Root traits benefitting crop production in environments with limited water and nutrient availability

Abstract

Background

Breeding for advantageous root traits will play a fundamental role in improving the efficiency of water and nutrient acquisition, closing yield gaps, and underpinning the ‘Evergreen Revolution’ that must match crop production with human demand.

Scope

This preface provides an overview of a Special Issue of Annals of Botany on ‘Root traits benefitting crop production in environments with limited water and nutrient availability’. The first papers in the Special Issue examine how breeding for reduced shoot stature and greater harvest index during the Green Revolution affected root system architecture. It is observed that reduced plant height and root architecture are inherited independently and can be improved simultaneously to increase the acquisition and utilization of carbon, water and mineral nutrients. These insights are followed by papers examining beneficial root traits for resource acquisition in environments with limited water or nutrient availability, such as deep rooting, control of hydraulic conductivity, formation of aerenchyma, proliferation of lateral roots and root hairs, foraging of nutrient-rich patches, manipulation of rhizosphere pH and the exudation of low molecular weight organic solutes. The Special Issue concludes with papers exploring the interactions of plant roots and microorganisms, highlighting the need for plants to control the symbiotic relationships between mycorrhizal fungi and rhizobia to achieve maximal growth, and the roles of plants and microbes in the modification and development of soils.

INTRODUCTION

Global agricultural production is limited by adverse soil conditions and a lack of water and nutrients (Mueller et al., 2012; White et al., 2013a, b; Beza et al., 2017; Lynch, 2019). The yield gap (i.e. the difference between the actual and potential crop yield) differs between crops and varies greatly across the world (Mueller et al., 2012; Beza et al., 2017). The global yield gaps for the major cereals, rice, maize and wheat, are estimated to be about 30–60 %, of which most is contributed by lack of water and nutrients (Neumann et al., 2010; Mueller et al., 2012). Since plants acquire water and nutrients through their roots, roots have a fundamental role in improving the efficiency of water and nutrient acquisition, closing yield gaps, and underpinning the ‘Evergreen Revolution’ that must match crop production with human demand (Lynch, 2007, 2013, 2019; White et al., 2013b; Wissuwa et al., 2016; Lammerts van Bueren and Struik, 2017; Luo et al., 2019).

Many recent studies have investigated the roles of root traits in resource acquisition by plants to identify appropriate root ideotypes and genetic targets for breeding the crop cultivars of the future (Lynch, 2007, 2013, 2019; White et al., 2013a, b; Meister et al., 2014; Wissuwa et al., 2016; Lammerts van Bueren and Struik, 2017; Luo et al., 2019; W. Wang et al., 2019). Plant traits for acquiring mineral nutrients can be divided into ‘mining’ (points 1, 2 and 6 below) and ‘foraging’ (points 3, 4 and 5 below) strategies (Lynch, 2007). Root traits influencing the acquisition of mineral nutrients by plants include (1) high-affinity/high-capacity transport systems for the uptake of mineral nutrients, which reduce rhizosphere nutrient concentrations and can accelerate the diffusion and solubilization of mineral nutrients, (2) modification of rhizosphere pH and the efflux of low molecular weight organic solutes and/or enzymes from roots that alter the properties of rhizosphere soil and increase the amounts of nutrients released into solution, (3) distribution of roots in the soil profile relative to the phytoavailability of nutrients and the proliferation of roots in nutrient-rich patches, which improve the efficiency of nutrient acquisition, (4) relative biomass allocation to roots (root/shoot biomass quotient) and root growth rate, which impact traits such as seedling vigour, (5) architectural and anatomical characteristics of the root system, such as the abundance and length of lateral roots and root hairs, root length density (root length/soil volume), specific root length (length/biomass quotient) and metabolic load, which are often related to the formation of aerenchyma, all of which affect the volume of soil explored by the root system and the surface area for the uptake of mineral nutrients, and (6) interaction with microbes either directly, through symbiosis with mycorrhizal fungi or N-fixing bacteria, or indirectly, by impacting the microbial community of the rhizosphere (Lynch, 2007, 2013, 2019; Richardson et al., 2011; Bennett et al., 2013; White et al., 2013a, b; Wissuwa et al., 2016; Brown et al., 2017; Lammerts van Bueren and Struik, 2017; W. Wang et al., 2019).

