During flower development, male and female organs are sensitive to high temperature, especially ≥30°C (Figure 4; Lavania et al., 2015). Heat stress severely affects flower bud initiation, and this sensitivity prevails for 10–15 days (Hedhly et al., 2009; Bita and Gerats, 2013) as reported in faba bean (Bishop et al., 2016), common bean (Vara Prasad et al., 2002), and soybean (Kitano et al., 2006). Heat stress influences the reproductive stage by decreasing the number and size of flowers, deforming floral organs, resulting in loss of flowers and young pods, and hence reduction in seed yield (Morrison and Stewart, 2002), as reported in chickpea and mungbean (Tickoo et al., 1996), common bean (Gross and Kigel, 1994; Suzuki et al., 2001), cowpea (Hall, 1992), pea (Stanfield et al., 1966), and peanut (Vara Prasad et al., 1999a). A mild heat stress during floral development severely reduced yield in faba bean (Bishop et al., 2016). The flowering stages are more susceptible to heat stress, and high temperatures are likely to coincide with gametophyte development and anther dehiscence in faba bean and some other legume species (Bishop et al., 2016).

Meiosis is an important stage in the sexual life cycle of a plant to allow the diploid sporophytic cells to produce haploid gametophytes (Thakur et al., 2010). After the inception of meiosis, the sensitivity of the male gametophyte to stress increases dramatically, with negative consequences for anthesis, pollen fertility, pollination, female fertility, early zygote development, and seed yield (Boyer and McLaughlin, 2007). In microsporogenesis of chickpea, meiosis and pollen development are most affected by heat stress (Devasirvatham et al., 2012a). Sexual reproduction and flowering, in particular, are extremely sensitive to heat stress, and often results in reduced crop productivity (Thakur et al., 2010; Bita and Gerats, 2013). Heat stress mainly accelerates the onset of anthesis, thereby initiating the reproductive stage prior to the accumulation of sufficient resources (Zinn et al., 2010; Bita and Gerats, 2013).

Male reproductive development in higher plants is very sensitive to heat stress at all growth stages (Bita and Gerats, 2013; Sage et al., 2015). In particular, high temperature stress results in a lower seed set due to male sterility in most legume crops, including chickpea (Devasirvatham et al., 2012a), common bean (Monterroso and Wien, 1990), cowpea (Warrag and Hall, 1983), and field pea (Jiang et al., 2015). In most legumes, the male gametophyte is more sensitive to high temperature than the female gametophyte (Devasirvatham et al., 2012a; Sage et al., 2015; Bhandari et al., 2016). Development of the male gametophyte (pollen grains) starts with the separation of reproductive tissue from the anther, followed by meiosis of the pollen mother cell, mitosis and microspore maturation, and the formation of mature pollen grains (Bita and Gerats, 2013). Specialized anther tissue has non-reproductive (tapetum for support, stomium for dehiscence) or reproductive functions (pollen mother cell for pollen formation). Male fertility depends on both the status of the tapetum and microspore development (Zinn et al., 2010; Bita and Gerats, 2013). Heat stress alters gene expression, which is possibly connected to tapetum degeneration and pollen sterility, in most plant species (Oshino et al., 2007; Endo et al., 2009). Sakata et al. (2010) suggested that understanding heat stress effects on pollen development will involve observations on carbohydrate turnover during this stage. Mature pollen grains are more tolerant to heat stress than any other stage of male gametophyte development (Hedhly, 2011). Tolerance of pollen grains to high temperature may be associated with its low plasma content, low metabolic activity to its protective structures, or its carbohydrate content and dynamics (Kaushal et al., 2013; Figure 5). Pollen grains penetrate the stigmatic surface, and pollen tube growth starts within the style and within the ovary toward the female gametophyte; the pollen tube growth rate is the first and most important characteristic to check under heat stress (Hedhly, 2011). Heat stress affects male sterility in most sensitive crop plants, by impairing pollen development to severely reduce yield (Wassmann et al., 2009; Bita and Gerats, 2013), as reported in cowpea (Ahmed et al., 1992), chickpea (Devasirvatham et al., 2012a, 2013; Kaushal et al., 2013), common bean (Gross and Kigel, 1994), groundnut (Vara Prasad et al., 1999b), soybean (Djanaguiraman et al., 2013), chickpea (Devasirvatham et al., 2013), field pea (Jiang et al., 2015), and faba bean (Bishop et al., 2016). Developing anthers are a strong resource sink and heat stress affects the development of tapetum cells and microspores, which involve DNA, carbohydrates, proteins, and lipids synthesis (Ma, 2005; Sage et al., 2015). Tapetal cells and microspores are separated symplastically from other anther tissue, and tapetal cells are metabolically highly active to nourish the growing microspores. The high transport and metabolic activity of the tapetum layer is indicated by the presence of some cell organelles such as plastids, mitochondria, peroxisomes, and endomembrane and cytoskeleton systems involved in processing and transporting metabolites (Bagha, 2014). Suzuki et al. (2001) found that heat stress caused early degeneration of the tapetum layer and disrupted ER in Phaseolus vulgaris.

