongside a Random Forest model trained on 2,880 records comprising soil pH, electrical conductivity, nitrogen (N), phosphorus (P), and potassium (K), soil moisture and temperature, as well as environmental and solar radiation variables collected under controlled greenhouse conditions. The outputs of both models are integrated at the decision level using a qualitative decision-level integration matrix, enabling the identification of relationships between disease incidence and soil properties, categorized into three qualitative levels: Deficient, Optimal, and Excess. The deep learning model achieved high performance, with precision and recall values above 97% and an F1-score exceeding 98%, exhibiting minimal confusion among visually similar disease classes. The Random Forest model reached an overall 93.7% of precision demonstrating stable and balanced classification across the three qualitative levels. Furthermore, qualitative integration matrix validated the integrated framework by revealing consistent associations between soil properties categories and specific leaf diseases. Overall, the results show that decision-level integration of visual and physiological data improves the system’s interpretability and enhances its ability to anticipate plant health risks, making the proposed approach an effective decision-support tool for precision agriculture.
Early and accurate detection of strawberry diseases is important for efficient crop management, yet overlapping visual symptoms can limit image-only classifiers. This study evaluates a hybrid convolutional neural network-Graph Attention Network (CNN-GAT) framework for four disease classes: angular leaf spot, blossom blight, powdery mildew, and leaf spot. A pre-trained ResNet18 backbone generates 128-dimensional image embeddings, which are connected through a sparse top-K cosine-similarity graph (K = 5). A GAT performs attention-weighted neighbourhood aggregation before multi-layer perceptron classification. The proposed model achieved 94% overall accuracy, compared with 89% for ResNet18 and 91% for EfficientNet under the reported experimental setting. Class-wise precision, recall, and F1-scores were balanced, with the highest F1-score for powdery mildew. Confusion-matrix analysis, t-SNE visualization, and ablation results support the contribution of relational refinement, particularly for visually similar disease classes. The source dataset consists mainly of mobile-phone images acquired under natural illumination in South Korean greenhouses, with approximately 20% drawn from public online sources. Because the archived experiment does not provide a separately verified strict-inductive graph evaluation or controlled perturbation testing, the 94% graph result is interpreted conservatively as transductive. Independent field, cross-device, and strict-inductive validation remain necessary before deployment-level robustness can be claimed.
Stomata are pivotal regulators of plant carbon-water balance, yet their complex, nonlinear dynamics—governed by intricate networks of ion transporters, signaling molecules, and metabolic reactions within guard cells—have long resisted intuitive understanding and rational manipulation. Here, we present a perspective on the development and application of a "digital twin" for stomatal guard cells, centered on the OnGuard family of quantitative systems models. This platform integrates decades of molecular, biophysical, and kinetic data into a rigorous computational framework, creating a "virtual proving ground" where environmental conditions can be simulated and genetic perturbations tested in silico. By encoding the fundamental constraints of mass and charge conservation alongside experimentally derived transporter kinetics, the model generates emergent behaviors that recapitulate complex physiological responses and, crucially, yields experimentally validated predictions that uncover previously hidden regulatory nodes. We discuss how this iterative cycle of prediction and validation has revealed unexpected homeostatic networks and context-dependent transporter functions. Finally, we argue that extending such modeling approaches to incorporate larger-scale physiological modules—including mesophyll photosynthesis, leaf hydraulics, and canopy-scale fluxes—will enable true multi-scale global simulation analysis, transforming our ability to predict and design crop performance under real-world conditions
The negative effects of climate change on wheat production put the world's food security in danger, and rising temperatures are predicted to have an increasingly negative impact on sustained wheat productivity. A comprehensive understanding of metabolome alterations will clarify the metabolic processes that govern environmental adaptation in wheat. Therefore, metabolome analysis can aid in the more precise selection of genotypes that can withstand stress in crop breeding. This study was carried out to determine the metabolite profiles of two popular Indian wheat genotypes under control and heat stress conditions. A total of 51 known metabolites were accumulated as determined from the chromatogram after the GC-MS analysis. Metabolites such as Vitamin E, acid, Octadecanamide, and Tetracontane were highly upregulated in the heat-treated samples. Besides, there was a high accumulation of some metabolites, such as Fumaric acid, tocopherol, and palmitaldehyde in the heat-treated Raj 3765, a well-known heat stress-tolerant wheat variety in India. The research found 12 and 8 upregulating metabolites in Raj 3765 and HD 2967, respectively. Metabolic pathway analysis revealed significant activities in the steroids and arginine biosynthesis, TCA cycle and pyruvate metabolism. Further experimental validation through targeted quantification and functional assays will be required to confirm the causal role of these metabolites in heat tolerance. These results, we believe, may provide useful metabolic indicators associated with stress in wheat, which might be employed as rapid and focused diagnostic instruments to identify germplasm that performs better at elevated temperatures.
