Traditional screening for drought tolerance refers to conditions in which the germplasm/breeding populations are exposed to varying moisture stress levels in field environments. While the control plots receive optimal irrigation throughout the crop cycle, in the stressed plots, water is withheld at a critical stage (i.e., reproductive phase) for a specific period and then released similar to control (irrigated) plots. The difference in pod yield between irrigated and drought-stressed plots is measured as a response to drought stress. The genotypes that show the least reduction in pod yield under stress are classified as tolerant to drought (Craufurd et al., 2003; Krishnamurthy et al., 2007; Kakani et al., 2015; Akbar et al., 2017; Abadya et al., 2021). This type of stress is categorized as a mid- or end-of-season drought. The occurrence, frequency, and intensity of stress in natural field environments are difficult to predict, i.e., the crop may face stress at any given time during the rainy season. This type of stress is defined as intermittent drought. In a situation like this, the genotypes/breeding populations are exposed to intermittent drought stress while the corresponding control plots receive optimal irrigation throughout the crop cycle. Genotypes with the least difference in pod yield between stressed and control (irrigated) plots are identified as tolerant to intermittent drought (Gangappa et al., 2006; Ratnakumar and Vadez, 2011; Hamidou et al., 2012; Vadez and Ratnakumar, 2016). Such screening methods are time- and resource-intensive, and subject to bias due to genotype-by-environment interaction effects. However, screening only for yield response under drought or high temperature does not give us information regarding the physiological and genetic mechanisms that may be involved in the observed yield under drought tolerance.

Water use efficiency (WUE) is a critical trait in breeding for drought- and heat-stress tolerance in peanuts. However, long-term transpiration is challenging to measure under field conditions. Technically, it requires using lysimeters, which is economically unfeasible for typical peanut breeding programs. Surrogates for WUE have been identified, including carbon isotope discrimination (Δ¹³C), specific leaf area (SLA), SPAD chlorophyll meter reading (SCMR), canopy temperature depression (CTD), normalized difference vegetation index (NDVI), and visual rating of leaf wilting (Wright et al., 1994; Ravi et al., 2011; Luis et al., 2016; Vadez and Ratnakumar, 2016). Still, the labor and time required to collect these measurements are prohibitive for large populations and in multiple environments. High-throughput phenotyping in plants is thus a significant bottleneck in breeding programs.

High-throughput plant phenotyping (HTPP) employs unmanned aerial vehicle (UAV), unmanned ground vehicle (UGV), robotics, various imaging technologies, and advanced data analytics to enable efficient and effective characterization of complex plant traits for screening germplasm or breeding populations. Most HTPP research has been focused on major row crops such as cotton, maize, soybean, and wheat, whereas HTPP research in peanuts only started in recent years. Table 1 summarized a list of existing HTPP studies where the predicted traits were or could be used to screen drought tolerance in peanuts.

Infrared thermal imaging of canopy temperature is currently the most accurate and direct method, as drought-induced stomatal closure causes a reduction in transpiration and, thus, a decrease in canopy temperature. Typically, a thermal camera would be mounted on a UAV along with other imaging sensors for high-throughput multi-modal imagery acquisition over a large area. Balota and Oakes (2017) first evaluated UAV-based red-green-blue (RGB) and near-infrared (NIR) imaging and handheld RGB and thermal imaging for HTPP of 26 peanut cultivars in a drought experiment. RGB color indices, NDVI, and CTD, were found to have strong to moderate correlations with visual leaf wilting rating, pod yield, sound mature kernel, and crop value at the end of water stress imposition. Aerial RGB color indices coupled with statistical learning models have been reported to achieve a 90% accuracy in predicting visual leaf wilting ratings (Sarkar et al., 2021). CTD can detect drought stress before visible leaf wilting occurs (Balota and Oakes, 2017). Other related HTPP studies in peanuts were focused on peanut canopy morphology. Although they may not provide early detection of drought and heat stress, peanut canopy architecture traits can potentially influence plant water use. Peanut canopy height has been quantified accurately by both Light Detection and Ranging (LiDAR) sensors on a high-clearance motorized cart (Yuan et al., 2019) and digital surface models derived from a UAV-based RGB imaging platform (Sarkar et al., 2020). In addition, leaf area index (LAI) and lateral growth can be predicted by training statistical and machine learning models on aerial vegetation indices (Sarkar et al., 2021). Deep roots can increase plant water uptake capability, contributing to drought and heat tolerance. Minirhizotron imaging has been used to infield HTPP of peanut root architecture, and the UNet-based semantic segmentation method has been effective and robust in detecting root pixels (Xu et al., 2020).

HTPP of other agronomic traits in peanuts also has been standardized to accelerate peanut breeding efforts. Pod yield has been the most important trait to measure in peanut breeding programs. UAV remote sensing-based vegetation indices of peanut canopy at critical phenological stages, such as the pod-filling stage, have shown their value for early yield prediction (Balota and Oakes, 2017; Jewan et al., 2022). For direct sensing of peanut pods, ground penetrating radar has shown the potential to explain yield variability up to 51% (Dobreva et al., 2021). In addition, HTPP of infield peanut pods after inversion presents a low-cost approach for pod yield prediction at the end of the growing season. Bidese et al. (2021) employed a push-cart system to collect top-viewing and side-viewing RGB videos of inverted peanut plants in the field. They explored Mask R-CNN-based peanut pod detection coupled with multivariate linear regression for pod yield prediction. The imaged scenes were highly complex, with heavy occlusions between peanut pods, leaves, and vines. The potential of this approach needs further investigation to account for pod size and variability in visibility. Disease incidence may become a confounding factor for screening of drought and heat tolerant peanut genotypes and affect subsequent data analysis and selection process. UAV multispectral imaging-derived vegetation indices accurately predict visual ratings of tomato spot wilt virus and bacterial wilt in peanuts (Patrick et al., 2017; Chen et al., 2020).