The first papers in this Special Issue consider the consequences of breeding for reduced shoot stature and greater harvest index during the Green Revolution on root system architecture and the importance of various root traits for resource acquisition in environments with limited water or nutrient availability. These are complemented with papers exploring the interactions of plant roots and microorganisms, highlighting the control of symbiotic relationships between plants and mycorrhizal fungi and rhizobia, and the roles of plants and microbes in the modification and development of soils.

THE EFFECTS OF BREEDING FOR SHORT PLANT HEIGHT ON ROOT ARCHITECTURE AND RESOURCE ACQUISITION

The development of shorter, lodging-resistant cultivars of arable crops that responded to the application of more fertilizer by delivering greater yields was a major innovation of the Green Revolution (Fageria et al., 2011; White et al., 2013b). Modern cultivars of the major cereals, such as wheat (Triticum aestivum), rice (Oryza sativa) and maize (Zea mays), and other arable crops, such as oilseed rape (Brassica napus), have all been bred for short stature using a variety of ‘dwarfing’ genes (Peng et al., 1999; Hedden, 2003; Evenson and Gollin, 2003). Crop breeders have tacitly assumed that selecting genotypes for greater yields will concurrently select for genotypes with root systems that support these yields. However, the effects of dwarfing genes on root architecture and resource acquisition have rarely been studied.

In this Special Issue, Li et al. (2019) compared the root traits of 323 wheat accessions grown in the glasshouse with their shoot traits recorded in the field. They observed that accessions with deeper roots generally had a lower canopy temperature and greater grain mass per plant than accessions with shallow roots, but they found no significant relationships between root dry weight and plant height, canopy temperature or grain mass per plant. Since deeper rooting and shorter plant height were inherited independently, Li et al. (2019) concluded that these provided distinct target traits for molecular-assisted breeding. This is consistent with the conclusions of other researchers for a variety of crops. Li et al. (2019) also identified several chromosomal loci (quantitative trait loci, QTLs) impacting root depth, plant height and grain mass per plant, each of which explained between 4.7 and 18.7 % of the phenotypic variation. These QTLs included genes previously reported to affect plant height, such as the Green Revolution gene Rht-D1 (Peng et al., 1999), and three QTLs central to the regulation of root depth and grain mass per plant that include members of the nitrate transport (NRT2) gene family. Li et al. (2019) noted that favourable alleles for traits affecting plant height and grain mass per plant have been selected more strongly than those for root depth in recent wheat breeding programmes.

Alleles of several genes affecting plant shoot stature have been deployed in cereal breeding programmes. These include not only Rht-B1 and Rht-D1, which control responses to gibberellins, but also genes affecting responses to other plant hormones (Nadolska-Orczyk et al., 2017). The Auxin Response Factor (ARF) proteins are thought to regulate the expression of auxin-responsive genes (Roosjen et al., 2018). In this Special Issue, J. Wang et al. (2019) have studied the effects of TaARF4 on root length and plant height in wheat. They report that the TaARF4 gene is expressed constitutively throughout the plant and the TaTRF4 protein is located in the nucleus. Ectopic expression of TaARF4-A in arabidopsis (Arabidopsis thaliana), together with in vitro biochemical analyses, suggest that TaARF4 affects root length and plant height through both abscisic acid (ABA) and auxin signalling cascades. The geographical distribution of allele frequencies and breeding histories indicate that a TaARF4-B allele (Haplotype I), which confers shallow rooting and reduced plant height, was selected in Chinese wheat breeding programmes.