Heat stress delinks source and sink strength leading to depletion of available carbohydrates at the reproductive stage of plant development, ultimately reducing fruit set and other yield attributes in chickpea (Nayyar et al., 2005; Kaushal et al., 2013) (Figure 5) and lentil (Bhandari et al., 2016; Sita et al., 2017). High temperature also influences early abortion of tapetal cells which leads to pollen sterility (Parish et al., 2012), structural abnormalities in developing microspore-associated tapetal degeneration due to deformity in ER (Peet et al., 2002), fertilization arrest and abrupt embryo development (Barnabás et al., 2008), reduced seed germination, loss of vigor, and reduced seedling emergence in many crop plants (Akman, 2009; Ren et al., 2009; Bita and Gerats, 2013). Heat stress results in premature abortion of tapetal cells causing the pollen mother cells to rapidlly progress toward meiotic prophase and undergo programmed cell death (PCD) resulting in pollen sterility (Sakata and Higashitani, 2008; Parish et al., 2012). For example, the structural abnormalities in developing microspores of snap bean anthers under heat stress were associated with degenration of tapetum as a result of malformations in the ER (Suzuki et al., 2001). Heat stress caused reduction in dehiscence of anthers, accompanied by closure of the locules, and thus decrease in pollen dispersal in several crop plants (Peet et al., 2002). Exposure to high temperature after fertilization can impair subsequent embryo development (Barnabás et al., 2008). The reproductive failures in chickpea due to high-temperature stress are the result of disrupted sucrose metabolism in leaves as well as anthers (Kaushal et al., 2013).

The female gametophyte in plants is also called the embryo sac and is mostly a seven-celled structure (Thakur et al., 2010). Female gametophyte development occurs over two stages referred to as megasporogenesis and megagametogenesis. The female gametophyte is less sensitive to heat stress than the male gametophytic (Kaushal et al., 2013, 2016). Elevated temperatures probably inhibit style length and consequently induce abnormalities in ovary development, as observed in chickpea (Srinivasan et al., 1999). Temperatures >30°C reduce stigmatic receptivity and stigmatic pollen germination (Harsant et al., 2013), stigma and style growth (Snider et al., 2011; Song et al., 2015), and ovule penetration (Saini et al., 1983). Heat stress abruptly affects almost all aspects of female gametophyte development, e.g., reduced stigma receptivity in chickpea at 40/30 and 45/35°C (Kumar et al., 2013), and reduced ovule number and viability in common bean at 30°C (Suzuki et al., 2001). The female gametophyte produces important cells within the ovule viz. egg, central cell and synergids, which are developed by mitotic divisions (Sage et al., 2015). Synergids produce attractants into the micropylar end that guide pollen tube growth to the ovule (Chae and Lord, 2011). Heat stress alters the secretion of pollen tube attractants (Higashiyama et al., 1998), and reduces penetration of the ovule by the pollen tube (Saini et al., 1983). The effects of heat stress on expansion, division, and differentiation of egg and synergids in female gametophytes have been reported in bean (Sage et al., 2015).

Both male and female plant parts coordinate to make certain the deposition of pollen when the stigma becomes receptive, and this involves appropriate positionining of anthers nearby to the stigma for capturing the pollen after dehiscence (Sage et al., 2015). Heat stress disrupts this coordination by changing the structural positioning of anthers related to the stigma, the timing of dehiscence of anthers, and maturation and recetivity of stigma/style due to alteration in cell division and elongation (Basra, 2000; Giorno et al., 2013; Sage et al., 2015). These changes ultimately prevent pollen deposition on the stigma and alter the fertilization process.

For establishment of seed, the pollen grains must interact with a receptive stigma, followed by pollen tube growth to reach the ovules for fertilization, and embryo and endosperm development (Barnabás et al., 2008). Some of these events may be impacted by the adverse environmental conditions frequently encountered by crop plants (Driedonks et al., 2016). High temperature may arrest fertilization by inhibiting the development of male (Jain et al., 2007) and female gametophytes (Snider et al., 2009). Reduced fertilization is a common problem associated with heat due to disruption of meiosis and fertilization in various species, such as chickpea, cowpea, and barley (Kaushal et al., 2013; Jagadish et al., 2014; Bac-Molenaar et al., 2015; Driedonks et al., 2016). Reduced fertilization efficiency due to heat stress has been attributed to increasing oxidative stress, reduced carbohydrates, ATP concentration in gynoecium and decreased leaf photosynthesis, in mungbean (Suzuki et al., 2001), soybean (Board and Kahlon, 2011), and chickpea (Kumar et al., 2013). High temperatures during pollen development limit fertilization and seed development (Porch and Jahn, 2001) by reducing the number of mature pollen grains available for pollination (Erickson and Markhart, 2002; Sato et al., 2002), which causes abnormal pollen development, and reduces the viability and germinability of available pollen grains (Firon et al., 2006; Sato et al., 2006; Jain et al., 2007).

Heat stress (>30°C) from early meiosis to pollen maturity reduces the viability of pollen grains in chickpea resulting in fertilization failure leading to reduced seed set (Saini and Aspinall, 1981; Kaushal et al., 2016). Heat stress results in abnormal anther morphology and limits anther dehiscence at anthesis (Dupuis and Dumas, 1990), and prevents the accumulation of carbohydrates in developing anthers and pollen grains, which accounts for poor pollen viability at anthesis (Porch and Jahn, 2001; Kaushal et al., 2013). Gross and Kigel (1994) reported that high temperatures of 27/32°C at sporogenesis reduced pollen viability and yield in heat-sensitive genotypes of bean, due to failed anther dehiscence, pollen sterility, low pod and seed set,. In soybean, high temperatures of 38/28°C during flowering reduced in vitro pollen germination. Pollen grains were deformed, with a thicker exine and a disintegrated tapetum layer (Djanaguiraman et al., 2013). In chickpea, heat stress late in the season produced more structural abnormalities in anthers and pollen grains such as changes in anther locule number, anther epidermis wall thickening and pollen sterility in sensitive genotypes ICC-4567, ICC-10685 (Devasirvatham et al., 2013). Temperatures of 35/20 and 40/25°C pre- and post-anthesis reduced pollen viability, pollen production per flower and percentage of pollen germination in chickpea (Devasirvatham et al., 2012b). The effects of heat stress on both male and female reproductive tissue in some legume crops are shown in Table 2.