In addition to the studies reviewed above, other HTPP technologies can facilitate breeding drought and heat tolerance in peanuts. Hyperspectral imaging provides both high resolutions in spatial and spectral dimensions for phant phenotyping applications (Sarić et al., 2022). Compared to a typical multispectral camera with five wide spectral bands (blue, green, red, red edge, and near-infrared), a visible-near-infrared (VNIR) hyperspectral camera produces hundreds of narrow spectral bands between 400 and 1000 nm wavelengths, which can reveal a far more detailed spectral signature of plant organs. New normalized difference vegetation indices (i.e., FOSBNDI-1, FOSBNDI-2, and COSBNDI) derived from UAV-based hyperspectral data were effective at predicting maize leaf water content at the V6 stage in conjunction with a machine learning model (Raj et al., 2021). In contrast to engineering spectral features, all spectral bands can also be exploited by statistical or deep learning methods (i.e., partial least squares regression or deep convolutional neural networks) for maize leaf water content prediction with automatic feature selection or learning (Ge et al., 2016; Rehman et al., 2020). Similar approaches can help rapidly screen peanut genotypes with high water uptake capability.

Regarding remote sensing of peanut yield and yield components, Bagherian et al. (2022) recently evaluated predicting biomass, pod count, and pod yield using UAV-based hyperspectral imaging and machine learning techniques for single peanut plants of an F₁ population in a mid-season drought experiment using rainout shelters. Eighteen days after the drought was found to result in the highest prediction accuracies for the three agronomic traits (R ² = 0.52–0.61). On the other hand, high photosynthetic capacity can act as a mechanism for drought and heat tolerance. Currently, photosynthetic parameters such as the maximum carboxylation rate of Rubisco (V _c,max ), maximum electron transport rate (J ₁₈₀₀ ), maximum electron transport rate supporting RuBP regeneration (J _max), maximal light-saturated photosynthesis (P _max), and chlorophyll content are often measured using a portable photosynthesis instrument (e.g., LI-COR LI-6800), which can be extremely time- and labor-intensive. Hyperspectral imaging and machine learning have been found effective in predicting photosynthetic parameters (Fu et al., 2019; Meacham-Hensold et al., 2020). This can be applied to track the temporal dynamics of photosynthetic activity in peanuts for drought experiments under rainout shelters and identify superior genotypes that quickly recover from drought stress. Chlorophyll fluorescence imaging provides a two-dimensional image instead of a point measurement using a chlorophyll fluorimeter. The resultant high spatial resolution can reveal spatial variability in photosynthetic performance on a single leaf or between leaves on a plant. Chlorophyll fluorescence imaging has been used to study drought and heat stress response in tomato, Arabidopsis, and wheat in controlled environments (Wang C. et al., 2018; Yao et al., 2018; Abdelhakim et al., 2021) and grain sorghum under field conditions (Herritt et al., 2020). Chlorophyll fluorescence measurements, such as maximum and operating quantum efficiencies of photosystem II photochemistry (F _V /F _M , Φ _PSII = (F _m’–F)/F _m’), can indicate plant drought and heat stress earlier than the occurrence of morphological changes such as leaf wilting. Table 2 lists some HTPP studies for drought and/or heat stress in crop species that could be applied to peanuts.

With the advancements in HTPP and increasing availability of high-dimensional sensor-based phenotypic datasets, phenomic-assisted selection has recently been proposed and evaluated. This analysis uses HTPP data instead of genomic data as input to the statistical models in genomic selection. The phenomic choice can achieve comparable predictive accuracy compared to genomic selection for crops such as wheat, soybean, and maize (Rincent et al., 2018; Parmley et al., 2019; Weiß et al., 2022). The advantages include low cost and robustness across different environments (Rincent et al., 2018). The phenomic selection performed more accurately for complex traits such as grain yield than traits controlled by a few genes (Zhu et al., 2022). Since drought and heat tolerance are considered complex traits, peanut breeders are expected to benefit from phenomic selection and reduce the labor and time required for screening diverse populations.

A similar approach, like discarding of ultrasusceptible types to drought stress, is also recommended for screening for heat stress tolerance in peanut (Akbar et al., 2017; Table 3). Developing a reliable index and identifying traits for acquired thermotolerance in peanuts is necessary for breeding heat-tolerant varieties. Several reports observed the genotypic variability in peanut’s heat tolerance for partitioning dry matter to pods and kernels, fruit set, membrane stability, and chlorophyll fluorescence (Srinivasan et al., 1996; Vara Prasad et al., 2001; Craufurd et al., 2003). The detached leaf assay method was used to screen the sixteen genotypes from US minicore accessions along with standard checks were evaluated for acquired thermoterance. Here, the change in the temperature sensitivity of chlorophyll accumulation was used as an indicator of acquired thermotolerance. However, in this study, there was no significance effect of thermotolerance on seed weight was observed, hence it was difficult to relate chlorophyll content with heat tolerance (Selvaraj et al., 2011; Table 3). In another study, peanut seedlings of diverse genotypes for heat tolerance were screened using temperature induction response techniques. About 2 days old peanut seedlings were exposed to sublethal temperature from 28°C to 54°C for 5 h, followed by the lethal temperature at 54°C for 3 h. The expression patterns of stress-responsive genes were analyzed in selected heat-tolerant genotypes; genes related to HSP90, DREB2A, and LEA4-2 were highly induced (Kokkanti et al., 2019) that can be used as markers for screening. Lipid peroxidation can cause severe membrane injury (MI) during abiotic stress. As such, it can be measured to assess the degree of stress in peanuts (Blum and Ebercon, 1981; Srinivasan et al., 1996; Bajji et al., 2002). Ribose, hydroxyproline, and saturated fatty acids were negatively correlated with MI, which can be used as stress tolerance parameters. Hence, there is a need to emphasize on the practical and robust screening methods to select for heat stress tolerance in peanut. Of late, studies have been carried out to utilize the thermal indices (growing degree days, phenothermal indices, heat use efficiency) for studying heat tolerance in peanut (Sukanth, 2022) and efforts are being made to map the heat tolerance related traits in groundnut (Sharma et al., unpublished).