In Europe, commercial semi-dwarf hybrid cultivars of oilseed rape are produced using alleles of the BZH gene. The bzh hybrids are more winter-hardy and drought-tolerant than normal hybrids (Pinochet and Renard, 2012) and have greater N uptake efficiency (NUpE) and physiological N utilization efficiency (NUtE) than normal hybrids (Miersch et al., 2016). In this Special Issue, Schierholt et al. (2019) have examined whether the bzh gene affects the relative size of the root system, as dwarfing genes appear to do in wheat (Wojciechowski et al., 2009; Hodgkinson et al., 2017). They observed that seedlings of bzh hybrids had shorter main roots and fewer lateral roots than normal hybrids when grown on agar, and that bzh hybrids had smaller root electrical capacitance, which they used as a proxy for root system size (Chloupek, 1977; Dietrich et al., 2013), than normal hybrids at low N supply, but not at high N supply, in the field. They concluded that the greater root/shoot biomass quotient and larger harvest index of bzh hybrids than normal hybrids contributed to their greater NUpE and NUtE, respectively. This conclusion is supported by previous studies of N use efficiency in oilseed rape (He et al., 2017). Using a genetic mapping population, Schierholt et al. (2019) found a QTL impacting root electrical capacitance close to the BZH locus in plants grown in various environments.

A further focus of the breeding programmes of the Green Revolution was to accelerate and improve the uniformity of crop establishment, particularly in relatively infertile or hostile soils. Crop establishment, which is promoted by juvenile root vigour, is often, but not always, positively correlated with final yield (Thomas et al., 2016; Lammerts van Bueren and Struik, 2017; White et al., 2018; Khokhar et al., 2019). Anzooman et al. (2019) have examined two seedling traits, the rate of coleoptile elongation and the angle between the closest seminal roots from the vertical, that might be used to screen wheat genotypes for the potential to establish in sodic soils, which is restricted both by large exchangeable sodium concentrations (ESP) and high bulk density. They report that although coleoptile elongation was reduced by increasing either ESP or bulk density it did not differ among the genotypes they studied, whereas the angle between the central seminal roots differed significantly between genotypes but was not affected by soil ESP or bulk density. Anzooman et al. (2019) observed that seedling emergence in crusted sodic soils was inversely correlated with the angle between the central seminal roots. Thus, they argue that, among the genotypes they studied, although coleoptile length was not a suitable trait to identify genotypes with better establishment on sodic soils, selecting for a steeper root system might be. They suggest that genotypes with a narrower angle between their central seminal roots might access water and nutrients at depth more rapidly, which would be beneficial in sodic soils. They also note that narrower angles between the central seminal roots, together with a greater abundance of seminal roots, have been proposed as traits for identifying wheat genotypes with greater drought tolerance (Manschadi et al., 2008; Christopher et al., 2013) and productivity in hostile soils (Khokhar et al., 2019).

The rapid and consistent production of adventitious roots from cuttings is important for the clonal propagation of many horticultural and forestry crops (Druege et al., 2019). The induction of adventitious roots is controlled by the interaction of many plant hormones including auxin, ethylene, jasmonate, strigalactones and cytokinins (Druege et al., 2019). In this Special Issue, Yang et al. (2019) have investigated the role of auxin in the formation of adventitious roots in cuttings of petunia (Petunia × hybrida), which is stimulated by incubating cuttings in the dark before planting and inhibited by low N status of stock plants. They observed that incubating cuttings in the dark results in increased expression of various genes related to auxin synthesis, transport and signalling (e.g. genes encoding acyl acid amido synthetases, PhPIN1, Aux/IAA auxin repressors, Auxin Response Factors, auxin-responsive SAUR proteins), transport of auxin from the upper shoot and the accumulation of auxin in the stem base. By contrast, reducing the N content of cuttings reduced the expression of auxin-signalling genes and had little effect on auxin concentration in the stem base. They present a working model for adventitious rooting in petunia in which incubation of cuttings in the dark stimulates adventitious rooting through auxin accumulation and signalling in the stem base and the inhibition of adventitious rooting by N limitation acts downstream of the auxin signal.

ROOT IDEOTYPES FOR ENVIRONMENTS WITH LIMITED WATER AND NUTRIENT AVAILABILITY

A lack of water is often the greatest constraint to plant growth and crop production (Mueller et al., 2012; Beza et al., 2017; Battisti et al., 2018). In Mediterranean-type environments, deep soil water is generally a more reliable source of water than surface water and plant species adapted to these conditions tend to have a deep root system to overcome drought (Pierret et al., 2016). It has been suggested that this trait might similarly benefit crops cultivated in regions with restricted water availability (Manschadi et al., 2008; Lynch, 2013; White et al., 2013b).