Translating QTL and GWAS signals into functionally validated gene targets remains a critical bottleneck in crop improvement, particularly in complex polyploid genomes such as Nicotiana tabacum (∼4.3 Gb, 2n = 48). To address this, we systematically characterised 15 candidate genes nominated for tobacco leaf chemistry traits, including nicotine, nornicotine, total sugar, reducing sugar, fructose, and sugar–alkali ratio. The candidate proteins, spanning 210–881 amino acids across seven subcellular compartments, exhibited functionally interpretable domain architectures, including bHLH-MYC, SnRK2 kinase, UDP-glycosyltransferase GT1, ALMT, and RING finger domains, in 10 of 15 proteins, establishing mechanistic hypotheses aligned with associated traits. Structural integrity was further supported by AlphaFold3 modelling, confirming high-confidence folds with binding pockets co-localised within conserved functional domains. At the regulatory level, promoter analysis revealed a universal TGACG-motif and ABRE elements across all candidates, while the auxin-responsive AuxRR-core was exclusive to the alkaloid-associated gene Nt16g00236, suggesting trait-specific transcriptional control. Haplotype analysis reinforced this specificity, identifying significant allelic associations in eight candidate genes, including five genes on chromosome 8 and three genes on chromosome 16. Directional antagonism was observed at the chromosome 8 loci, where nicotine- and nornicotine-elevating haplotypes coincided with reduced sugar–alkali ratio, indicative of pleiotropic carbon–nitrogen partitioning. Complementing these findings, tissue-specific and stress-responsive expression profiling revealed shoot- and root-preferential patterns consistent with trait biology, temporally distinct drought-induction waves, and a cold-induced root-to-shoot expression switch in Nt22g03479, all corroborated by qRT-PCR validation. Collectively, these findings establish a multi-evidence functional framework for candidate gene prioritisation, advancing marker-assisted breeding and targeted experimental validation in tobacco improvement.
Mungbean is a short-duration grain legume valued for its high protein content, biological nitrogen fixation capacity, and adaptability to low-input farming systems. Despite its importance for food and nutritional security, its productivity remains below its genetic potential due to a narrow genetic base, susceptibility to multiple biotic and abiotic stresses, and a complex trait architecture influenced by genotype-by-environment interactions. These constraints are further intensified by increasing climate variability, which exacerbates yield instability across diverse agro-ecological regions. Recent advances in genomics have transformed mungbean breeding by enabling the identification of genes and genomic regions associated with agronomic and stress-resilient traits. However, the translation of these discoveries into routine breeding applications remains limited. This review provides a critical synthesis of current progress in mungbean improvement, with an emphasis on evaluating the strengths, limitations, and practical applicability of key genomic tools, including quantitative trait locus (QTL) mapping, genome-wide association studies (GWAS), marker-assisted selection (MAS), and genomic selection (GS). While QTL mapping and GWAS are effective for detecting trait-associated loci, their predictive efficiency is often constrained by population structure, environmental interactions, and limited effect sizes. In contrast, GS offers greater potential for improving complex polygenic traits and enhancing genetic gain across environments, although it requires large training populations and high-quality phenotypic datasets. The concept of climate resilience is broadened to include multi-stress interactions, environmental variability, and genotype stability across environments. The integration of genomic tools with physiological screening, high-throughput phenotyping, and multi-environment trials is essential for developing stable cultivars. This review proposes a climate-smart, data-driven breeding framework that integrates conventional and modern approaches to accelerate the development of high-yielding, nutritionally enhanced, and climate-resilient mungbean varieties for sustainable agriculture.