Peanut shows different water needs at different developmental stages. The water demand is the highest at mid-pod filling stage because the peanut canopy covers all the ground and maintains open the stomata to maintain high photosynthesis to fill the growing pods (Stansell and Pallas, 1985; Rowland et al., 2012a; Rowland et al., 2012b). Understanding the main effects of drought on plant growth and yield may unfold the physiological basis of drought tolerance.

The drying of the soil due to drought and the subsequent reduction in leaf water potential and cell turgor leads to the inhibition of cell division and elongation that results in slower leaf growth rates aimed at reducing transpiration at the canopy level (Figure 2) (Ribaut et al., 2009; Avramova et al., 2015). To preserve water in the soil and maintain an acceptable leaf water potential, peanuts tend to decrease stomatal conductance (g _s ) and transpiration resulting in reduced photosynthesis (Reddy et al., 2003; Pilon et al., 2018). Reduced leaf area expansion and lower photosynthesis per leaf area lead to a decline in canopy carbon assimilation that will reduce biomass accumulation and yield (Reddy et al., 2003). Plant traits that preserve soil moisture, such as high-water use efficiency (WUE) due to rapid stomatal closure, could increase drought tolerance (Devi et al., 2010; Devi and Sinclair, 2011; Shekoofa et al., 2015; Sinclair et al., 2017; Zhang et al., 2022). Contrarily, there are peanut cultivars that can maintain adequate plant water status and escape drought by collecting more water due to a more complex or deep root system (Rowland et al., 2012a; b; Zhang, et al., 2022). Utilization of genomic and genetic resources for developing peanuts for harsh environment conditions are illustrated in Figure 1.

When drought is maintained for long periods, and CO₂ assimilation is reduced, the excess light not used for photosynthesis tends to produce reactive oxygen species (ROS) such as superoxide anion and hydrogen peroxide, H₂O₂ (Akcay et al., 2010; Laxa et al., 2019). ROS accumulation has been related to lipid peroxidation and thylakoid membrane damage (Lauriano et al., 2000; Quilambo, 2004). It decreases the effectiveness of Photosystems I and II (PSI and PSII) by producing non-stomatal limitations of drought that decrease photosynthetic efficiency resulting in reduced yield (Pilon et al., 2018). Peanut cultivars can tolerate these effects of drought by accumulating antioxidant substances that can reduce the accumulation of ROS to maintain higher photosynthetic rates during drought (Figure 2) (Akcay et al., 2010; Li et al., 2021).

Drought not only decreases plant growth and yield through a decrease in leaf and canopy photosynthesis (Reddy et al., 2003; Pilon et al., 2018) but also decreases nitrogen uptake by inhibiting the biological nitrogen fixation (BNF) in the roots (Pimratch et al., 2007). Nitrogen fixation is more sensitive to drought than photosynthesis because drought increases oxygen permeability to the nodule, inhibiting the activity of nitrogenase, the enzyme that catalyzes BNF (Parsons and Sunley, 2001). Drought also reduces the transport of the amino acid products of BNF from the nodule to the shoot, accumulates amino acids in the nodules, and creates a feedback mechanism that inhibits BNF (Peoples et al., 1986; Vessey et al., 2005). Thus, selecting genotypes with higher BNF under drought is another target to improve peanut yields (Sinclair et al., 1995; Sinclair, 2011; Devi et al., 2013). Peanut plants may therefore adapt diverse physiological attributes to balance productivity and stress tolerance as detailed herewith.

Higher temperatures can disrupt the physiological processes in plants, including a reduction in the rate of photosynthesis, degradation of chloroplast proteins, damage to PSII, lower relative water potential, ROS accumulation, and increase in lipid peroxidation (Crafts-Brandner and Salvucci, 2002; Dutta et al., 2009; Hasanuzzaman et al., 2020a). Heat stress affects the growth of male and female reproductive organs by impairing pollen tube growth, pollen viability, germination, egg viability, and fertilization (Figure 3). Microsporogenesis (3–6 days before flowering) and fruit set are two critical stages of peanut development which is affected by high temperatures (Craufurd et al., 2002, 2003). The late flowering to early seed setting stage was observed to be highly susceptible to high temperatures in peanuts (Prasad et al., 1999). However, the time of flower initiation at temperatures higher than 40°C/28°C day and night is the primary determinant of pod number in peanuts (Craufurd et al., 2000).

The impact of the elevated temperature is devastating as it affects membrane stability, inactivates chloroplast and mitochondrial enzyme function, causes protein degradation, reduces carbon metabolism, and alters cytoskeleton organization (Bita and Gerats, 2013) (Figure 3). Under heat stress, the thylakoid membrane and photosystem II (PSII) are severely damaged, disrupting the electron transport system and ATP synthesis during photosynthesis (Wang Q. L. et al., 2018). At temperatures higher than 38°C/32°C, the leaf chlorophyll content is reduced, which leads to less photosynthesis and low sugar content (Liu and Hang, 2000). Heat stress also affects the water status in the plant cells due to osmotic perturbation caused by low photosynthetic capacity, reduced sugar content, and higher transpiration rate (Hemantaranjan et al., 2018). High temperature under limited water conditions causes a reduction in relative water content and total water absorption rate, which contributes to total yield loss (Ashraf and Hafeez, 2004) (Figure 3). Stomatal conductance (g _s ) is directly correlated to the rate of photosynthesis and transpiration rate. During heat stress, stomatal density and stomatal conductance were found to be decreased in susceptible peanut varieties (Dash et al., 2020).