The importance of a deep root system for increasing yields in dry environments has been reported previously for common bean (Phaseolus vulgaris; Polania et al., 2017). In this Special Issue, Berny Mier y Teran et al. (2019) screened 112 wild accessions from across Mesoamerica, together with 11 domesticated cultivars, for root and shoot traits related to drought tolerance. They observed that drought promoted a deeper root system and that beans originating in drier regions had deeper roots and produced more biomass than those originating in wetter regions. They also observed that, although domesticated cultivars had larger shoot and leaf biomass, larger root systems and deeper roots than wild accessions, wild accessions had a greater ability to continue growing in dry conditions than domesticated cultivars. Finally, they identified chromosomal molecular markers associated with root depth and biomass production in dry conditions that have been selected during domestication and suggest genetic targets for improving drought tolerance of common bean.

There is evidence that reducing transpiration, whether during the day or the night, can increase the yields of crops, such as soybean (Glycine max), wheat and sorghum (Sorghum bicolor), in Mediterranean-type, terminal-drought environments by preserving soil moisture for the entire crop growth period (Schoppach et al., 2017). Previous research on wheat has indicated that reducing root hydraulic conductance can contribute to reducing transpiration rates (Schoppach et al., 2014). In this Special Issue, Sadok and Schoppach (2019) report that auxin concentrations in roots, but not auxin or ABA concentrations in leaves, are strongly and negatively correlated with both daytime and nocturnal transpiration, plant hydraulic conductance and responses of transpiration to vapour pressure deficit among eight wheat genotypes with contrasting water-saving strategies and root hydraulic properties. They argue that auxin might reduce root hydraulic conductance and, thereby, transpiration by narrowing the diameter of central metaxylem vessels or by reducing the expression of genes encoding aquaporins. They observe that root auxin concentration is genotype-dependent and propose that this trait might be selected in breeding for greater wheat yields in dry environments.

In general, plants acclimate to a reduced supply of N or P by partitioning more biomass to roots and increasing their root/shoot surface area quotient (Hermans et al., 2006; White et al., 2013b). Since water uptake and losses from a plant must balance, this necessitates either a reduction in root hydraulic conductivity or greater water losses through stomata. Water flows across barley (Hordeum vulgare) roots are dominated by a transcellular pathway mediated by aquaporins, rather than a solely apoplastic pathway (Ranathunge et al., 2017). In this Special Issue, Armand et al. (2019) report a reduction in both transpiration and root hydraulic conductivity of barley plants grown hydroponically when they receive an inadequate N or P supply. They show that the activity of plasma-membrane aquaporins, but not the expression of genes encoding these aquaporins, is reduced in roots when plants receive suboptimal N or P supply, and that increased formation of Casparian bands and suberin lamella could also contribute to a reduced root hydraulic conductance under these conditions. A similar increase in root/shoot surface area quotient and response in root hydraulic conductance was reported following a reduction in K supply to barley plants (Coffey et al., 2018). Armand et al. (2019) argue that the modulation of root hydraulic conductivity allows plants receiving suboptimal nutrition to maintain balanced water flows whilst increasing their root/shoot surface area quotient.

In addition to the general increase in root/shoot biomass and root/shoot surface area quotients in response to limited P supply, plants often produce roots with more cortical aerenchyma and greater capacity for P uptake, adopt a ‘topsoil foraging’ phenotype that is complemented by the proliferation of lateral roots in P-rich patches and the production of long root hairs, secrete organic acids and phosphatases into the rhizosphere, and develop associations with mycorrhizal fungi (Richardson et al., 2011; White et al., 2013a, b; Lynch, 2019). In this Special Issue, Klamer et al. (2019) have quantified the importance of root hairs for P acquisition and growth of maize by comparing the rth2 mutant, which has very short root hairs, with its corresponding wild type in an alkaline soil with a limited or adequate P supply and a limited or sufficient water supply. They observe that the effects of reducing P and water supply on root and shoot growth and plant P content were additive and that the rth2 mutant accumulated less biomass and P than wild-type plants when stressed. Similar observations have been made on other plant species, including barley (Brown et al., 2012), confirming the importance of root hairs for P acquisition and growth of crops. To maintain an appropriate tissue P concentration, the rth2 mutant developed more lateral roots and had greater specific root length than wild-type plants, but did not release more organic carbon or phosphatase activity into the rhizosphere. Associations with mycorrhizal fungi might also compensate for a lack of root hairs, but mycorrhiza were absent from the roots of the plants studied by Klamer et al. (2019).