Temperature fluctuations during seed filling drastically reduce yield in legumes such as common bean (Rainey and Griffiths, 2005), pea (McDonald and Paulsen, 1997), chickpea (Kaushal et al., 2013; Kumar et al., 2013), mungbean (Kaur et al., 2015), lentil (Barghi et al., 2012; Bhandari et al., 2016; Sita et al., 2017), and cowpea (Ahmed et al., 1992). Seed filling is the completion of growth and development in crop plants, which involves transport processes to import constituents and biochemical processes related to the synthesis of carbohydrates, proteins, and lipids in seeds (Yang and Zhang, 2006; Awasthi et al., 2014). High-temperature stress causes yield loss in legumes (Canci and Toker, 2009; Kumar et al., 2016) and other crops due to poor seed development (Hall, 2004). Moreover, heat sensitivity differs for different crop species (Sung et al., 2003); on average, a one-degree rise in temperature will reduce plant yield by at least 10%. Under high temepratures, seed filling is accelerated, to reduce the duration of this stage to limit the yield potential (Boote et al., 2005). The reduction in starch accumulation was suggested to be the primary reason of yield reduction since starch acumulation accounts for substantial dry weight of the seeds. The reduction in seed weight in response to heat stress during the early stages of seed filling can be attributed to fewer endosperm cells (Nicolas et al., 1985), while during the later stages, heat stress impairs starch synthesis by limiting the supply of assimilates to the seed (Blum, 1998) or directly affecting the synthetic processes in the seed (Yang et al., 2004).

The number of endosperm cells determined early in grain fill, and the final size of the cells influence the extent of starch and protein accumulation in each seed, the rate and duration of grain fill also affect the accumlation of the seed reserves (Egli, 1998; Barnabás et al., 2008).

Reductions in various yield attributes due to heat stress has been reported in many crops such as cowpea (Hall, 1992), pea (Guilioni et al., 1997), common bean (Vara Prasad et al., 2002; Rainey and Griffiths, 2005), peanut (Vara Prasad et al., 1999a, 2000), soybean (Board and Kahlon, 2011), lentil (Barghi et al., 2012), and chickpea (Krishnamurthy et al., 2011; Kaushal et al., 2013; Kumar et al., 2013).

High-temperature stress reduces seed size due to the insufficient accumulation of photosynthates during seed filling (Kumar et al., 2016). A few days of heat stress (30–35°C) during seed filling accelerates senescence, decreases seed set and seed weight, and reduces yield in legumes (Siddique et al., 1999; Kumar et al., 2016). High yield losses have been reported in snap bean under heat stress (Tsukaguchi et al., 2003). Gutiérrez-Rodríguez et al. (2003) studied the biomass and yield of bean plants raised in two different seasons, i.e., winter and summer, and found that the winter-sown crop had 41 and 38% higher biomass and yield, respectively than the summer-sown crop. In soybean, even short-term exposure to stressful temperatures above 40°C reduced seed production and yield (Kitano et al., 2006; Board and Kahlon, 2011; Djanaguiraman et al., 2011). Vara Prasad et al. (2006) reported that increasing temperatures from 32/22°C to 36/26°C and 40/30°C, reduced seed yield in sorghum (Sorghum bicolor) by 10 and 99%, respectively. High-temperature stress increased the percentage of shriveled seed and reduced seed size in common bean (Vara Prasad et al., 2002) and groundnut (Prasad et al., 2003). In chickpea, Jumrani and Bhatia (2014) reported that increased temperatures during the reproductive stage severely reduced yield (by 10, 23, 64, and 78%) at different temperature ranges (26/16, 30/18, 34/20, and 38/28°C), respectively. Kaushal et al. (2013) observed a 63–64% yield reduction in chickpea exposed to 32/20°C. In similar studies, chickpea yields declined by 34–50% (Gan et al., 2004) and 34% (Wang et al., 2006) when exposed to temperatures >32/20°C. Other studies have reported inhibitory effects of high temperature on yield in pea (McDonald and Paulsen, 1997), cowpea (Ismail and Hall, 1999; Thiaw and Hall, 2004), peanut (Prasad et al., 2003), soybean (Puteh et al., 2013), field pea (Vijaylaxmi, 2013), faba bean (Kirra et al., 2014), mungbean (Kaur et al., 2015), and lentil (Bhandari et al., 2016; Sita et al., 2017).