The efficient partitioning and allocation of carbon assimilate and photosynthates from source to sink are essential for plant growth and yield. Heat stress reduces carbon assimilation and partitioning and affects the plant source-to-sink relationship. Seed weight and yield depend on net photosynthesis and re-translocation of water-soluble carbohydrates from vegetative organs during seed filling (Fischer, 2011). The partitioning of dry matter allocation to shell relative to seed was higher in peanuts during higher temperatures. Higher temperature affects the photosynthates partitioning to the pod leading to low pod yield. The tolerant peanut genotypes were found to have higher crop growth rate and pod growth rate under heat-stress conditions than in the non-stress environment. Further, the photoperiod × temperature interaction influences the partitioning of dry matter into pods in peanuts (Nigam et al., 1994; Nigam et al., 1998; Akbar et al., 2017). Heat stress also damages the enzymes involved in nutrient metabolism and disrupts nutrient acquisition (Hungria and Kaschuk, 2014) (Figure 3).

Plants can tolerate the adverse effects of heat stress to some extent by modifying physiological and biochemical processes such as solute accumulation, osmotic adjustment, cellular homeostasis maintenance, and redox balance (Janská et al., 2010). One of the strategies for heat-stress tolerant cultivars is to have higher ceiling temperatures for pollen germination. Since heat stress significantly impacts peanut flowering, genotypes with a higher pollen germination ceiling temperature tend to be heat tolerant (Kakani et al., 2002; Chaudhary et al., 2022). Another effect of heat tolerance on peanuts or any other crop is membrane injury; increased unsaturated fatty acid levels caused by heat stress can disrupt membrane permeability, causing disrupted cellular homeostasis (Marcum, 1998). Photochemical changes during photosynthesis and reactive oxygen species (ROS) production due to heat stress can also affect membrane integrity (Bita and Gerats, 2013), causing membrane leakage. The heat-tolerant peanut varieties can be screened using an electrolyte leakage test or membrane injury test (MIT) by evaluating cell membrane thermostability (CMT) (Lauriano et al., 2000; Yeh and Lin, 2003; Akcay et al., 2010). Craufurd et al., 2002 reported the lower membrane injury in heat-tolerant peanut varieties. The higher heat tolerance is observed to be associated with higher gs value, photosynthesis rate, and stomatal conductance (Awasthi et al., 2014). The higher carotenoid content helps in preventing photo-oxidation of chlorophyll during stress. A study showed the heat tolerance peanut cultivars exhibited the higher carotenoid and higher chlorophyll level in the leaf (Dash et al., 2020). The protective mechanism of heat tolerance is associated with the activation of enzymatic and non-enzymatic ROS scavenging compounds. The higher activity of antioxidants such as SOD (Super oxide dismutase), CAT (Catalases), APX (Ascorbate peroxidase), and GR (Glutathione reductase) has been reported to provide plant thermostability (Kumar et al., 2013). Higher levels of total soluble sugar improve the thermotolerance of legume plants primarily by increasing water relations and gas exchange activities, thereby enhancing vegetative and reproductive growth (Ahmad et al., 2021a). Phytohormones such as abscisic acid (ABA) play a vital role in the stress tolerance of the plant by regulating its physiological processes. Plant growth regulators such as ABA, TU (Thiourea), GABA (Gamma-Amino butyric acid), and brassinosteroids help in enhancing the accumulation of soluble sugar, osmoprotectants, antioxidant enzymes, and gas exchange traits during heat stress tolerance (Ahmad et al., 2021b). Further, heat shock proteins (HSPs) play a crucial role in thermotolerance by maintaining protein structure and membrane integrity. Gene expression profile revealed that HSPs and heat shock factors (HSFs) are involved in tolerance during heat stress in legume crops (Zhang et al., 2015; Ma et al., 2016).

Plant genetic resources are the basic raw materials to empower crop improvement programs. The peanut gene pool consists of cultigen (Arachis hypogaea) with its many landraces, improved cultivars, and more than 83 wild species of the genus Arachis (Gregory et al., 1980). Cultivated peanut is an allotetraploid (2n = 4× = 40) that originated from natural hybridization of two diploid species, Arachis duranensis (A-genome) and Arachis ipaensis (B-genome) followed by spontaneous polyploidization of the hybrid and its subsequent domestication by Neolithic proto-farmers 6-10,000 years ago (Bertioli et al., 2016). Cultivated peanut germplasm is classified into two main subspecies: A. hypogaea and A. fastigiata. The subsp. hypogaea contains two botanical varieties: hypogaea (Virginia type) and hirsuta, while subsp. fastigiata contains four botanical varieties: fastigiata (Valencia type), peruviana, aequatoriana, and vulgaris (Spanish type). All six botanical varieties have unique morphological characteristics that separate them from one another (Krapovickas et al., 1994). Worldwide over 15,000 peanut accessions are preserved in the national and international genebanks, including 1823 accessions in N.I. Vavilov Research Institute of Plant Industry, Russia; 14,320 in ICRISAT, India; 7,432 in USDA, Griffin; and 9103 in China (Pandey et al., 2012). Assessment of such a large gene pool for agronomically beneficial traits is economically not feasible and also subject to high genotype by environment interaction. Smaller subsets representing a diversity of the entire collection of given species preserved in a genebank are ideal resources to evaluate for agronomic and stress tolerance traits.

