Multiple benefits of legumes for agriculture sustainability: an overview

Food security, lowering the risk of climate change and meeting the increasing demand for energy will increasingly be critical challenges in the years to come. Producing sustainably is therefore becoming central in agriculture and food systems. Legume crops could play an important role in this context by delivering multiple services in line with sustainability principles. In addition to serving as fundamental, worldwide source of high-quality food and feed, legumes contribute to reduce the emission of greenhouse gases, as they release 5–7 times less GHG per unit area compared with other crops; allow the sequestration of carbon in soils with values estimated from 7.21 g kg−1 DM, 23.6 versus 21.8 g C kg−1 year; and induce a saving of fossil energy inputs in the system thanks to N fertilizer reduction, corresponding to 277 kg ha−1 of CO2 per year. Legumes could also be competitive crops and, due to their environmental and socioeconomic benefits, could be introduced in modern cropping systems to increase crop diversity and reduce use of external inputs. They also perform well in conservation systems, intercropping systems, which are very important in developing countries as well as in low-input and low-yield farming systems. Legumes fix the atmospheric nitrogen, release in the soil high-quality organic matter and facilitate soil nutrients’ circulation and water retention. Based on these multiple functions, legume crops have high potential for conservation agriculture, being functional either as growing crop or as crop residue. Graphical abstract . .


Introduction
Global population will hit 9.6 billion people by 2050 [108] and will face global challenges among which achieving food security, lowering the risk of climate change by reducing the net release of greenhouse gases into the atmosphere and meeting the increasing demand for energy are the most critical ones.In particular, the impact of climate change and associated biotic and abiotic stresses to which crop systems will be increasingly exposed pose serious implications for global food production [119].
To meet these challenges, a policy framework needs to be developed in which the sustainability of production/ consumption patterns becomes central.In this context, food legumes and legume-inclusive production systems can play important roles by delivering multiple services in line with sustainability principles.Indeed, legumes play central roles [112]: (1) at food-system level, both for human and animal consumption, as a source of plant proteins and with an increasingly importance in improving humans health [106]; (2) at production-system level, due to the capacity to fix atmospheric nitrogen making them potentially highly suitable for inclusion in low-input cropping systems, and due to their role in mitigating greenhouse gases emissions [53]; and (3) at croppingsystem levels, as diversification crops in agroecosystems based on few major species, breaking the cycles of pests and diseases and contributing to balance the deficit in plant protein production in many areas of the world, including Europe [43,48,72,78,116].
Leguminosae family comprises 800 genera and 20,000 species [54] and represents the third largest family of flowering plants.Some legumes are considered weeds of cereal crops, while others are major grain crops; these latter species are known as grain legumes, or pulses, 1 and represent the focus of this review.For some of these species, the trends for word acreage and yield are available, as reported in Table 1.
Despite the growing trends observed during the 50-year period between 1974 and 2014 for some warm-season legumes (e.g.soybean, cowpea, dry bean, groundnut, and pigeon pea), the acreages of several temperate legumes (e.g.pea, faba bean, lupin, french bean and vetch) have declined worldwide with differences between world Regions (Table 3).In any case, food legumes occupy a minimal part of arable land, mostly dominated by cereal crops [99]; soybean represents the most important and cultivated legume, acreage of which reached 117.72 million ha in 2014 (steadily increased over years, see also Table 3), which is about that of the other grain legumes, but still far below the major cereals (e.g.rice, wheat, maize).Such trend is mainly associated to the expansion of more specialized and intensive production systems [82]. 1 Soybean and groundnuts are not defined by FAO as 'pulse crops' .
Market forces stimulating specialization of cropping systems as non-marketable benefits of diversification, like cultivation/introduction of legumes in the farming system, do not deliver immediate and/or apparent profits [82].This is, however, not equally perceived throughout the globe, and there is indeed a remarkable diversity in grain legumes' production trends across the world (Table 3).
The European decline in grain legume's production is not mirrored by other regions of the world such as Canada or Australia, where legume's cultivation has been increasing over the last few decades.In these areas, monoculture of cereals, which relies on frequent summer-fallowing and use of mechanical tillage, has been replaced by extended and diversified crop rotations together with the use of conservation tillage [122].Furthermore, supply chains and markets are inadequately developed for most legume crops (see also [66], for France) with the exception of soybean, for which the global market is well developed [85].Nevertheless, soybean areas in Europe are constrained by climatic factors although there is considerable potential to develop new varieties suitable to flourish under cool growing conditions [123].The low diffusion of legumes' cultivation is also due to reduced and unstable yields and susceptibility to biotic and abiotic stress conditions; the average yields for unit area have increased (soybean +86%, lentil +77%, groundnut +75%, chickpea +70%) less than cereal crops (+104%, on average) (Table 1).Moreover, legume cultivation depends not only on the effect of farmers' choices, although they play a central role for such decision, but also on policymakers who have the responsibility to provide effective strategies to support the integration of legumes into cropping systems.This aspect is particularly relevant if the overall objective for future agricultural systems is to promote sustainability, improve resource use efficiency and preserve the environment [82].