During seed filling, carbohydrates, proteins, and lipids accumulate in developing seeds (Thakur et al., 2010). Heat stress alters the activities of carbon metabolism enzymes, starch accumulation, and sucrose synthesis by down-regulating specific genes in carbohydrate metabolism (Ruan et al., 2010). Plant hormones such as ABA and cytokinins play an important role in the regulation of seed filling (Brenner and Cheikh, 1995). These phytohormones are involved in the determination of sink size and strength, and the capacity of the seed to accumulate biomass (Thakur et al., 2010). Auxins, gibberellins, and ABA mediate cell division, enlarge endosperm cells, and regulate the direction and rate of assimilate flow from source to sink tissues (Hansen and Grossmann, 2000). Heat stress can influence seed filling by changing the concentration and amount of phytohormones as well as the expression of enzymes (Thakur et al., 2010). Low leaf photosynthetic rates during seed filling in heat-stressed plants are a major cause of reduced seed size (Singh, 1987; Leport et al., 1998). The accumulation of various seed components (mainly starch and proteins) may be inhibited by heat stress due to the inactivation of enzymatic processes involving starch (Ahmadi and Baker, 2001) and protein synthesis (Triboï et al., 2003).

Auxins regulate reproductive processes; in plants, a naturally occurring auxin is indole-3-acetic acid (IAA) (Ozga et al., 2017). However, in legume species, particularly those in the Fabaceae family such as pea, grass pea (Lathyrus sativus L.), lentil and faba bean, also contain the naturally occurring chlorinated form of auxin, 4-chloroindole-3-acetic acid (4-Cl-IAA), which is biologically more active than IAA in auxin bioassays (Reinecke, 1999; Ozga et al., 2017). Heat stress suppresses auxin biosynthesis and signaling in developing anthers, resulting in pollen abnormalities (Higashitani, 2013; Ozga et al., 2017). Similarly, gibberellins play an important role in stamen and pollen development (Plackett et al., 2012). Some studies have revealed that jasmonic acid signaling is required for stamen development and fertility because stamen development can be restored only in jasmonic acid biosynthesis mutants by exogenous jasmonic acid (Yan et al., 2007). Elevated temperature stress affects ethylene biosynthesis/signaling pathways in developing anthers, which leads to reduced anther dehiscence. Pollen development and pollen germination can be enhanced by pre-treatment with an ethylene-releasing agent, ethephon (Firon et al., 2012).

At the stage of fruit set, high temperature reduces auxin flux through the pedicel, allowing ethylene-facilitated pedicel abscission and fruit abscission/loss (Ozga et al., 2017). Developing seeds of pea and pericarp contains GAs and auxins (4-Cl-IAA and IAA) (Rodrigo et al., 1997; Ozga et al., 2017). Heat stress leads to seed abortion by altering seed-derived auxins and other seed signaling molecules transported to the pericarp, potentially having a negative effect on pericarp growth and facilitating pedicel abscission.

Elevated temperatures during seed filling and maturation can increase the proportion of seeds that are shriveled and abnormal at physiological maturity and result in seeds that exhibit reduced germination and seedling vigor in soybean (Egli et al., 2005). Furthermore, in legumes such as soybean, heat stress leads to the retention of chlorophyll in mature seeds, which can reduce seed quality (Teixeira et al., 2016). Low leaf photosynthetic rates during seed filling in heat-stressed plants are a major cause of reduced seed size (Awasthi et al., 2014). The accumulation of various seed components (mainly starch and proteins) may be inhibited by heat stress due to inactivation of enzymatic processes involving starch (Ahmadi and Baker, 2001) and protein synthesis (Triboï et al., 2003). Reduced seed weight was associated with reduced starch biosynthesis enzyme activities (ADP-glucose pyrophosphorylase and soluble starch synthase) in the endosperm during seed filling when the temperature increased above a threshold level (Singletary et al., 1994). Heat stress also reduces invertase activity associated with reduced carbon degradation (from sucrose to hexose) and partitioning (to starch synthesis) within endosperm, rather than being associated with limited carbon supply to the seed (Ozga et al., 2017). The legume embryo, being a strong terminal sink for sucrose, is not vascularly connected to the maternal seed coat tissue (Hardham, 1976). In faba bean, Weber et al. (1996), proposed a model for invertase-mediated unloading of sucrose for legume embryos during early seed development. Heat stress interrupts seed invertase activity and may alter nutrient portioning and seed growth and maturation in legumes (Ozga et al., 2017). During seed development and maturation, hormone regulation plays an important role in legume (Jameson and Song, 2016). Heat stress reduces cytokinin levels in seed leading to reduced seed cell numbers and growth rates (Emery et al., 2000; Jameson and Song, 2016). According to Yang et al. (2016), treatment with CK (6-benzylaminopurine) diminishes the inhibitory effect of heat stress on seed filling rate, division rate of endosperm, endosperm cell number, and seed weight in soybean. Heat stress regulates GA biosynthesis and catabolism in developing seeds to reduce GA-associated seed growth and development processes (Ozga et al., 2017). High-temperature stress increases the levels of ethylene, leading to reduced growth and enhanced senescence and abscission of various plant organs (Kukreja et al., 2005; Abeles et al., 2012). Heat stress induces ethylene, which can reduce photosynthesis and grain filling rates, and cause embryo abortion in some crops such as wheat (Rajala and Peltonen-Sainio, 2001; Hays et al., 2007). The effects of heat stress on different growth hormones at various reproductive developmental stages in legumes are listed in Table 3.