Reduced subsets in the form of core (Frankel, 1984) and mini core (Upadhyaya and Ortiz, 2001) collections provide smaller subsets of germplasm that could be used to mine valuable traits from entire germplasm collections more efficiently instead of screening germplasm as a whole collection. Several such sets are reported for efficient utilization of genetic resources in peanut breeding and genetics (Ding et al., 2022; Dwivedi et al., 2008; Holbrook and Stalker, 2003; Holbrook and Dong, 2005; Upadhyaya et al., 2002a and; 2002b). The U.S. peanut mini-core collection has been effectively used for the identification of interesting alleles and traits for breeding programs for traits related to drought tolerance (Wang H, et al., 2016; Wang M. L. et al., 2016; Zhang et al., 2019; Li et al., 2022; Patel et al., 2022; Zhang et al., 2022). End-of-season drought tolerance was reported in 15 accessions after evaluating ICRISAT peanut mini core collection and selected based on pod yield, SPAD, and SLA measurements (Upadhyaya, 2005).

Assessment of 150 peanut genotypes under rainout shelters showed significant differences in pod yield, relative water content (RWC), SLA, leaf dry matter content (LDMC), chlorophyll fluorescence (CHF), Δ¹³C, photosynthesis, and stomatal conductance (g_s), and resulted in 13 accessions resistant to midseason drought stress. In addition, gas exchange parameters were measured regularly during the drought and recovery to monitor dynamic changes in photosynthesis and g_s under stress. Genotypes with high yield, Δ¹³C, photosynthesis, and g_s under stress were classified as water spenders while genotypes with equally high yields but with low Δ¹³C and g_s and moderate photosynthesis under drought stress were classified as water savers (Zhang et al., 2022). The previous reports on screening US peanut mini core collection across three irrigation treatments over 2 years and two field locations unfolded five accessions (PI 502120, PI 493329, Line 8, Georgia-06G, AU-NPL-17) as resistant to drought. These accessions had high SPAD, flowering, and paraheliotropism (Selvaraj et al., 2009; Belamkar, 2010). Paraheliotropism refers to condition in plants wherein the plants orient their leaves parallel to incoming rays of light. Elsewhere several germplasms tolerant to drought were reported, which showed significant differences in root depth, length, and density. The tolerant germplasms (#11, #34, #49, A596, Datangyou, Fenghua 1, Huayu 17, Huayu 21, Huayu 22, Huayu 25, Huayu 27, Ji 0212-4, Jihua 2, Jihua 4, L19, L121, L146, Luhua 14, NC6, Rugaoxiyangsheng, Shanhua 11, Tai 0125, Tai 0005, Taihua 4, Tangke 8, Xianghua 2008, Xianghua 55, Xuhua 13, Yuanza 9102, Yuanza 9307, Yueyou 7, Zhonghua 8) display increased root to shoot ratio and the enhanced root length and density, particularly in the deep soil, in comparison to those grown under normal growth conditions. The dragon-type (runner) peanuts, such as ‘A596’ and ‘Rugaoxiyangsheng,’ were more tolerant, followed by Virginia, Spanish, intermediate, and Valencia peanuts (Yang et al., 2019).

A screen of 247 ICRISAT accessions tested under water-deficit environments in Africa and India identified a few most adapted genotypes [ICG 5891, ICG 6057, ICG 9777] across moderate- and high-temperature stressed environments. It showed significant genotype × environment interaction (Hamidou et al., 2012). Field assessment of advanced breeding lines under irrigated conditions during the postrainy season in heat-stressed (air temperature 35°C and above during flowering) and non-stressed (air temperature below 35°C irrigated) environments unfolded large variability for pod yield and physiological traits among genotypes in a heat stress environment. A pod yield reduction of 1.5%–43.2% was recorded under a heat-stressed climate. The genotypes under heat stress either yield poorly stable or increased pod yield under high-temperature stress. The heat-tolerant genotypes are identified based on the stress tolerance index. GJG 31, ICGV 87846, ICGV 03057, ICGV 07038, and GG 20 showed an increase in pod yield of 9.0%–47.0% at high temperatures, with a 0.65%–3.6% increase in pod growth rate. ICGV 06420, ICGV 87128, ICGV 97182, TCGS 1043, and ICGV 03042 were stable for pod yield and recorded a 0.25%–3.1% increase in pod growth rate. Thus, pod yield, hundred-seed weight, and pod growth rate under heat stress can be used to select heat-stress tolerant genotypes. Based on stress tolerance indices and pod yield performance, ICGVs 07246, 07012, 06039, 06040, 03042, 07038, and 06424 were identified as heat-tolerant genotypes and are used as parents in breeding programs in India (Akbar et al., 2017).

Developing reliable indices and traits for acquired thermotolerance in peanuts is necessary for breeding heat-tolerant varieties. Several reports observed the genotypic variability in peanut’s heat tolerance for partitioning dry matter to pods and kernels, fruit set, membrane stability, and chlorophyll fluorescence (Srinivasan et al., 1996; Vara Prasad et al., 2001; Craufurd et al., 2002).

Wild Arachis species originated in South America, selected during evolution in a range of environments and biotic stresses, which provided opportunities for the evolution of a rich source of allele diversity for resistance to several pests, including fungal diseases and drought tolerance (Bertioli et al., 2021). Thoppurathu et al. (2022) conducted A transcriptome analysis for A. duranensis (drought tolerant) and Arachis stenosperma (drought susceptible) revealed A. duranensis genotype had a higher number of transcripts related to DNA methylation or demethylation, phytohormone signal transduction and flavonoid production, transcription factors, and responses to ethylene, indicating that it is tolerant to drought stress. Exposing A and B genome diploids under progressive drying to examine curves of vapor pressure deficit (VPD) against a fraction of transpirable soil water (FTSW) revealed that A. magna and A. duranensis accessions had superior ability to regulate transpiration under water deficit stress (Leal-Bertioli et al., 2012).

In summary, although we have large number of accessions of peanut germplasm collection, a very limited numbers have been identified as abiotic stress tolerant genotypes. Some of those tolerant genotypes have been used in breeding programs, and others are available for further investigation as shown in Table 4.