Grain legumes impacts on atmosphere and soil quality
Among the many important benefits that legumes deliver to society, their role in contributing to climate change mitigation has been rarely addressed.Legumes can (1) lower the emission of greenhouse gases (GHG) such as carbon dioxide (CO 2 ) and nitrous oxide (N 2 O) compared

Greenhouse gas emissions
The introduction of legumes into agricultural rotations help in reducing the use of fertilizers and energy in arable systems and consequently lowering the GHG emissions [52].N fertilizer savings across Europe [51], in rotations including leguminous crops, range around 277 kg ha −1 of CO 2 per year (1 kg N = 3.15 kg CO 2 , [42]. It has been reported that half of the CO 2 generated during NH3 production would be reused if the NH3 was converted to urea.This is, however, only a time shift of CO 2 release in the atmosphere since, once the urea is applied to the soil, the hydrolyzation activity by urease will release CO 2 originally captured during urea production [39].Considering an efficiency of 2.6-3.7 kg CO 2 generated per kilogram of N synthesized, the annual global fertilizer leads to a release of 300 Tg of CO 2 into the atmosphere each year [42].Some studies indicate that at global scale, the amount of CO 2 respired from the root systems of N 2 -fixing legumes could be higher than the CO 2 generated during N-fertilizer production [42].However, it is important to emphasize that the CO 2 respired from nodulated roots of legumes comes from the atmosphere through the photosynthesis activity.Conversely, all the CO 2 released during the process of N-fertilizer synthesis derives from fossil energy, thus determining a net contribution to atmospheric amount of CO 2 [42].N 2 O represents about 5-6% of the total atmospheric GHG, but it is much more active 2 than CO 2 [21].Agriculture represents the main source of anthropogenic N 2 O emissions (about 60%; [84], due to both animal and crop production [38]).A majority of these emissions result from the application of nitrogen fertilizers [84]: every 100 kg of N fertilizer about 1.0 kg of N is emitted as N 2 O [42], although different amounts depend on several factors including N application rate, soil organic C content, soil pH, and texture [78,88].Denitrification processes are the most important source of N 2 O in most cropping and pasture systems [76,88,102].
In the recent years, several studies have focalized on the role of legumes in the reduction of GHG emissions.Jeuffroy et al. [44] demonstrated that legume crops 2 N 2 O absorbs approximately 292 times as much infra-red radiation per kilogram as CO 2 .

Table 3 Trend for Region Δ acreage (%) during the 50-year period starting from 1974 to 2014 for legume crops included in FAOSTAT classification [23]; the major three cereal crops are also reported, for comparison
In Table 2 Nevertheless, it is important to highlight that the influence of legumes in reducing GHG depends also on the management of agro-ecosystems in which they are included.When faba bean was grown as mono cropping, it led to threefold higher cumulative N 2 O emissions than that of unfertilized wheat (441 vs. 152 g N 2 O ha −1 , respectively); conversely, when faba bean was mixed with wheat (intercropping system), cumulative N 2 O emissions fluxes were 31% lower than that of N-fertilized wheat [96].Anyway, the benefits derived from the introduction of legumes in crop rotations become significant when commercially relevant rates of N fertilizer are applied [42].
The mitigation in terms of GHG emissions is also obtained by adopting sustainable agricultural systems, such as conservation tillage and conservation agriculture systems, which are suitable for the cultivation of both grain and green-manure legumes (see "Grain legumes and conservation agriculture" section).
In conclusion, it is noteworthy to underline that field tests and experimental analyses on GHG emissions, and in particular on N 2 O, provided quite different results [89] due to the influences of differences of several variables, including climatic, soil and management conditions [45,78,88].
In general, N 2 O losses from soils covered with legumes are certainly lower than those from both N 2 O fertilized grasslands and non-legume crops, as also indicated by Jensen et al. [42] who report a mean of 3.22 kg N 2 O-N ha −1 , calculated from 67 site years of data.In addition, there is no direct association between N 2 O emissions and biological nitrogen fixation [42], since organic N from legume residues is decomposed, mineralized and rapidly immobilized by microorganisms [78].Emissions of N 2 O could occur either during nitrification or due to denitrification, being affected by timing of mineralized N supply [20]: the asynchrony between N supply and utilization from the following crops enhances N loss, especially in winter/early spring in cold wet soils [64].