An extreme example of root foraging and mining of P-rich patches is found in plant species, such as lupins (Lupinus spp.) and members of the Proteaceae family, that produce short, determinate, ‘cluster’ roots that develop to exude copious amounts of organic acids in response to a local P supply (Lamont, 2003; Lambers et al., 2012). Brooker et al. (2015) suggested that species with complementary strategies for resource acquisition might act synergistically to exploit soil resources more efficiently than either alone. Thus, plants that are less adapted to environments with low P availability might benefit from co-cultivation with species with cluster roots that can acquire P effectively from infertile soils. In this Special Issue, Fajardo and Piper (2019) have examined the interactions between two species of southern South American Proteaceae (Embothrium coccineum and Gevuina avellana) with the ability to form cluster roots and two species of Nothofagaceae (Nothofagus betuloides and Nothofagus pumilio) that do not. Plants of the four species were grown in pots, either alone or in pairs, in a nutrient-rich ‘nursery’ or a nutrient-poor ‘tephra’ substrate. After three growing seasons, Fajardo and Piper (2019) observed that (1) Proteaceae had more cluster roots, greater cluster root biomass and better growth when cultivated in the nursery substrate than in tephra, (2) co-cultivation of a Proteaceae plant with any other plant (including a conspecific) led to greater leaf P concentrations and enhanced growth, and (3) co-cultivation of Nothofagaceae with another plant did not improve their growth or survival. Although previous studies have indicated that Proteaceae from both south-western Australia and southern South America can improve the growth and survival of other plant species, leading to a high species diversity in these ecosystems (Lambers et al., 2010, 2012, 2018), the data presented by Fajardo and Piper (2019) do not support the general hypothesis that Proteaceae facilitate the growth and survival of other plant species. In contrast, Fajardo and Piper (2019) argue that cluster roots provide southern South American Proteaceae with a competitive advantage over species that do not possess cluster roots on both infertile and fertile soils.

Aerenchymatous tissue is induced in response to both nutrient deficiencies and hypoxia in the rhizosphere (Lynch, 2013, 2019; Voesenek and Bailey-Serres, 2015; Yamauchi et al., 2018). The formation of aerenchyma increases specific root length and thereby allows greater exploration of the soil volume for a given root biomass (Lynch, 2013, 2019). It also enables better oxygenation of root tissues and is essential to the survival of plants in waterlogged soils (Voesenek and Bailey-Serres, 2015; Yamauchi et al., 2018). The formation of aerenchyma is a coordinated process that often involves programmed cell death followed by cell wall degradation (Voesenek and Bailey-Serres, 2015; Yamauchi et al., 2018). In sugarcane (Saccharum spp.), cell wall degradation during the formation of aerenchyma is largely restricted to the hydrolysis of β-glucan and some pectins in the middle lamella, and the formation of a polymer composite around the air spaces (Leite et al., 2017). In this Special Issue, Grandis et al. (2019) have examined the role of cell wall hydrolases in the development of aerenchyma in roots of sugarcane by assaying gene expression, protein abundance and enzymatic activities along the apical 5 cm of the root. They provide evidence that the formation of aerenchyma probably begins with the de-acylation of pectins (by acetyl esterases) followed by their degradation (by endopolygalacturonases, β-galactosidases and α-arabinofuranosidases). Subsequently, there is degradation of callose (by β-glucanases) and, at the same time, there is a partial degradation of xylan (by α-arabinofuranosidases, β-xylosidases and α-xylosidases) and xyloglucan (by xyloglucan endotransglycosilase/hydrolases), modification of xyloglucan–cellulose interactions (by expansins), and a partial degradation or rearrangement of cellulose (by cellulases and β-glucosidases). These observations support the hypothesis that the degradation of cell walls during the formation of aerenchyma is tightly controlled by the sequential production of specific cell wall hydrolases.