Breeding for drought and heat stress tolerance is extremely challenging due to the complexity associated with various stress-adaptive mechanisms, uncertainty in the onset and intensity of stress, and large genotype × environment interactions. Conventional crossing and selection to improve drought and heat stress tolerance have been met with limited success. The focus on yield, lack of simple physiological traits as a measure of tolerance, and complex inheritance (polygenes with small effects) contributed to the low genetic gain in stress tolerance breeding. A better understanding of the physiological basis of stress tolerance contributes to identifying and manipulating traits associated with yield in water-deficit stressed field environments (reference needed). A systematic characterization of germplasm and breeding lines resulted in genetically diverse germplasm varying in response to drought and heat stress (Table 5). Such genetic stocks showed large variations for physiological traits such as SLA, chlorophyll content, amount of water transpired, WUE and harvest index in a drought-stressed environment. Both empirical (yield in the stressed environment) and trait-based selection (SLA, SCMR) have led to the development and release of drought-tolerant peanut cultivars in India and Australia (Rachaputi, 2003). The trait-based selection, however, did not show a consistent superiority over the empirical selection for drought tolerance (Nigam et al., 2005). Integrating physiological traits or their surrogates in the selection scheme would be advantageous in selecting segregants that utilize water more efficiently and partition photosynthates more effectively into economic yield. A few drought-tolerant cultivars of wheat bred by trait-based breeding without any yield penalty have been released in Australia where drought is intense and terminal (Rebetzke et al., 2002; Richards et al., 2002; Condon et al., 2004). A combination of trait-based selection in an early stage of breeding and yield assessment at a later stage of cultivar development in target environments is needed to select for abiotic stress adaptation and yield in peanuts. High yield potential and higher resistance are difficult to target together; therefore, to avoid the yield penalty, cultivars with high yield potential were bred to moderate levels of stress tolerance (Nigam et al., 1991). A few commercial heat-tolerant varieties have been released through conventional breeding, such as 55-437, 796, ICG 1236, ICGV 86021, ICGV 87281, and ICGV 92121 (Craufurd et al., 2003).

Crop wild relatives are the source of variation for stress tolerance and productivity traits. Advanced backcross populations originating from synthetic amphidiploid as donors for wild alleles detected several QTLs with positive effects on pod/seed size and adaptation traits in water-limited environments (Essandoh et al., 2022). Progenies derived from a cross between synthetic allotetraploid (A. duranensis × A. batizocoi) and cultivated peanut improved photosynthetic traits and yield under water-deficit stress (Dutra et al., 2018). A root transcriptomic study involving drought tolerant (A. duranensis) and susceptible (Arachis stenosperma) wild species unfolded 1465 differentially expressed genes (DEGs) under drought stress and 366 polymorphic SNPs among DEGs. Three SNPs differentiated the two species and may be useful for selecting drought-tolerant lines (Thoppurathu et al., 2022). In addition, advanced backcross populations involving several synthetics (ISATGR 121250, ISATGR 278-18, ISATGR 265-5, ISATGR 40) and peanut cultivars (ICGV 91114, ICGV 87846, TMV 2, Tifrunner) exhibited considerable variability for morpho-agronomic traits (Sharma, 2017). They could be a good resource for screening for abiotic stress adaptation and use in peanut breeding. Synthetics are tetraploid derivatives originating from a cross between two diploid Arachis species from secondary gene pool with different genomes (Mallikarjuna et al., 2012).

Advances in peanut genomics unfolded several QTLs and markers (SSRs, SNPs) and candidate genes associated with drought tolerance surrogate traits in peanut. For example, NDVI effectively predicts biomass and yield, while CTD is associated with transpiration efficiency and carbon dioxide assimilation. These markers explained between 6.6% and 20.8% phenotypic variation, with most markers identified on the A subgenomes and respective homeologous chromosomes on the B subgenomes (Shaibu et al., 2020). Such markers, upon validation, may be deployed in marker-assisted breeding for drought tolerance in peanuts. A number of NAM- and MAGIC-based populations are being developed in peanuts which may provide useful genetic and genomic resources to study and implement genomic-assisted breeding for enhanced resistance to drought and heat stress in peanuts (Holbrook et al., 2013; Varshney, 2016; Gangurde et al., 2019). Efforts are underway to apply genomic selection (GS) for predicting the phenotypes by studying their genotypic architecture in multi-environment breeding trials (Pandey et al., 2020), but study of GS related to abiotic stress tolerance is not yet defined in peanut. Marker-assisted breeding has successfully introduced resistance to nematodes, rust and leaf spot and improved oil quality in peanuts (Varshney et al., 2014; Bera et al., 2018; Ballén-Taborda et al., 2022).

Ascorbate peroxidase (APX), an antioxidant enzyme, contributes to ROS scavenging by decreasing hydrogen peroxide (H₂O₂) under environmental stresses. A comprehensive GWAS unfolded 166 AhAPX genes in the peanut genome, grouped into 11 main clusters, and have roles in oxidoreductase activity, catalytic activity, cell junction, cellular response to stimulus and detoxification, biosynthesis of metabolites, and phenylpropanoid metabolism. AhAPX4/7/17/77/82/86/130/133 and AhAPX160 showed significantly higher expression in diverse tissues/organs, while AhAPX4/17/19/55/59/82/101/102/137 and AhAPX140 were significantly upregulated under drought and cold stress, and phytohormones treatments. Functional characterization and validation of the AhAPX and SNPs may accelerate breeding programs to develop stress-tolerant peanut cultivars (Raza et al., 2022).