There are limited studies on the response of stage-specific functional physiology from flowering and post-flowering in legumes during high-temperature stress. Though the susceptibility to heat stress in plants varies with plant development, the reproductive stage due to its sensitive organelle makeup is bound to experience greater impact and surrender to temperature vagaries. The response depends upon the species and genotype, with profuse inter- and intra-specific differences (Sakata and Higashitani, 2008; Bita, 2016). Heat stress alters photosynthesis and respiration to shorten the life cycle and thus to reduce the plant productivity (Barnabás et al., 2008). A reduction of photosynthesis will in due course deplete the energy reserves and limit the availability of resources for reproduction in parental and gametophytic tissues (Sumesh et al., 2008). Heat stress often hastens the onset of anthesis, to start the reproductive phase of development before ample resources are gathered (Zinn et al., 2010). Several genes are alterted under high-temperature stress to result in degenration of tapetum and pollen sterility in many plant species (Oshino et al., 2007; Endo et al., 2009). Elevated temperatures target the enzymes involved in carbohydrate metabolism (e.g., cell wall, vacuolar invertase, and sucrose synthase) and sugar-transporters to reduce the pollen viability (Hedhly, 2011). Particularly, enzymes invertase and sucrose synthase isomorphs are down-regulated, which affects the turnover of starch and sucrose in pollen grains to decrease accumulation of soluble carbohydrates (Hedhly, 2011).

Male sterility has been observed in many heat-stressed food legumes, such as chickpea (Kaushal et al., 2013) and mungbean (Kaur et al., 2015), and impaired pollen development has been a vital reason linked to yield losses due to heat stress (Wassmann et al., 2009). Anthers developing under high temperature showed cell-proliferation arrest, distended vacuoles, altered chloroplast development and mitochondrial abnormalities (Sakata et al., 2010). Heat stress decreases accumulation of carbohydrates in pollen grains and stigmatic tissue by changing partitioning of the assimilates and the proportion between symplastic and apoplastic loading of the phloem (Taiz and Zeiger, 2006), which affects pollen viability (Kaushal et al., 2013). Heat stress decreases the activity of sucrose synthase and many cell wall and vacuolar invertases in developing pollen grains; as a result, the turnover of sucrose and starch turnover is impaired to reduce the accumulation of soluble carbohydrates in chickpea (Sato et al., 2006; Kaushal et al., 2013). Similar findings have been reported in chickpea (Devasirvatham et al., 2012a; Kaushal et al., 2013), lentil (Bhandari et al., 2016), and mungbean (Kaur et al., 2015). Heat stress also decreases the starch, protein and total oil yield in many crop species such as soybean (Rotundo and Westgate, 2009; Thuzar et al., 2010), and has been linked to high temperatures during seed development (Banowetz et al., 1999). Thus, for crop production under elevated temperatures, it is highly desirable to know which developmental stages and plant processes are most sensitive to heat stress, as well as whether high day or high night temperatures are more detrimental.

Among the components of a plant cell, plasma membranes are considered the most heat-sensitive, as they are the primary sites of injury under heat stress (Blum, 1988; Wise et al., 2004). Elevated temperature severely affects membrane structure and function, thereby increasing the fluidity of membranes due to protein denaturation and increased unsaturated fatty acids, causing a phase transition from solid gel structure to flexible liquid crystalline structure (Wahid et al., 2007). Due to the presence of double bonds in fatty acids (unsaturated state), these are less tightly packed into a membrane (Horváth et al., 2012), which facilitates the activation of lipid-based signaling cascades, elevated Ca²⁺ influx and reorganization of cytoskeletal (Ruelland and Zachowski, 2010; Bita and Gerats, 2013). Heat stress injury can be determined by loss of membrane integrity due to structural modifications of component proteins, which increase membrane thermostability and leakage of organic and inorganic ions from cells (Salvucci and Crafts-Brandner, 2004). Therefore, an electrolyte leakage value serves as an indicator of membrane damage and reflects stress-induced changes and has been used to evaluate membrane thermostability under high-temperature stress conditions (Liu and Huang, 2000; Xu et al., 2006). The effects of heat stress on membranes have been reported in various legume crops. In soybean, enhanced membrane permeability and electrolyte leakage was noticed under heat stress (Lin et al., 1984), which decreased the capacity of the plasma membrane to hold water and solutes. Similarly, membrane injury was noticed in chickpea genotypes, especially sensitive genotypes, at 40/30°C, which was further intensified at 45/35°C (Kumar et al., 2013). Chickpea is the most heat sensitive legume, as per observations based upon membrane thermostability and PSII function, compared with other grain legumes such as pigeon pea, groundnut, and soybean (Devasirvatham et al., 2012b). Other cool-season legumes such as faba bean and lentil have also been evaluated for membrane thermostability, which is closely related to plant heat tolerance (Ibrahim, 2011; Barghi et al., 2012). Membrane thermostability has been successfully employed to assess thermotolerance in many food crops worldwide.

Structural changes in thylakoid membranes with moderately high temperature stress have been observed using the freeze-fracture technique (Gounaris et al., 1984; Sharkey, 2005). The three major heat-sensitive sites in photosynthetic machinery are the photosystems, mainly photosystem II (PSII) with its oxygen-evolving complex (OEC), and the ATP generating and carbon assimilation processes (Mohanty et al., 2007; Murata et al., 2007). Photosystems I and II show damage under high temperature, with PSII more sensitive in chickpea (Kaushal et al., 2016). PSII in the electron trasnport chain of light reaction (Heckathorn et al., 2002) and RuBisCO activase in the carbon fixation cycle (Crafts-Brandner and Salvucci, 2000) are both sensitive to high temperature (Sinsawat et al., 2004; Kaushal et al., 2013). Heat stress damages the chlorophyll and photosynthetic apparatus by producing ROS (Guo et al., 2007; Bita and Gerats, 2013). In chickpea, Kumar et al. (2013) reported that damage to chloroplast membranes, mainly due to deterioration of photosynthetic pigments, reduced photosynthesis under high-temperature stress. A reduction in chlorophyll under elevated temperature has been reported in beans (Petkova et al., 2007) and chickpea (Kumar et al., 2013).