Genotypes of various crops have been selected for cultivation on land subject to periodic waterlogging and hypoxia. The perennial C4 grass Urochloa humidicola (syn. Brachiaria humidicola) is grown as a forage crop across tropical America on acidic soils that are prone to waterlogging. Genotypes tolerant to waterlogging have been developed by selecting for root traits that enable greater tissue aeration, such as more aerenchyma and the presence of a suberized exodermis (Jiménez et al., 2015). In this Special Issue, Jiménez et al. (2019) have investigated the influence of nutrient supply and hypoxia on root traits, leaf nutrient concentrations and growth of U. humidicola in a series of pot experiments. As expected, growth was restricted when plants had insufficient nutrients, but, although growth was reduced by hypoxia when nutrient supply was adequate, growth was not reduced by hypoxia when plants received a suboptimal nutrient supply. Plants grown in hypoxic conditions had more roots and shorter roots with narrower steles and greater air volume. Urochloa humidicola appeared to be constitutively adapted to restrict oxygen loss from roots with a well-developed, suberized exodermis and lignified sclerenchyma. Consistent with the hypothesis that cations such as Ca²⁺ might traverse the root to the xylem via an apoplastic pathway (White, 2001; Barberon et al., 2017), shoot Ca, Mg and K concentrations (but not shoot N, P and S concentrations) were reduced by hypoxia.

Plant roots acquire N largely in the form of nitrate or ammonium (Hawkesford et al., 2012). Although nitrate is readily accumulated by plants, large concentrations of ammonium in the rhizosphere are toxic to most plant species (Britto and Kronzucker, 2002). In waterlogged soils, the relative abundance of ammonium increases relative to that of nitrate because of reduced nitrification by soil microbes (Britto and Kronzucker, 2002; Nguyen et al., 2018). Paddy rice acquires most of its N as ammonium, but exposure to large concentrations of ammonium inhibit root growth. Previously, Xuan et al. (2013) reported that the transcription factor INDETERMINATE DOMAIN 10 (IDD10) regulated the expression of many ammonium-responsive genes, including AMMONIUM TRANSPORTER 1;2 (AMT1;2), in rice. In this Special Issue, Xuan et al. (2019) now show that the expression of many CALCINEURIN B-LIKE (CBL) genes and CBL INTERACTING PROTEIN KINASE (CIPK) genes in rice roots are sensitive to the presence of ammonium and that IDD10 activates the transcription of CIPK9 and CIPK14 directly. They also reveal that the heightened inhibition of root growth by the presence of ammonium in the idd10 mutant is independent of AMT1;2 activity, that root growth of the cipk9 and cipk23 mutants also show hypersensitivity to ammonium, and that overexpression of CIPK9 prevents hypersensitivity of root growth to ammonium in the idd10 mutant, thereby demonstrating that inhibition of root growth by ammonium can be independent of ammonium uptake and ameliorated by the activities of IDD10, CIPK9 and CIPK23.

IMPROVING CROP NUTRITION THROUGH INTERACTIONS WITH SOIL MICROBES

In addition to the anatomical, morphological and biochemical traits of roots that improve nutrient acquisition directly, the interactions of roots with microorganisms can also improve nutrient acquisition. These interactions include both intimate symbiotic relationships with mycorrhizal fungi and N-fixing bacteria as well as interactions with beneficial microbes that perform specific biochemical processes or modify the physical and chemical properties of the rhizosphere (Smith and Read, 2008; Richardson et al., 2011; White et al., 2013a, b; Huang et al., 2014; van der Heijden et al., 2015; Jin et al., 2017; Brundrett and Tedersoo, 2018; Sachs et al., 2018).