US peanut researchers at Clemson University reported lipid metabolism traits associated with heat tolerance which could be useful in determining lipid biomarkers to develop climate-resilient varieties. A targeted effort is underway in the US to develop new heat-resilient peanut cultivars using a range of heat-sensitive to heat-tolerant varieties such as “Bailey,” “Georgia 12Y,” “Phillips,” “Sugg,” “Tifguard,” and “Wayne” and a breeding line SPT06-07 (Zoong Lwe et al., 2020). The development of molecular marker linkage maps and identification of markers and QTLs for target traits paved the way to develop efficient breeding methods to generate new, improved heat-tolerant peanut cultivars. The availability of peanut genome sequences and advanced genomics tools will aid in efficiently utilizing genetic resources toward a generation of sustainable crop yield.

The classical breeding methods employed to enhance drought and heat stress tolerance have had limited success. Advances in next-generation sequencing and phenomics, availability of genome sequences, and advances in bioinformatics and biotechnological tools may open new windows of opportunities to improve abiotic stress adaptation in food crops, including peanuts.

Transgene and genome editing methods make up the core of the genetic engineering tool kit. These technologies alter a genome to create modified cell lines, new cultivars possessing valuable traits, or learn novel information about cellular processes or development. A transgene is a unit of genetic material inserted permanently or transiently into a cell where it is expressed to confer a phenotype. Efforts have been made to transfer genes of interest into the peanut genome via Agrobacterium-mediated transformation or particle bombardment. However, the successful production of transgenic plants has been limited since only a few genotypes were found to be transformable with relatively high efficiency. In peanuts, many factors, including a lack of efficient protocols to regenerate whole plants (Sharma and Anjaiah, 2000; Sharma and Ortiz, 2000; Geng et al., 2012; Chu et al., 2013) and bacterial interactions with peanut cells (Gelvin, 2003) may restrict transformation success and regeneration via Agrobacterium. The recalcitrance nature of many peanut cultivars to Agrobacterium-mediated transformation and regeneration is a challenging bottleneck for future peanut-based technology development. Therefore, there is a pressing need to explore suitable genotype-independent transformation methods, such as in planta transformation, which may avoid time-consuming tissue culture and regeneration processes.

Nevertheless, the technology has successfully deployed to create transgenic events in peanuts with enhanced drought, salt, and aflatoxin tolerance. Transgenic peanuts containing AtDREB1A confer tolerance to drought. Assessment of these events under varying moisture stress regimes and vapour pressure deficits (VPDs) yielded up to 24% improvement in seed yield largely due to increased harvest index and higher seed filling, and displayed 20%–30% lower pod yield reduction than WT under drought stress (Bhatnagar-Mathur et al., 2014). Another study led by Qin et al. (2011) reported that regulating the expression of the IPT gene by a water-deficit inducible promoter (P_SARK) performed much better, maintained higher photosynthetic rates and stomatal conductance (g _s ), produced significantly more biomass, and yield under reduced irrigation conditions in greenhouse and field environments. Transgenic peanut plants overexpressing the AtAVP1 were tolerant to both drought and salt stress, produced high biomass, and maintained higher photosynthetic and transpiration rates under reduced irrigation and saline conditions in the greenhouse. Additionally, transgenic peanuts expressing the transcription factors AtNAC2 and MuNAC4 from Arabidopsis and Macrotyloma showed high tolerance to drought, salt, and moisture stress and high yield in stressed environments (Pandurangaiah et al., 2014; Patil et al., 2014). Similarly, the expression of the mtlD (mannitol-1- phosphate dehydrogenase) in transgenic peanut plants conferred drought tolerance (Bhauso et al., 2014a; Bhauso et al., 2014b; Patel et al., 2017), and the overexpression of GmMYB3a into transgenic peanut plants displayed better physiological parameters with improved drought tolerance (He et al., 2020).

Since stress tolerance is a multigenic trait involving different signaling cascades, developing transgenic lines with more tolerance traits by transferring more than one gene is needed (Venkatesh et al., 2018). Co-expression of multiple genes in transgenic plants has shown improved stress tolerance compared to transgenic plants with single-gene. Using modified MultiSite Gateway approach (Vemanna et al., 2013) to simultaneously stack Alfin1, PgHSF4, and PDH45 genes driven by individual promoters and terminators into a single vector resulted in transgenic peanut plants with improved stress tolerance, higher growth, and productivity than WT plants under drought-stress conditions (Table 6). Another successful example was that two-antifungal plant defensins MsDef1 and MtDef4.2 and two host-induced gene silencing of aflM and aflP genes were cloned into four binary vectors. These vectors were mobilized into Agrobacterium, resulting in transgenic peanuts with a near-immunity of aflatoxin contamination (Sharma et al., 2018). This gives hope that day is not far off to stacking genes cascade with suitable promoters for developing peanuts that combine aflatoxin resistance, tolerate drought as well productive. Conventional breeding has had limited success to achieving resistance to aflatoxin because of multiple mechanisms (in-vitro seed colonization, pre-harvest aflatoxin contamination and aflatoxin production) controlling aflatoxin contamination, phenotyping bottlenecks to measure different components of resistance, large genotype × environment interactions, and issues associated with pre- and post-harvest management of peanuts (Pandey et al., 2019).

Overcoming tradeoffs is a significant breeding challenge when combining stress tolerance and crops productivity as many genes of minor effects are involved. Identifying gene variants with diverse functions to overcome tradeoffs should receive a greater investment of time and resources to balance crop growth, stress tolerance and productivity (Dwivedi et al., 2021).