Higher temperature reduces the photosynthetic rate by decreasing leaf chlorophyll and nitrogen contents. In soybean, heat stress (38/28°C) significantly reduced chlorophyll content and, as a result, sucrose content. High-temperature stress reduces carbohydrate synthesis and carbohydrate transport from leaves; as a result, carbohydrates are diverted into vegetative organs at the expense of reproductive organs (Plaut et al., 2004; Suwa et al., 2010; Zhou et al., 2016). Heat stress negatively affects photosynthesis, carbohydrate synthesis, and flower and bud numbers, and ultimately leads to reduced sucrose content, the primary end product of photosynthesis translocated to reproductive organs (Lalonde et al., 1999). Leaf photosynthesis directly affects sucrose import into reproductive organs (Boyer and McLaughlin, 2007). Sucrose import and utilization are affected by invertase activity (breaks down sucrose), which regulates carbon allocation and sugar signaling (Roitsch and González, 2004), and could affect flower and fruit set due to high-temperature stress (Zhou et al., 2016), as observed in chikcpea (Kaushal et al., 2013). The effects of high temperature on the process of photosynthesis in some legume crops are listed in Table 4.

Heat stress is frequently associated with rapid loss of water from the plant surface resulting in dehydration (Koini et al., 2009). Heat-induced hikes in transpiration and water movement is a necessary tool for plant survival under extreme temperatures (Kolb and Robberecht, 1996; Hasanuzzaman et al., 2013). Increased transpiration during the day siphons out excess moisture from plants resulting in reduced turgor pressure and ultimately disturbed key physiological processes (Tsukaguchi et al., 2003). High-temperature stress influences plant water relations due to the faster depletion of water from soil profiles which affects soil temperatures and transpiration. High-temperature stress indirectly affects osmotic adjustments through impaired photosynthesis (especially damage to PSII), increased respiration, reduced leaf osmotic potentials, and decreased sugar concentrations (Huve et al., 2005; Vara Prasad et al., 2008). In snap bean (Phaseolus vulgaris), under heat stress, loss of water during the day time was more common because of increase in trasnpiration than night time, resulting in generation of water deficit stress (Tsukaguchi et al., 2003). Leaf transpiration rate increases once the threshold temperature is reached increase leaf cooling under heat stress (Levitt, 1980). High stomatal conductance under heat stress enhances transpirational heat dissipation in tolerant chickpea genotypes as long as soil water is available (Kaushal et al., 2013). However hastening drought stress will have further physiological implications, not least on photosynthesis (Liu et al., 2004). On the other hand, under severe heat stress, stomatal conductance decreases markedly, as in tobacco (Tan et al., 2011) to agagravate the damage to leaves.

Drought and heat stress conditions in the semi-arid tropics restricted nitrogen fixation efficiency (Naya et al., 2007). Elevated temperatures can affect N₂ fixation directly or indirectly. Direct inhibition by temperature is a consequence of decreased nodule development (Dart and Mercer, 1965; Piha and Munns, 1987; Junior et al., 2005), functionality (Hernandez-Armenta et al., 1989) and accelerated nodule senescence whereas indirect inhibition is related to temperature effects on root hair formation, reduction of nodulation sites (Frings, 1976), and modified adherence of bacteria to root hairs (Frings, 1976). Heat stress impacts on rhizobia have been thoroughly studied (Lira et al., 2005). The maximum temperature for rhizobial growth ranges from 32 to 47°C (Hungria and Vargas, 2000). Rahmani et al. (2009) established that heat tolerance in Bradyrhizobium directly affects the symbiotic efficiency between Bradyrhizobium and the soybean host. All stages of legume–rhizobium symbiosis are susceptible to high temperature (Hungria and Vargas, 2000; Nehra et al., 2007; Yadav and Nehra, 2013). Hungria and Franco (1993), studied the effect of high-temperature exposure on nodulation and efficiency of N₂ fixation in common beans; under high-temperature treatment (35 and 38°C/8 h/day), nodules formed but were inefficient at N₂ fixation. The control plants (grown at 28°C), when exposed to even higher temperatures (40°C/8 h/day) at flowering, had reduced nitrogenase activity and N₂-fixation efficiency. No nodules formed in peanut at 40°C or soybean at 37°C (Hungria and Vargas, 2000). Therefore, the selection of temperature tolerant N₂-fixing rhizobial strains may be used as an efficient tool for mitigating temperature stress (Yadav and Nehra, 2013).

Traditional breeding programs focus on developing cultivars with high yield traits under non-stress conditions. Such efforts have helped to enhance crop production per unit area and increased total agricultural production (Warren, 1998). Plant improvement for heat stress tolerance through genetic engineering is an economically better solution for crop production under stressful conditions (Blum, 1988). Heat sensitivity varies across developmental stages which makes the development of thermotolerant crops a challenging task (Driedonks et al., 2016). While breeding approaches have made significant advances in generating heat-tolerant lines in various crops, the genetic basis and range of heat tolerance largely remain unrevealed, especially in legumes. Development of new varieties is time-consuming and costly; therefore, understanding heat tolerance mechanisms would facilitate in developing strategies for screening germplasm of various legumes for traits related to heat tolerance. Using and exploring wild varieties along with landraces in breeding will enhance genetic diversity in crops (Driedonks et al., 2016).