Achieving uniform seed placement and maximizing wheat yield require an integrated consideration of key operational factors, including tillage method, crop residue level, and forward speed. Although the individual effects of these factors have been widely investigated, their interactive impacts on drill performance under field conditions remain insufficiently understood. Therefore, this study aimed to evaluate the combined effects of tillage systems, residue coverage, and forward speeds on wheat planting quality and grain yield. A field experiment was conducted at the Research Farm of the University of Agricultural Sciences and Natural Resources of Khuzestan using a split-split plot arrangement within a randomized complete block design (RCBD) with three replications. Treatments consisted of three tillage methods-conventional (moldboard plow + disc), conservation (combined tillage), and conservation (chisel packer)-three residue levels (0%, 40%, and 80%), and three forward speeds (4, 7, and 10 km h(-1)). The results showed that tillage method and residue level significantly affected planting depth uniformity, seed distribution uniformity (p < 0.05), and uncut residue (p < 0.01). In addition, the interaction of tillage, residue level, and forward speed had a significant effect on grain yield (p < 0.01). Combined tillage achieved the highest planting depth uniformity (87.66%) and grain yield (7326 kg ha(-1)). The 40% residue level resulted in superior yield performance compared to 0% and 80% residue levels. Moreover, the lowest forward speed (4 km h(-1)) produced the highest average grain yield (6990 kg ha(-1)). Overall, the findings suggest that optimizing wheat planting performance requires the integrated management of tillage practices, residue retention, and operational speed. Such optimization can enhance seeding precision and improve crop productivity under conservation agriculture systems.
The effective management of agricultural data is critical for enhancing transparency and security in modern farming, yet traditional centralized systems face risks of data tampering more than ever before., high storage costs, and insufficient trust in multi-party collaboration. This study proposes a blockchain and IoT-integrated system that introduces two key innovations: (1) a multi-role trust mechanism enforced by smart contracts—where reporters, validators, and administrators interact through cryptographically verifiable workflows, eliminating reliance on any central authority; and (2) a hybrid on-chain/off-chain storage architecture combining Ethereum with the InterPlanetary File System (IPFS), which reduces storage costs by over 99.9% compared to pure on-chain storage while preserving data immutability. This system bridges IoT-driven data collection and blockchain-enabled trust verification and offers a tamper-resistant platform for agricultural workflows. A citrus orchard case study is used in this paper to validate its effectiveness in real-time monitoring, disease detection, and automated task management. However, limitations in scalability and farmer usability highlight the needs for distributed optimization and user-centric interfaces.
Continuous monocropping often leads to progressive soil functional degradation and yield decline, yet the mechanisms by which severe nutrient depletion and stoichiometric imbalance reshape microbial functional reorganization in the rhizosphere remain poorly understood. Here, using a pot-based consecutive cropping system of Coix lacryma-jobi as a model, we integrated soil metabolomics, metagenomics, and physicochemical analyses to elucidate the assembly and successional trajectories of the rhizosphere ecosystem. Our results show that severe nutrient depletion—characterized by the marked decline of bioavailable phosphorus and potassium—was a major stoichiometric constraint associated with microbial succession, rather than soil acidification or autotoxic effects alone. This multidimensional nutrient limitation was accompanied by a fundamental shift in microbial functional strategies from nutrient acquisition to resource conservation. Specifically, microbial functional convergence increased, characterized by the upregulation of membrane transport systems and nitrogen metabolism pathways. Network topology analysis further revealed that Mesorhizobium emerged under nutrient stress as a key hub taxon coordinating microbial necromass recycling. This hub taxon was tightly coupled with glucosamine, a marker of microbial cell wall turnover, and membrane transport-related functional modules, indicating a progressive transition of community metabolism toward necromass scavenging as a means to sustain function under nutrient-depleted conditions. Collectively, our findings show that consecutive cropping forces the rhizosphere ecosystem into a necromass-recycling-oriented adaptive state, providing a mechanistic framework for understanding how nutrient depletion and stoichiometric imbalance shape microbial functional evolution in agricultural systems.