Genome editing involves transgenes or occasionally only proteins with or without an RNA, which can modify existing genetic material in a targeted manner, creating insertions, deletions, or base modifications. This technology has provided an alternative approach to plant breeding and has been efficient in producing new cultivars and genetic resources within a relatively short period. The clustered regularly interspaced short palindromic repeats (CRISPRs) and CRISPR-associated protein (Cas) type II systems provide methods for rapidly and efficiently editing plant genomes. This is accomplished through an RNA (CRISPR) guided nuclease (Cas) induced targeted double-strand DNA break, which can be repaired through several pathways that may lead to mutation. Furthermore, genomic DNA can be modified by tethering various enzymes to a nuclease-deficient Cas protein, which may also introduce targeted mutations (Xie et al., 2015; Samanta et al., 2016; Li et al., 2019). Specifically, CRISPR/Cas9-related technologies have tremendously impacted functional genomics by enabling selective and specific alteration of genomic DNA sequences in vivo (Li et al., 2019; Scheben et al., 2017; Zhang et al., 2019). Genome editing has various applications in plants, including basic and applied biological research for developing advanced biotechnology products (Zhang et al., 2019) through forward and reverse genetics, targeted gene insertion, promoter modification, and splice variant generation. This is a valuable technique for functional gene analysis or trait alteration, and its effectiveness has been demonstrated in many plant species.

Moreover, gain-of-function mutations through this same approach have been reported in plants (Wang et al., 2022). Additionally, identifying other type II Cas proteins, such as Cas12a (Cpf1), is another important avenue in genome editing research as it opens additional genomic regions to modification due to alternative PAM: TTTV (V=C, G, A) utilization. The recent development of Cas protein variants fused to a variety of enzymes has also widened the application of CRISPR/Cas technology from inducing indels, gene insertion, and gene modification through targeted double-strand DNA breaks to produce targeted single base changes through a system known as base editing (BE) or manipulating the gene expression through the promoter region.

Although CRISPR/Cas tools have been successful in a wide variety of plant species, their application in peanuts is currently limited. Realizing the full potential of CRISPR/Cas-based genome editing in peanuts will require the development of a toolbox of validated CRISPR/Cas constructs and protocols for their utilization. Several research programs have been focused on establishing these systems in peanuts by developing gene editing and base editing technology to include vectors using CRISPR/Cas9 and CRISPR/Cas12a variants to induce indels, make DNA alterations through base editing, or regulate gene expression. Work has progressed to develop delivery methods and validate construct functionality. For example, we have developed several constructs for genome editing using the peanut FAD2 genes as proof-of-concept experiments. Two CRISPR/Cas9 constructs, pDW3872, and pDW3877, have induced indels in FAD2 with an efficiency of up to 32% and 24%, respectively (unpublished). Two base editing constructs, pDW3873, and pDW3876, were developed using nCas9 fused to the cytosine deaminases APOEBEC1 or PmCDA1, respectively. These constructs successfully induced C to T changes with an overall efficiency of up to 21% and 42%, respectively (unpublished). Additional constructs (pDW3882 and pDW3886) expressing the enzymes AsCpf1 or LbCpf1 were investigated for their editing efficiency. The latter was more effective in peanuts than the former (unpublished). These preliminary results demonstrate that genome editing using CRISPR/Cas systems is feasible in peanuts.

Overall, using transgenes and genome editing technology in peanuts comes down to developing genotype-independent transformation protocols, identifying genes of interest and proof of concept using suitable genome editing constructs. Application of these advances will greatly accelerate genetic improvement in peanuts leading to the efficient generation of new lines with desirable traits, which will benefit peanut producers, industry, and consumers. On the other hand, the presence of foreign genes in transgenic plants triggers biosafety regulations. However, a comparison of transgenic Golden rice GR2E and conventional rice showed no statistically significant differences in the concentrations of phytic acid or the levels of trypsin inhibitor and no differences in pest and disease reactions between them (Mallikarjuna Swamy et al., 2019; Mallikarjuna Swamy et al., 2021). Comparative assessment of transgenic wheat containing the sunflower gene, HaHB4 conferring drought tolerance and improved yield in driest environments, is nutritionally equivalent to non-transgenic wheat lines (González et al., 2019; Miranda et al., 2022). Because CRISPR/Cas9-based genome editing plants can be transgene-free by crossing edited plant offspring, they are assigned a non-regulated status (Ahmad et al., 2021), such as the common button mushroom modified by the CRISPR system obtained a non-regulated status in 2016 (Waltz, 2016). ‘Sanatech Seed’ has launched the world’s first genome-edited high-GABA tomato with enhanced nutritional benefits for consumption in Japan. This tomato contains high levels of gamma-aminobutyric acid (GABA), an amino acid believed to aid relaxation and help lower blood pressure (https://www.isaaa.org/kc/cropbiotechupdate/article/default.asp?ID=18668).

The evidence to date and predictions suggest the overall negative effect of climate change on agricultural production especially when more food, nutritious and safe, is required to feed the growing word population. In addition, the food and feed produced today is less nutritious, and faces an increased risk to contamination by mycotoxin producing fungi due to climate change.

Peanut at reproductive stage is ultra-susceptible to drought and heat stress, causing substantial loss to production and nutritional quality. The drought-stressed peanut is prone to aflatoxin contamination. Thus, aflatoxin contaminated peanut is hazardous to human and animal health, in addition to adversely impacting peanut trade.

Novel sources of resistance to drought in peanut gene pool has led researchers thoroughly investigate the physiological and molecular basis of stress tolerance, while the genes and markers associated with stress tolerance detected significant marker trait associations, which after validation may be deployed in genomics-assisted breeding in peanut. Greater resources are needed to unfold the genetic and molecular basis of heat stress tolerance as this trait in the past received less attention compared to drought research in peanut. A few reports indicate the feasibility of gene transfer by transgenic technology, with some events showing no growth-defense tradeoff, suggesting the transgene(s), a valid technology, to rapidly integrate stress tolerance gene(s) without yield penalty. Advances in developing peanut-based construct and their editing efficiency demonstrate that genome editing using CRISPR/Cas system is feasible in peanut. Public perception about the use of genetically modified and/or gene edited crops is gradually changing in favor of for food and feed uses and also for commercial cultivation, as the evidence to date suggests no significant changes in proximate composition between genetically engineered and conventionally-bred produce, except for the trait introduced.