Soil properties
Cultivation and cropping may cause significant SOC losses through decomposition of humus [18].Shifting from pasture to cropping systems may result in loss of soil C stocks between 25 and 43% [101].
Legume-based systems improve several aspects of soil fertility, such as SOC and humus content, N and P availability [42].With respect to SOC, grain legumes can increase it in several ways, by supplying biomass, organic C, and N [27,53], as well as releasing hydrogen gas as by-product of BNF, which promotes bacterial legume nodules' development in the rhizosphere [49].
In sandy soils, the beneficial effect of grain legumes after three years of study was registered in terms of higher content of SOC compared with soils with oats (7.21 g kg −1 DM, on average).Specifically, cultivation of pea exerted the most positive action to organic carbon content (7.58 g kg −1 , after harvest, on average), whereas narrow-leaved lupin had the least effect (7.23 g kg −1 , on average) [30].In southern America (Argentina), the intercropping of soybean with maize at different rates favoured a SOC accumulation of 23.6 g C kg −1 versus 21.8 g C kg −1 of the sole maize; the greatest potential for enhancing SOC stocks occurred in the 2:3 (maize:soybean) intercrop configuration [11].Furthermore, just only amending the soil with soybean residues allows to obtain an increase of 38.5% in SOC [11].
Thanks to BNF, legumes also affect significantly soil N availability; by using legumes as winter crops in ricebean and rice-vetch combination, rice residue N content is enhanced by 9.7-20.5%,with values ranging from 1.87 to 1.93 g N kg −1 soil [120].It needs to be underlines that a majority of studies on the role of legumes for soil N fertility have investigated the shoot N content.In this regard, Carranca et al. [15] found that 7-11% of total legume N was associated with root and nodules and an allocation of 11-14 kg N fixed t −1 belowground dry matter, representing half the amount of total aboveground plant.
In intercropping cowpea-maize, Latati et al. [50] found an increase in P availability at rhizosphere level associated with significant acidification (−0.73 U) than in sole cropping.Wang et al. [115], assessing properties related to N and P cycling in the rhizosphere of wheat and grain legumes (faba bean and white lupin) grown in monoculture or in wheat/legume mixtures, found that the less-labile organic P pools (i.e.NaOHextractable P pools and acid-extractable P pools) significantly accumulated in the rhizosphere of legumes.However, the P uptake and the changes in rhizosphere soil P pools seem to depend also on legume species.Compared with the unplanted soil, the depletion of labile P pools (resin P and NaHCO 3 -P inorganic) was the greatest in the rhizosphere of faba bean (54 and 39%) with respect to chickpea, white lupin, yellow lupin and narrow-leafed lupin [31].Of the less-labile P pools, NaOH-P inorganic was depleted in the rhizosphere of faba bean, while NaOH-P organic and residual P were most strongly depleted in the rhizosphere of white lupin [31].
Also in North Rift, Kenya Region, in well-drained, extremely deep, friable clay, acid humic top soil, the effects of cultivation and incorporation of lupine and garden pea were significant in terms of soil-available P with respect to fallow, with lupine showing higher P availability than pea (from 20.3 to 31.0%higher).
Although there is a general agreement on the influence of grain legumes on rhizosphere properties in terms of N supply, SOC and P availability, the magnitude of the impact varied across legume species, soil properties and climatic conditions.Among these, soil type represents the major factor determining plant growth, rhizosphere nutrient dynamics and microbial community structure.The pattern of depletion and accumulation of some macro-and micronutrients differed also between cropping systems (i.e.monoculture, mixed culture, narrow crop rotations) as well as among soil management strategies (i.e.tillage, no-tillage).

Role of grain legumes in cropping systems
Legumes could be competitive crops, in terms of environmental and socioeconomic benefits, with potential to be introduced in modern cropping systems, which are characterized by a decreasing crop diversity [24,80] and an excessive use of external inputs (i.e.fertilizers and agrochemicals).