Genetic diversity is crucial for breeding program to develop improved varieties. Estimating genetic diversity and population structure among genotypes of a given population is compulsory to select the most divergent parents in breeding programs. This study evaluated 144 sugarcane genotypes using 20 microsatellite markers. The markers produced 482 alleles, and 88.59% (434) of the alleles were polymorphic. The number of alleles per locus ranged from 14 to 40, with a mean of 24.1, while polymorphic information content values ranged from 0.42 to 0.92, with an average value of 0.67 for all markers evaluated. In this study, high genetic diversity within the population was recorded with an average value of number alleles of 1.75, effective alleles of 1.53, Shannon's information index of 0.45, Nei's gene diversity of 0.30, and polymorphic percentage loci of 79.25%. Among the populations, India showed the highest observed number of alleles (1.86), the effective number of alleles (1.60), Shannon's information index (0.49), Nei's gene diversity (0.34), and unexpected heterozygosity (0.34). Thus, the genotypes from India could be a source of divergent parents to use in sugarcane breeding programs. Cluster analysis using the UPGMA method showed four main clusters, further grouped into eight sub-clusters. Population structure analysis also identified six subpopulations within the genotype set. AMOVA analysis revealed that 96% and 4% of the total variability was attributed to within- and between-population variation, respectively. The principal coordinate analysis showed that the distribution of genotypes in the scatter plot was considerably dispersed, indicating a wide genetic diversity between genotypes originating from different populations. The overall results offer valuable information for future genetic study of the sugarcane genotypes and better utilization of sugarcane germplasm resources in sugarcane breeding.
Global warming poses a critical threat to rice production, with elevated temperatures at the reproductive stage causing irreversible yield losses. The premium aromatic cultivar 'Khao Dawk Mali 105' ('KDML105') is widely cultivated in Thailand but is highly heat-sensitive. To assess its vulnerability to high temperuare, rice plants were grown under controlled environments at 32 degrees C, 36 degrees C, and 40 degrees C from panicle initiation to maturity. Daytime warming >= 36 degrees C delayed flowering, shortened culm and panicle length, and inhibited uppermost internode elongation, thereby resulting in incomplete panicle exsertion. Anther development was strongly affected, with reduced size, collapse of structural integrity, and poor pollen release at 36 degrees C, progressing to deformation and complete sterility at 40 degrees C. Spikelet fertility declined sharply with rising temperature, culminating in 100% sterility at 40 degrees C, confirming the narrow thermal threshold of 'KDML105'. At the molecular level, elevated temperature downregulated proline-GABA pathway genes (OsP5cs, OsProDH, OsOat, Osbadh2), suppressed biosynthesis of 2-acetyl-1-pyrroline (2-AP), and increased volatilization losses, thereby diminishing aroma. Together, these results show that long-term elevated daytime temperature simultaneously compromises reproductive success and fragrance quality. Panicle traits and anther morphology serve as sensitive indicators for screening heat tolerance, while targeted breeding and adaptive agronomic strategies are essential to protect yield and grain quality in aromatic rice under future climate scenarios.
The agricultural sector faces mounting challenges from climate change and crop losses caused by biotic and abiotic stresses, with traditional breeding and chemical controls offering limited solutions. Advances in artificial intelligence (AI) have enabled the de novo design of protein binders exhibiting high specificity and stability, presenting new opportunities for crop protection and stress tolerance. Here we review core AI-driven methodologies—including diffusion models and protein language models—that facilitate the rational design of mini-proteins targeting pathogen effectors and plant immune components, demonstrating potential for precise molecular interventions to enhance disease resistance and abiotic stress resilience. These approaches have the potential to simplify or navigate differently the regulatory complexities often associated with transgenic organisms. While traditional delivery systems still facing a critical challenge, emerging cell-penetrating peptide (CPP) systems offer a non-viral, targeted solution for protein translocation, potentially simplifying regulatory pathways by enabling transient expression without genomic integration. While promising, successful agricultural application requires expanding plant-specific structural data and developing compatible delivery systems. This emerging paradigm integrates computational protein design with plant biotechnology, offering a transformative strategy for sustainable crop improvement amid global food security challenges.