Arbuscular mycorrhizal fungi (AMF) can contribute to plant nutrition by mining the soil for poorly phytoavailable nutrients, such as P and micronutrients, and providing these to host plants (Smith and Read, 2008; van der Heijden et al., 2015; Brundrett and Tedersoo, 2018). Plant species differ in the extent to which they are colonised by AMF (Bennett et al., 2013; van der Heijden et al., 2015; Brundrett and Tedersoo, 2018; Martín-Robles et al., 2018) and the association of AMF with roots is determined by a complex set of interactions that are governed by both plant and fungal genetics (van der Heijden et al., 2015; Lanfranco et al., 2018). Interestingly, the benefits of symbiosis with AMF appear to have been reduced during the domestication of major crops (Lehmann et al., 2012; Bennett et al., 2013; Martín-Robles et al., 2018). In this Special Issue, Plouznikoff et al. (2019) have identified several QTLs involved in the colonization of tomato roots by AMF, each explaining between 3.8 and 8.7 % of the phenotypic variation in the frequency of colonization, the intensity of colonization, or the abundance of hyphae, arbuscules or vesicles/spores in roots, using a population of recombinant inbred lines derived from a cross between the cultivated tomato (Solanum lycopersicum) and a wild relative (Solanum pimpinellifolium). They observed that the frequency, intensity and arbuscular abundance of AMF colonization were highly correlated and that seven of the eight QTLs they identified affected both intensity and arbuscule abundance. Solanum pimpinellifolium usually provided the allele increasing the expression of the trait, which might be used in breeding programmes to increase AMF colonization of cultivated tomato. Candidate genes underpinning these traits likely affected C, N and P metabolism, membrane transport processes and plant defence.

The arbuscular mycorrhizal (AM) symbiosis begins with reciprocal molecular signalling between AMF and the roots of the host plant (Bucher et al., 2014; Huang et al., 2014; Gobbato, 2015; Lanfranco et al., 2018). There have been many studies documenting the changes in gene expression that occur during the penetration of root cells by AMF and following establishment of a functional AM symbiosis, but few studies have examined the changes in gene expression that occur in roots before their physical contact with AMF (Bucher et al., 2014; Camps et al., 2015; Lanfranco et al., 2018). In this Special Issue, Tian et al. (2019) report that considerable transcriptional reprogramming in wheat roots occurs in response to the presence of the AM fungus Rhizophagus irregularis (BGC-HEB07D) without any physical contact. Molecular signals from the AM fungus altered the expression of more than 2000 genes in wheat roots, many of which could not be attributed biological functions. However, many of the differentially expressed genes whose functions could be ascribed had roles in strigalactone and flavonoid signalling between roots and AMF, intracellular signalling, nutrient and metabolite transport, the penetration of root cells by AMF and the development of arbuscules. The expression of many of these genes also changes their expression in roots of barrel medic (Medicago truncatula) in response to signalling molecules (Myc factors) from AMF (Camps et al., 2015) and differs between uncolonized and fully colonized wheat roots (Li et al., 2018).

A substantial amount of the N required by legumes can be supplied through an effective symbiosis with rhizobia (Sachs et al., 2018; Ferguson et al., 2019). However, the carbon costs of this symbiosis can amount to 15–35 % of the photosynthate produced by the host plant (Minchin and Witty, 2005; Kaschuk et al., 2010). Thus, the symbiosis must be tightly controlled for maximal benefit to the plant (Sachs et al., 2018; Ferguson et al., 2019). Although hypernodulating mutants of legumes have a guaranteed N supply, the large number of root nodules can increase the metabolic burden and reduce the carbon available for plant growth. In this Special Issue, Silva Lopes et al. (2019) have investigated the consequences of hypernodulation for plant water relations. They report that a hypernodulating mutant (nod4) of soybean had smaller roots than a non-nodulating mutant (nod139) and a reduced water uptake capacity, which resulted in lower leaf water potential and depression of photosynthesis at midday. Thus, they conclude that, in addition to a being direct drain of carbon resources, hypernodulation might also reduce photosynthesis indirectly by affecting plant water status. This implies that the extent of nodulation should be considered carefully in legume breeding programmes, especially when breeding for dry environments.

In addition to directing the intimate symbioses between plants and mycorrhizal fungi or N-fixing bacteria, root exudates also influence the general microbial community in the rhizosphere and its biological activities (Richardson et al., 2011; Bennett et al., 2013; White et al., 2013a, b; Huang et al., 2014; Jin et al., 2017). In this Special Issue, Xu et al. (2019) have studied the effects of root exudates on the decomposition of soil organic matter through a ‘rhizosphere priming effect (RPE)’, which is thought to be mediated by changes in microbial abundance and community structure. The effects of root exudates on microbial communities can be both direct (e.g. by providing carbon substrates for growth) and indirect (e.g. by altering pH or increasing the bioavailability of nutrients). Xu et al. (2019) compared two near-isogenic lines of wheat varying in citrate efflux. They observed that increasing P supply resulted in greater plant growth and greater microbial biomass, soil respiration, RPE and concentrations of water-extractable P, Fe and Al in the rhizosphere of both genotypes. However, the genotype with greater citrate efflux had greater shoot growth, greater RPE and greater concentrations of soluble P, Fe and Al in its rhizosphere, but less microbial biomass in its rhizosphere, than the genotype with less citrate efflux, irrespective of P supply. Xu et al. (2019) suggest that increased citrate efflux from roots led to greater P and Fe bioavailability, which improved plant growth and led to a greater RPE independent of microbial biomass.