Grain legumes into crop-sequences
In the recent years, many studies have focused on the sustainable re-introduction of grain legumes into crop rotations, 4 based on their positive effects on yield and quality characteristics on subsequent crops [46,82,103].However, assessment of the rotational advantages/disadvantages should be based on a pairwise comparison between legume and non-legume pre-crops [82].Some experimental designs involving multi-year and multispecies rotations do not provide information on yield benefits to the subsequent species in the rotation sequence.Therefore, it is difficult to formulate adequate conclusions [2].
The agronomic pre-crop benefits of grain legumes can be divided into a 'nitrogen effect' component and 'break crop effect' component.The 'nitrogen effect' component is a result of the N provision from BNF [77], which is highest in situations of low N fertilization to subsequent crop cycles [82].The second one (break crop effect) includes non-legume-specific benefits, such as improvements of soil organic matter and structure [34], phosphorus mobilization [98], soil water retention and availability [2], and reduced pressure from diseases and weeds [87].In this case, benefits are highest in cereal-dominated rotations [82].
Several authors have reviewed the yield benefits of legumes for subsequent cereal crops.
In Australia, Angus et al. [2] reported higher yield of wheat after legumes (field peas, lupins, faba beans, chickpeas and lentils) than those of wheat after wheat.In particular for a wheat-wheat yield of 4.0 t ha −1 , the mean grain legume-wheat yield was 5.2 t ha −1 (+30% on average).Other studies from Australia quantified yield benefits compared to pure cereal crop sequences at 40-50% for low N levels and 10-17% for high N levels [3].
In Europe yields benefits of grain legumes have been shown to strongly depend on climatic factors which affect N dynamics in soils [52].In temperate environments, cereals yield is on average 17 and 21% higher in grain-legume based systems than wheat monocropping, under standard and moderate fertilization levels, respectively [40].Conversely, yield benefits are lower in Mediterranean climates where water availability is the limiting factor to cereal yields [46,61,62].
The yield advantage to subsequent cereal crops provided by legumes depends also on the species and amounts of fixed N [114,121].Field pea and faba bean 4 According to Angus et al. [2], crop-sequences experiments can be classified into rotation experiments and break crop experiments.Rotation strictly defined, refers to a recurring sequence of crops, forages and fallows, or more loosely defined, to a cropping sequence that contains fallows, or crops and forages in addition to the locally dominant species.A break crop generally refers to a single alternative crop followed by the dominant species.
accumulate about 130 and 153 kg N ha −1 in their aboveground biomass, respectively [77] and significant quantities (30-60% of the accumulated total N) may also be stored in belowground biomass [77].Differences in BNF patterns are also found between the same species.For example, Mokgehle et al. [69] compared 25 groundnut varieties for plant BNF at three differing agro-ecologies in South Africa, highlighting N-fixed range between 76 and 188 kg ha −1 , depending also on soil and environmental conditions as well as on N-uptake.Other factors influencing BNF include salinity and sodicity (alkalinity) of soils, as observed in chickpea [83], common bean [22] and faba bean [109].
It is, however, rather difficult to quantify the legume dependent increase in N uptake in subsequent crops, versus other sources of N [46,77].In temperate environments of Australia, measurements of the additional N-nitrate available to wheat crops following legumes instead of cereals, averaged around 37 kg N ha −1 [17].In Denmark, nitrogen uptake in crops that follow legume crops has been reported to increase by 23-59% after field pea and narrow-leafed lupin on different soil types [40], but only 14-15% for durum wheat following vetch in a semi-arid Mediterranean environment [28].Increased N uptake of crops after grain legumes reached up to 61% or 36 kg ha −1 for a vetch-barley rotation in Cyprus [74].Further, some legume residues have beneficial effects on some quality aspects of the subsequent crops in southern Italy [104].
Among other beneficial effects brought about by legumes, the production of hydrogen gas (H 2 ) as a by-product of BNF greatly affects the composition of the soil microbial population, further favouring the development of plant growth-promoting bacteria [2].Some grain legumes, including chickpea, pigeon pea and white lupin can mobilize fixed forms of soil P through the secretion of organic acids such as citrate and malate and other P mobilizing compounds from their roots [36].Among grain legumes, white lupin most strongly solubilize P, a function that can be facilitated by its proteoid roots that may englobe small portions of soil [2].Glasshouse experiments using a highly P-fixing soil showed better wheat growth following white lupin than soybean [37], suggesting that the cereal was able to access P made available by the previous white lupin break crop.'Break crop' effects also include increased soil water content, since the break-crop stubble can affect retention of soil water and infiltration and retention of rain water [47].A species-specific response has also been documented.Soil profiles after pea field can be wetter than after a wheat crop [2].In Saskatchewan, Canada, Miller et al. [68] reported that post-harvest soil water status up to 122 cm-depth was 31 and 49 mm greater for all legumes (field pea, lentil and chick pea) with respect of wheat under loam and clay soils, respectively.This was primarily due to increased plant water use efficiency.Lentil in rotation with cereals has been shown to increase total grain production by increasing residual soil water in dry areas of Saskatchewan [25].
In general, grain legumes are not susceptible to the same pests and diseases as the main cereal crops (nonhost), resulting suitable as break crops in wheat-based rotations [121].Grain legumes as break crops can also contribute to weed control [97] by contrasting their specialization and helping stabilizing the agricultural crop weed community composition [7].
Despite the described beneficial effects, there are still concerns on the introduction of grain legumes into cropping sequences.Cropping systems that include legume crops in farm rotations must be supported by best cropmanagement practices (e.g.N fertilization rates and timing, soil management, weeding, irrigation), which often do not match standard techniques normally applied by farmers.For example, some possible risks in terms of nitrate leaching associated to grain legumes cultivation can be counteracted by including cover crops in the system [33,81].Additional reasons may explain why grain legumes are not very common in high-input cropping systems.These include (1) their low and unstable yields [16,86]; (2) inadequate policy support [14]; (3) lack of proper quantification (and recognition) of long-term benefits of legumes within cropping systems [82].However, other efforts could be addressed, for example, to breeding programs for improved crop cultivars, to better sustain livelihood and increase the economic return to farmers.Indeed, during last years significant progresses in breeding for quality traits for food [110] and feed uses [79], as well as for resistances to biotic [91] and abiotic stresses [4] are being achieved, but several others, many of which are controlled quantitatively by multiple genes, have been more difficult to achieve.