High-temperature tolerance through conventional breeding is one approach to minimizing the damaging effects of heat stress on crop yield (Krishnamurthy et al., 2011). Breeding programs are generally conducted in a climactic region having similarity to where the crop will be grown. For relatively hot regions, selection of breeding lines occurs under hot conditions (Mickelbart et al., 2015). This technique has been reasonably successful considering that crops grown in warmer regions are often more tolerant of high temperatures than those in cooler regions (Kugblenu et al., 2013; Gaur et al., 2014). The chickpea genotype ICCV 92944, which is heat tolerant in screening experiments, has been released in three countries (as JG14 in India, Yezin6 in Myanmar and Chinadesi2 in Kenya) (Gaur et al., 2014). Two faba bean varieties (Shendi and Manami) with heat tolerance have been released in Sudan (Gaur et al., 2014). A new variety of cowpea has been produced with higher grain yield during high temperatures in the reproductive stage (Ehlers and Hall, 1998). Many heat-tolerant genotypes of legumes have been developed using various conventional breeding methods. By using rapid generation advancement methods, heat-tolerant index and earlier empirical methods, tolerant chickpea genotypes have been developed (Gaur et al., 2008; Krishnamurthy et al., 2011). Heat-tolerant common bean has been developed using the stress tolerant index (STI), geometric mean (GM) and recurrent selection techniques (Porch, 2006). Sultana et al. (2012) developed heat-tolerant genotypes of lentil using rapid initial growth habit and earliness. Mungbean, pea and snap bean have also been made tolerant to heat stress using the temperature-induction response and pedigree methods, respectively (Porch and Hall, 2013; Bindumadhava et al., 2017). Other crops such as groundnut and cowpea have been developed for improved performance under elevated temperatures using varied conventional breeding methods namely solute leakage, chlorophyll fluorescence and STI (in the case of groundnut), cross combination, breeding, pedigree breeding/backcrossing, and pedigree method (cowpea only) (Patel and Hall, 1990; Hall, 1992, 1993, 2011; Craufurd et al., 2003; Lucas et al., 2013).

While conventional breeding has been successful in developing heat-tolerant lines, the physiological and genetic basis of improvement remains unsure. This prevents the identificationof molecular biomarkers which would help in screening germplasm for enhanced heat tolerance and permit effectual breeding of this complex trait. Moreover, in conventional breeding, the potential gain in tolerance to heat stress is restrained by low genetic diversity (Paran and van der Knaap, 2007). Genetic diversity exists for heat tolerance in legumes (Kumar et al., 2016; Bindumadhava et al., 2017). Legume breeding programs, with various classical breeding methods, have potential in the application of technology that could promote their global production.

Genetic and quantitative trait locus (QTL) mapping has become a successful method for detecting specific chromosome segments that have candidate genes for heat tolerance (Argyris et al., 2011; Zhang et al., 2012). To improve knowledge regarding heat tolerance at the genetic level, attempts have been made to identify QTLs in segregating mapping populations. Till now a wide range of QTLs governing heat tolerance has been discovered in cereal crops (Chen et al., 2008; Zhang et al., 2008; Kumar et al., 2010; Mason et al., 2010; Wei et al., 2013), but very few heat-tolerant QTLs have been reported so far in legumes mainly including cowpea (Lucas et al., 2013; Pottorff et al., 2014) and azuki bean (Kaga et al., 2003; Vaughan et al., 2005). QTLs for several traits related to heat tolerance have been identified, such as increased chlorophyll fluorescence and reduced canopy temperature during vegetative and reproductive stages in wheat (Vijayalakshmi et al., 2010; Lopes et al., 2012). Reduced canopy temperatures show that efficient water uptake is ultimately associated with deep rooting, and increased chlorophyll fluorescence reflects heat-tolerant photosynthesis (Pinto and Reynolds, 2015). Studies have been conducted on the effect of heat stress on reproductive characters, mainly pollen germination, pollen tube growth, grain filling, grain weight, fruit set and post-anthesis senescence of leaves (Driedonks et al., 2016). A QTL study on rice (Oryza sativa) recently focused on spike fertility under heat stress (Ye et al., 2015). This study was based on earlier work (Ye et al., 2012) and confirmed that a recessive QTL on chromosome 4 is present, which is responsible for a 15% increase in rice spikelet fertility under high temperatures (Ye et al., 2015). The use of a multiparent advanced generation inter-cross (MAGIC) population may lead to the introduction of more genetic variation and identification of thermotolerant genes for spikelet fertility (Ye et al., 2015).