Transposable elements (TEs) are mobile genetic elements that are widespread in nearly all eukaryotic genomes, often constituting a substantial portion of the DNA sequence. However, due to mutation accumulation and epigenetic control, the vast majority of TEs have either permanently lost their ability to transpose or remain dormant for extended periods, with only a small fraction retaining activity in contemporary genomes. Active TEs can introduce novel insertions into the genome, serving as a crucial source of genetic variation. They not only influence plant growth and development but also enhance the adaptability of plants to adverse environments. Moreover, TEs have valuable applications in various aspects of genetic research. Studies have demonstrated that external factors such as biotic stress, abiotic stress, tissue culture, and hybridization can activate their mobility. This paper summarizes current research progress on TE activity,with a focus on the regulatory mechanisms governing their mobilization, the diverse impacts of active TEs on plants, their utility in genetic research, and the primary methodologies used to investigate TE activity. By synthesizing these aspects, we aim to deepen the understanding of TE-host genome interactions, offer an integrated perspective on genome dynamics and evolution, and facilitate broader application of TEs in genetic research.
The radicle of maize (Zea mays L.) is the first nutritional organ to emerge during seedling growth. However, the regulatory mechanisms underlying the pentose phosphate pathway (PPP) in radicle growth remain poorly understood. In this study, we investigated the effects of PPP inhibition on maize radicle development by treating maize seeds with the PPP inhibitor (6-amino nicotinamide, 6-AN). We evaluated radicle phenotypes and key energy and redox indicators, and performed integrated transcriptomic and metabolomic analyses to identify differentially expressed genes, metabolites, and affected pathways. Our results showed that 6-AN treatment significantly reduced radicle length and altered the NADPH/NADP+ and NADH/NAD+ ratios, with a marked decrease in reactive oxygen species (ROS) levels, total antioxidant capacity (T-AOC), and activities of SnRK1 and NADPH oxidase (NOX). These findings suggest that PPP inhibition leads to a significant reduction in energy and redox status. Additionally, transcriptomic and metabolomic analyses revealed that redox-related pathways were significantly affected. Notably, ROS levels in the radicle were restored to some extent by the addition of H2O2 and NADPH, confirming that PPP regulates radicle growth by modulating ROS homeostasis through NADPH. This study highlights the critical role of the PPP in regulating radicle development via ROS balance.
Rice blast, caused by Magnaporthe oryzae (M. oryzae), is one of the most destructive diseases of rice and poses a critical threat to global food security. Breeding rice varieties with broad-spectrum and durable resistance through the deployment of resistance (R) genes remains the most effective and economical control strategy. In this study, we identified a highly blast-resistant rice variety, Xiushui114 (XS), by screening diverse rice accessions across three blast nurseries from 2016 to 2019. Pathogenicity assays demonstrated that XS exhibits resistance to more than 16 M. oryzae strains collected from major rice-growing regions, indicating a broad-spectrum resistance phenotype. Genetic analyses revealed that XS carries multiple R genes. Using a map-based cloning approach and sequence analysis further confirmed the presence of several known blast R genes, including Pizh, Pik-KA, Pita, and Pib, in XS. Resistance-spectrum evaluation indicated that Pizh contributes predominantly to the strong resistance observed in XS. Together, these findings establish XS as a valuable germplasm resource for improving rice blast resistance and provide new insights for breeding programs aiming to develop durable, broad-spectrum blast-resistant cultivars.
Salinity is a major abiotic stress that threatens global food security, particularly in arid and semi-arid regions. Rice (Oryza sativa L.) is a stable food for over half of the world's population. Rising soil salinity, driven by climate change and unsustainable irrigation practices, is expected to affect nearly half of the world's arable lands by 2050. This review focuses on the role of the hst1 gene, a mutant form of OsRR22, in enhancing salinity tolerance in rice. The hst1 gene plays a crucial role in maintaining ion homeostasis, scavenging reactive oxygen species (ROS), and regulating stress-responsive signaling pathways, all of which are critical for mitigating salt-induced damage in rice. Under salt stress, the hst1 mutation enables rice plants to accumulate less sodium (Na+) and more potassium (K+), thus preserving cellular function and enhancing photosynthetic efficiency. Additionally, hst1 activates antioxidant enzymes such as superoxide dismutase (SOD), catalase (CAT), and glutathione reductase (GR), which protect cells from oxidative damage caused by ROS. The gene also modulates transcription factors that regulate downstream stress-response genes, contributing to improved growth and yield under saline conditions. By understanding the physiological and molecular mechanisms underlying hst1-mediated salt tolerance, this review provides insights into breeding strategies for developing salt-tolerant rice varieties. Such advancements are essential for sustaining rice production in saline-affected regions and ensuring global food security in the face of climate change and population growth.