Grain legumes in intercropping
Intercropping systems consist in simultaneous growth of two or more crop species on the same area and at the same time [13].Intercropping is widely used in developing countries or in low-input and low-yield farming systems [73].Despite several recognized beneficial aspects of intercropping such as better pest control [60], competitive yields with reduced inputs [70,107], pollution mitigation [63], more stable aggregate food or forage yields per unit area [100], there are a number of constrains that make intercropping not common in modern agriculture, such as example the request of a single and standardized product and the suitability for mechanization or use of other inputs as a prerogative in intensive farming system [13].It is therefore necessary to optimize intercropping systems to enhance resource-use efficiency and crop yield simultaneously [55], while also promoting multiple ecosystem services (see also [13]).Most recent research has focalized on the potential of intercropping in sustainable productions and in particular on grain legumes that can fix N 2 through biological mechanisms (BNF).Indeed, legumes are pivotal in many intercropping systems, and of the top 10 most frequently used intercrop species listed by Hauggaard-Nielsen and Jensen [32], seven are legumes One of the basic spatial arrangements used in intercropping is strip intercropping, in which two or more crops grow together in strips wide enough to permit separate crop production using inputs but close enough for the crops to interact.The current challenge is how to determine an optimal intercropping width to maximise the resources use efficiency and, consequently, the crop productivity.In a maize-bean strip intercropping, Mahallati et al. [65] suggested that strip width of 2 and 3 rows was superior compared with monoculture and other strip intercropping combinations in terms of radiation absorption, radiation use efficiency and biological yields of both species, also allowing to an improve of total land productivity and land equivalent ratio (1.39 and 1.37).Gao et al. [26] showed a total yield increase of 65 and 71% in a system of 1 and 2 rows of maize (planted at a higher density in intercropping) alternated with 3 rows of soybean compared with both crops grown as monoculture.However, Liu et al. [59] showed a reduction in the photosynthetically active radiation and R:FR ratio at the top of soybean canopy intercropped with maize -under two intercropping patterns: 1 row of maize with 1 row of soybean; 2 rows of maize with 2 rows of soybeanleading to increased internode lengths, plant height and specific leaf area (SLA), but reduced branching of soybean plants.In order to gain sufficient light in the most shaded border rows of the neighbouring, shorter crops, efforts could be addressed to (i) the selection of highly productive maize cultivars with reduced canopy height and LAI; (ii) the increase of the strip width under a higher fraction of direct PAR; (iii) the selection of crops and cultivars suitable under the shade levels that likely occur in strip-intercropping systems with maize [71].
The increase in N availability in intercrops hosting legumes occurs because the competition for soil N from legumes is weaker than from other plants.Moreover, non-legumes obtain additional N from that released by legumes into the soil [56,117] or via mycorrhizal fungi [113].Legumes can contribute up to 15% of the N in an intercropped cereal [57], thus increasing biomass production and carry-over effects [75], reducing synthetic mineral N-fertilizer use and mitigating N 2 O fluxes [9,96].However, the adoption of grain legume intercropping systems should benefit from the identification of suitable legumes that are less susceptible to N fertilizerinduced inhibition of BNF-that is, legumes that sustain higher %BNF in the presence of increasing soil mineral N. To this purpose, Rose et al. [90] indicated that faba bean is more suitable as intercrop than chickpea when supplementary N fertilizer additions are required, with about 40%BNF and 29%BNF maintained in faba bean and chickpea, respectively, supplying both crops with 150 kg N ha −1 .
BNF represents the most common plant growth stimulating factor that can also improve crop competition with respect to weeds in both organic and sustainable farming systems [10].Grain legumes are weak suppressors of weeds, but mixing species in the same cropping system could represent a valid way to improve the ability of the crop itself to suppress weeds [41,94].In a wheat-chickpea intercropping system (20 cm spacing without weeding treatment) it was observed a 69.7% reduction in weed biomass and 70% in weed population as compared to un-weeded monocrop wheat at 20 cm spacing [6].Similar results on weed smothering have been obtained by Midya et al. [67] in rice-blackgram (20 cm) intercropping system although the deferred seeding of blackgram in rice field (30 cm) with one weeding may be recommended for both better yield and weed suppression.
Direct mutual benefits in cereal-legumes intercropping involve below-ground processes in which cereals while benefiting of legumes-fixed N, increase Fe and Zn bioavailability to the companion legumes [118].
Physiology, agronomy and ecology can simultaneously contribute to the improvement of intercropping systems, allowing to enhance crop productivity and resource-use efficiency, so making intercropping a viable approach for sustainable intensification, particularly in regions with impoverished soils and economies where measured benefits have been greatest [93].But to realize these goals, major efforts in research programs still remain.For example: (1) breeding for intercrops; (2) better understanding of the interactions between plants and other organisms in crop systems, focusing on the roles of above-and belowground interactions of plants with other organisms; (3) improving agricultural engineering and management, i.e. developing new machinery that can till, weed and harvest at small spatial scales and in complex configurations to encourage the uptake of intercropping without greater demands for labour [58]; (4) adoption of a wider 'systems thinking' through the enactment of schemes, including payment for ecosystem services [105].