Quantitative trait locus can also be dedicated to the investigation of natural populations. As observed earlier, linkage mapping may be able to detect crucial genes and QTLs. However, the restricted number of generations and recombination events often results in QTLs covering a comparatively large region and the identification of genes involves a tedious process of fine mapping (Driedonks et al., 2016). Therefore, fine mapping is generally inefficient for the detection of candidate genes (Bergelson and Roux, 2010). Different studies on QTLs revealed multiple QTLs per trait, ranging from two in azuki bean and rice (improved pollen viability and spikelet number under high temperatures, respectively) to 34 in barley for traits related to heat stress. As such, heat tolerance depends on a variety of factors and QTLs, which differ among the crops (reviewed in Jha et al., 2014). Kaga et al. (2008) identified HQTL-1 and HQTL-2 in azuki beans involving traits for pollen viability. In cowpea, many QTLs have been detected, in particular Hbs-1, Hbs-2, and Hbs-3 for heat-induced browning of the seed coat (Pottorff et al., 2014), afot 1.1, afot 1.2, afot 1.3 and afot 2 for flower opening (Andargie et al., 2013), and Cht-1, Cht-2, Cht-3, Cht-4, and Cht-5 for male heat sterility (Lucas et al., 2013). In pigeon pea, qPD4.1 have been detected for pods per plant, and qFL4.1 and qFL5.1 for flowering (Kumawat et al., 2012). Currently, association mapping is acquiring popularity as a trait mapping technique which complements conventional QTL mapping (Yu et al., 2008; Jha et al., 2017). Recently, GWAS (genome-wide association study) was carried out in a panel of 300 accessions to scrutinize the marker-trait association for thermotolerance in chickpea (Thudi et al., 2017). Therefore, to accelerate the transfer of heat tolerance causative gene(s)/QTL(s) in major grain legumes, available molecular markers can be used in marker-assisted breeding programs (Jha et al., 2017). Futher, unrivaled improvements in next-generation sequencing (NGS) has paved way to unfold the complex genomic regions which are important in regulating complex traits (Elshire et al., 2011; Edwards and Snowdon, 2013). Genotype-by-sequencing (GBS) is one such technology that produces large number of SNP markers (Elshire et al., 2011), which are applied to develop genetic maps and decipher complex traits in legumes (Jaganathan et al., 2015; Kujur et al., 2015; Tayeh et al., 2015; Verma et al., 2015). The rising availability of refrence genome sequences in many grain legumes such as mungbean (Kang et al., 2014), soybean (Schmutz et al., 2009), groundnut (Bertioli et al., 2016; Chen et al., 2016), chickpea (Jain et al., 2013; Varshney et al., 2013), adzuki bean (Kang et al., 2015; Yang et al., 2015), pigeonpea (Varshney et al., 2012a), and common bean (Teixeira et al., 2005; Schmutz et al., 2014), provide great endevours to focus on important agricultural traits including thermotolerance.

Quantitative trait locus analysis in heat-sensitive and tolerant crops is gaining attention. The main advantage of QTL-based approaches is that they allow loci linked to heat stress to be identified (Bita and Gerats, 2013). Identification of adaptive QTLs for heat stress is one way of understanding tolerance mechanisms, and various studies have been conducted to detect genetic markers for various abiotic stresses, including heat stress (Roy et al., 2011). Markers linked to QTLs could be used to enhance thermotolerance in available germplasm. Currently, QTL identification for thermotolerance is being carried out using different traits, such as thousand grain weight (TGW), canopy temperature depression (CTD), grain filling duration (GFD), yield (Pinto et al., 2010), and traits related to senescence (Vijayalakshmi et al., 2010).

Association genetics has recently been used to assist in QTL detection in various crop species (Ahuja et al., 2010). The markers associated with QTLs, once isolated, the candidate QTLs can be introgressed in elite lines via MAS technology. The traits are usually controlled by small effect QTLs or multiple pleiotropic genes which are the main drawback of generating tolerant genotypes for heat stress (Bita and Gerats, 2013). Marker-assisted recurrent selection (MARS), pyramiding various QTLs from a large number of populations in the same genetic background or GS (Genomic Selection) techniques can be used to overcome this (Tester and Langridge, 2010; Varshney et al., 2012b). The MAS approach, however, for complex traits such as heat tolerance are not efficient due to the genotype–environment and gene–gene interactions, which eventually lead to reduced breeding efficiency (Collins et al., 2008). When characters like heat stress tolerance are involved, recurrent selection is an adequate method in plant breeding. There is a small probability of obtaining a superior genotype in multiple crosses, which combines all of the required alleles. The substitute is recurrent selection to accumulate gradually, through recombination cycles, the desired and available alleles in different parents (Donà et al., 2013). The main focus of recurrent selection is to enhance the frequency of desirable alleles for favorable traits, conserving genetic variability.

‘Omics’ technologies, such as genomics, proteomics, transcriptomics, and metabolomics, have revolutionized research in plant sciences (Yuan J.S. et al., 2008). The enormous progress in the field of “omics” has made possible to elucidate different candidate genes involved in response to complex abiotic stresses in crop plants (Vij and Tyagi, 2007; Urano et al., 2010; Deshmukh et al., 2014; Kujur et al., 2015). These technologies involve various disciplines, and new advances in these areas have markedly contributed to a better understanding of the molecular and genetic basis of the heat stress response that has been a crucial bottleneck for molecular and transgenic breeding (Reddy et al., 2012). As the technology has progressed, omics approaches have improved over the last decade (Deshmukh et al., 2014). Research in recent years has provided an understanding of the function of proteins, metabolites, and many key genes and molecular processes involved in plant responses to heat stress. The mechanism of heat stress tolerance, however, is quite complex because of the multiple genes and post-transcriptional regulation influence (Ramalingam et al., 2015). Moreover, gene expression is affected by stress conditions due to alterations in plant proteome and metabolome composition. Therefore, to understand plant stress tolerance, omics technology has become mandatory (Ramalingam et al., 2015).