Complex interactions between parent geological materials, plants and rhizosphere organisms have contributed to the formation of soils (Retallack, 2001; Kennedy et al., 2006; Hazen and Ferry, 2010) and their chemical, physical and biological properties (Huang et al., 2014; Jin et al., 2017). In this Special Issue, Reith et al. (2019) have elucidated the interactions between eucalypts and rhizosphere bacteria that produce B horizons composed of clays and carbonates in ecosystems of south-western Australia. They observed that the development of the B horizon beneath Eucalyptus incrassata proceeded in four stages: (1) red pods of phyllosilicate clay, characterized by enrichment in Fe and Al and increases in bulk density, accumulate around rootlets developing from the deeper lateral roots of mallee eucalypts that (2) develop into brown columnar pods that (3) coalesce to form (4) a continuous clay pavement in mature woodland. The red pods were associated with abundant fungi and bacteria. The red pods were enriched with Actinobacteria and the brown pods were enriched with Firmicutes, while Acidobacteria, Proteobacteria and Verrucomicrobia were enriched in sites distant from the pods. Microbes from these phyla are known to form biominerals, including phyllosilicates, carbonates and Fe oxides. Reith et al. (2019) also observed that the B horizon beneath Proteaceous woodlands (composed of Banksia attenuata, Banksia menziesii and Grevillea spp.) in Western Australia form in a similar way, as do B horizons beneath birch (Betula spp.) and spruce (Picea spp.) woodlands in Europe and mixed coniferous forests in the USA. These observations support a general hypothesis that B horizons arise through interactions between specific taxa of woody plants, which provide the mineral elements required for biomineralization and support a characteristic microbial community, and rhizosphere microorganisms, which perform the biomineralization itself. Reith et al. (2019) speculate that the formation of a subterranean clay pavement through which only tree roots have penetrated might allow them exclusive access to deep water sources and promote their survival in arid environments.

CONCLUDING REMARKS

Roots acquire the water and mineral nutrients required for crop production. Breeding for beneficial root traits will, therefore, underpin the ‘Evergreen Revolution’ that must continue to match crop production with human demand in the future. There has been considerable effort to identify appropriate root ideotypes for sustainable agricultural systems and their genetic basis. The first papers in this Special Issue highlight the changes in root architecture that arose as a consequence of breeding for reduced plant height and greater harvest index during the Green Revolution. Fortunately, reduced plant height and root architecture are inherited independently and can be improved simultaneously to increase the acquisition and utilization of carbon, water and mineral nutrients. The next papers emphasize the importance of deep roots for the acquisition of water in dry environments and the need for (top)soil foraging for nutrients, such as P, that are relatively immobile in the soil. Since water and different mineral nutrients have contrasting distributions in the soil volume, it is important for plants to sense both their nutritional status and the spatial availability of water and nutrients in the soil to enable them to respond appropriately, whether anatomically, morphologically, physiologically or biochemically. Several papers in this Special Issue have examined the responses of plants to combined abiotic stresses and data from these studies might provide an insight into the hierarchy of responses to water and nutrient limitations to inform crop breeding. Plant traits improving resource acquisition directly can be complemented by both intimate and indirect interactions between plant roots and soil microorganisms. The concluding papers in this Special Issue highlight the importance for plants in controlling the symbiotic relationships between mycorrhizal fungi and rhizobia to achieve maximal growth, and the roles of plants and microbes in the modification and development of soils.

FUNDING

Work at the James Hutton Institute is supported by the Rural and Environment Science and Analytical Services Division (RESAS) of the Scottish Government.