Grain legumes and conservation agriculture
Legumes have some characteristics particularly suitable for sustainable cropping systems and conservation agriculture, and making them functional either as growing crop or as crop residue.Conservation agriculture is based on minimal soil disturbance and permanent soil cover combined with rotations [35].As previously described, major advantages of legumes include the amount of nitrogen fixed into the soil and the high quality of the organic matter released to the soil in term of C/N ratio.Some legume species have also deep root systems, which facilitate nutrients solubilization by root exudates and their uptake/recycling as well as water infiltration in deeper soil layers.
Many countries already rely on conservation agriculture.Brazil has implemented conservation agriculture systems using soybean as legume crop.Grain legumes like lentil, chickpea, pea and faba bean play a major role in conservation agriculture in North America, Australia, and Turkey.In Australia, some advantages of minimum tillage for grain legumes have been quantified for waterlimited environments.Some studies indicate that the majority of grain-legumes producers use direct seeding after a legume pre-crop [1].This change from conventional tillage (CT) to reduced or no tillage (NT) systems (with at least 30% of the soil surface covered) would lead to significant positive impacts on SOC [18].In contrast, other results indicate that such positive effects are limited to the first 20 cm depth, while little or no difference between CT and NT in total SOC can be seen lower down the soil profile [5,111].Such findings suggest that C stock changes in the soil are mainly dependent to the net N-balance in the system.With high N harvest index legumes, SOC stocks are not preserved due to the high amount of N taken off from the field into the grain [42].Conversely, the effect of legumes on soil carbon sequestration is more detectable for forage, green-manures and cover-crops which return to the soil large amounts of organic C and N [52].Boddey et al. [12] indicate that vetch under no tillage may increase SOC stocks under NT (0-100 cm) at a rate between 0.48 and 1.53 Mg C ha −1 per year [42].
The implementation of practices of conservation tillage could significantly reduce the GWP, especially when a grain legumes is added to the rotation.In Mediterranean agro-ecosystems, Guardia et al. [29] compared three tillage treatments (i.e.no tillage: NT, minimum tillage: MT, conventional tillage: CT) and two crops (i.e.vetch, barley) and recorded the emission of N 2 O, CH 4 and CO 2 during one year.Authors found a significant 'tillage × crop' interaction on cumulative N 2 O emissions with vetch releasing higher N 2 O amount than barley only in CT and MT, whereas similar fluxes were observed under NT.This was attributable to the soil water-filled pore space, dissolved organic carbon content and denitrification losses, in spite of the presumable predominance of nitrification.In any case, the most sustainable crop and tillage treatments in terms of GWP were represented by the non-fertilized vetch and NT, due to higher carbon sequestration, lower fuel consumption and the absence of mineral N fertilizers [29].In subtropical Ultisol, under legume cover crops, NT soil exhibited increased N 2 O emissions with respect to CT soil (531 vs. 217 kg CO 2 eq ha −1 year −1 ); however, emissions of this gas from NT soil were fully offset by CO 2 retention in soil organic matter (−2063 to −3940 kg CO 2 ha −1 year −1 ) [8].Moreover, NT soil under legume cover crops behaved as a net sink for GHG (GWP ranged from −971 to −2818 kg CO 2 eq ha −1 year −1 ) [8].
The expansion of ecological-based approaches like conservation agriculture opens opportunities to food legumes to be profitably included in sustainable cropping systems.There are still major challenges for conservation agriculture that need to be overcome, including the development of effective methods for weed control (see also [92]) that can avoid the use of herbicides or tillage.Overall conservation agriculture is an environmentally sustainable production system that may boost the incorporation of grain legumes within large and small-scale farming.