Agriculture has been a cornerstone of human civilization and continues evolving to meet the growing global population’s demands. In an era of rapid technological advancement, the integration of smart agriculture, also known as smart farming or precision farming, has become essential for sustainable agricultural practices. This approach uses the Internet of Things (IoT) and Artificial Intelligence (AI) to improve data collection and analysis, facilitating real-time monitoring of crop health, soil conditions, irrigation, and comprehensive farm management. By incorporating IoT, farmers can make informed decisions, optimize resource use, and improve crop yields. This study explores the scope of integrating IoT into crop management to optimize traditional crop-related issues and to improve the quality, production, and marketing process. Our in-depth analysis illustrates the IoT architecture, the IoT in smart farming, and how IoT sensors and devices collect real-time data on soil moisture, temperature, humidity, and other environmental factors, which are then processed using cloud computing platforms and stored accordingly. After conducting an exhaustive search with sorted keywords aligned with our research interests, we have identified more than a thousand research articles. A further filtering approach is applied based on inclusion and exclusion criteria to select a significant number of well-suited articles, which are thoroughly analyzed in the development of this systematic review paper. The growing number of literature in this field has created a vast area for exploration. In response, we have thoroughly studied all the aspects of IoT and have proposed a novel taxonomy for IoT to systematically classify and analyze key aspects like Sensors, Actuators, Development Board, and Communication. We examine the application of IoT in crop management, emphasizing its benefits, challenges, and future potential. In addition, we analyze key IoT components, such as sensors, data analytics, and decision-making tools, and explore their role in enhancing agricultural sustainability, economic viability, and food security. Finally, we concluded our study by addressing the challenges and outlining future directions for further advancement.
In plant chassis, peroxisomes, as key hubs of metabolism, play important roles in fatty acid β-oxidation, reactive oxygen species (ROS) metabolism and photorespiration. Fatty acid β-oxidation breaks down fatty acids through a multi-step enzymatic reaction, providing precursors for energy metabolism and phytohormone synthesis; ROS metabolism maintains redox balance through the antioxidant system and enhances plant resilience; and photorespiration relies on the synergistic action of peroxisomes with other organelles to optimize the efficiency of carbon and nitrogen utilization. Notably, peroxisomes possess a clear and simple targeting signaling system and a flexible abundance regulation mechanism, which provide unique advantages for their metabolic engineering. Recent studies have achieved effective control of fatty acid accumulation and promoted crop lipid yield by modulating the activity and transport mechanism of specific enzymes within the peroxisome. In addition, by reconfiguring the metabolic pathways in peroxisomes, the researchers successfully optimized the photorespiratory pathway, which significantly improved the energy utilization efficiency and enhanced the stress tolerance of crops. Meanwhile, the modification of ROS metabolism also showed the potential to enhance the antioxidant capacity and environmental adaptability of plants. These advances not only deepen our understanding of the peroxisome function, but also provide new strategies and technical means for crop breeding improvement and efficient secondary metabolite synthesis, which are expected to have broader application prospects in agricultural biotechnology and natural product biosynthesis in the future.
Sorghum (Sorghum bicolor (L.) Moench) remains a crucial crop for food security and livelihood resilience in arid and semi-arid regions. This study reviews existing literature to identify thematic and geographical gaps in sorghum research, focusing on climate adaptation, farmer-led practices, input systems, and institutional dynamics. The goal is to critically analyze patterns in adaptation strategies, production challenges, and future research directions that affect sorghum sustainability under changing environmental and socio-economic conditions. A systematic review method was used to analyze peer-reviewed studies. Key themes covered agronomic adaptation, genetic diversity across ecological zones, policy barriers, and innovations in disease management. The findings reveal regional differences in research focus, with limited long-term data from the Middle East, Latin America, and Eastern Europe. Interdisciplinary studies remain few, especially those linking climate projections with behavioural adoption of technologies. Improvements in soil water retention with broad bed furrows, along with yield declines under delayed sowing. The study underscores the importance of localized, farmer-centred innovations, better institutional coordination, and region-specific data collection. Promoting a research agenda that combines agronomy, policy, and social factors can strengthen sorghum's future amid rising climate uncertainty and support scalable, inclusive agricultural development strategies.