Conclusion
The roles and importance of grain legumes in a context of sustainability in agriculture could be enhanced by the emerging research opportunities for the major topics discussed above.
A major task in the future will be the selection of legume species and cultivars which could be effectively introduced across different cropping systems.An important point concerns balancing yield, which gives economic return, with the environmental and agronomic benefits.
Some priority areas seem emerge.Nitrogen fixation activity of grain legumes should be evaluated in relation with soil, climatic, plant characteristics and management conditions to find the suitable approach to achieve the best improvements.With this respect, the ability of the host plant to store fixed nitrogen appears to be a major component of increasing nitrogen fixation input.A particular focus should be paid also to the study of abiotic stress limitations and in particular water deficit, salinity and thermal shocks require extensive investigation.
Legumes that can recover unavailable forms of soil phosphorus could be major assets in future cropping systems.Consequently, those legumes which are able to accumulate phosphorus from forms normally unavailable need to be further studied, since phosphorus represents an expensive and limiting resource in several cropping systems.
Because of the growing request for plant products, i.e. protein and oils, and to the increased economic and environmental pressures on agro-eco systems, it emerges that grain legumes would play a major role in future cropping systems.

Table 1 Trends for word acreage (million ha) and yield (t ha −1 ) for legume crops included in FAOSTAT classification start- ing from 1974 to 2014 [23]; the major three cereal crops are also reported, for comparison InTable 2 ,
for each legume crop, item name and code as well as FAO definitions are reported a Data are referred to year 2013 (2014 data not available)

Table 2 Definition of legume crops focused in Table 1 and corresponding item name in FAOSTAT Legume crop Scientific name Corresponding FAO item name and code FAO definition
, for each legume crop, item name and code as well as FAO definitions are reported a Data are referred to year 2013 (2014 data not available) -7 times less GHG per unit area compared with other crops.Measuring N 2 O fluxes, they showed that peas emitted 69 kg N 2 O ha −1 , far less than winter wheat (368 kg N 2 O ha −1 ) and rape (534 kg N 2 O ha −1 ).
2 O-N ha −1 , respectively).The same authors highlighted that grain legumes significantly reduced their emission factors suggesting that legume-fixed N is a less-emissive form of N input to the soil than fertilizer N.