Case study 2: Improving water and nutrient acquisition in dryland cereals in West Africa
Sorghum (Sorghum bicolor [L.] Moench) and pearl millet (Pennisetum glaucum [L.] R. Br.) are the fifth and sixth most important cereal crop in the World (FAOSTAT 2007; FAO 2014)⁠. They are well adapted to arid and semi-arid environments found in sub-Saharan Africa and India. In these regions, they are usually grown one cycle a year during the rainy season by smallholder farmers and represent a major source of micronutrient protein for humans and fodder for livestock. They are usually sown at low density (around 10,000 plants per hectare,FAO 2012) before or right after the first rain of the season and grown with no or low inputs simply because water and fertilizers are inaccessible and/or unaffordable (Matlon 1990; vom Brocke et al. 2010). Although sorghum and pearl millet prevail in these environments where other mainstream crops tend to fail, their cultivation is increasingly threatened by high temperature and intermittent drought caused by climate change and soil degradation caused by loss of nutrients and soil organic carbon, a phenomenon accentuated by desertification and erosion (Jones et al. 2013; Sultan, Defrance & Iizumi 2019). Therefore, improving the resilience of sorghum and pearl millet is particularly important for food security in arid and semi-arid regions of Africa and for adaptation of African agriculture to future climates.
Sorghum and pearl millet root systems are characterized by a single primary root and multiple post-embryonic nodal roots that originate from the mesocotyl (Tsuji et al. 2005; Chopart, Sine, Dao & Muller 2008; Singh et al. 2010; Faye et al. 2019). Primary, seminal and nodal roots all have lateral roots. In pearl millet, three different types of lateral roots have been identified that differ by length, diameter and internal structures (Passot et al. 2016), as well as by their growth dynamics (Passot et al. 2018). In sorghum, large variability in crown root angle and root area was observed in diversity panels, backcross nested association mapping (BCNAM) populations and recombinant inbred lines (Mace et al. 2012; Joshi et al. 2017). Similarly, diversity in primary root growth and root branching (Passot et al. 2016), and root length density (Faye et al. 2019) was observed in pearl millet. Despite this observed root phenotypic diversity and the fact that several authors have proposed root phenes as breeding targets for improvement of sorghum and pearl millet (Gemenetet al. 2016; Joshi et al. 2016), selection strategies involving root phenes have not been deployed to improve their cultivation so far.
For sorghum and pearl millet grown in sub-Saharan Africa where soils are generally deep and sandy with low water retention, deep rooting might be particularly interesting for tolerance to drought stress (Joshi et al. 2016). In maize, reducing crown root number or lateral root density was associated with deeper root growth, greater water capture at depth and improved plant water status and yield under drought (Zhan, Schneider & Lynch 2015; Gao & Lynch 2016). This response is linked to a carbon allocation mechanism in which plants that develop fewer roots are able to invest more carbon in individual roots that can grow deeper (Lynch 2013, 2018; Van Oosterom et al. 2016). Another means for a plant to grow deeper roots is to develop nodal roots with steeper growth angle (Lynch 2013). In sorghum, QTLs for steep nodal root angle co-located with QTL previously identified for stay-green and were associated with grain yield (Mace et al. 2012; Borrell et al. 2014).
Reduced root cortical cell file number (CCFN) and cortical cell size (CCS) were also hypothesized to reduce the metabolic cost of soil exploration (Lynch 2018). Maize lines with lower CCFN and greater CCS showed reduced root respiration, increased root growth at depth, better shoot growth and water status, and significant increase in yield under drought as compared to lines with more CCFN and less CCS (Burton, Brown & Lynch 2013; Jaramillo, Nord, Chimungu, Brown & Lynch 2013; Chimungu, Brown & Lynch 2014a,b). Similarly, maize lines with more root cortical aerenchyma (RCA) showed decreased root respiration, increased deep rooting, shoot growth and yield (Zhu, Brown & Lynch 2010; Chimungu et al. 2015). Moreover, simulations using the functional-structural plant model OpenSimRoot showed that more RCA, less CCFN and lower CCS had beneficial effects on plant biomass after 42 days of growth under nitrogen (N), P and potassium (K) stresses (Postma et al. 2017; Lynch 2019). RCA formation was observed both in sorghum and pearl millet indicating that this phene could be explored for diversity, genetic control and used for improving drought tolerance (Promkhambut, Polthanee, Akkasaeng & Younger 2011; Jaffuel et al. 2016; Passotet al. 2016). No information exists however on CCFN and CCS diversity nor their impacts in sorghum and pearl millet drought response.
If drought tolerance is often dependent on the ability for the plant to capture water, it is equally dependent on the way this water is used to produce biomass, i.e. its transpiration efficiency. Strategies to improve transpiration efficiency have sometimes relied on reduced xylem conductance capacity. In wheat for instance, reduced xylem diameter and the associated reduced root hydraulic conductance resulted in more conservative plants that yielded 11% more grains under drought conditions (Richards & Passioura 1989). In fact, annual crop plants adapted to drought stress environments tend to favour smaller xylem diameter as a water conservation strategy (Henry, Cal, Batoto, Torres & Serraj 2012; Kadam, Yin, Bindraban, Struik & Jagadish 2015; Grondin, Mauleon, Vadez & Henry 2016). Sorghum lines with higher number of xylem vessels showed higher transpiration rate which suggest that reducing xylem vessel number in this crop could lead to water saving strategies (Salih et al. 1999). The ability for a plant to restrict transpiration when the vapor pressure deficit (VPD) is above a certain threshold (transpiration restriction) represents other means for improving transpiration efficiency (Sinclairet al. 2017). In sorghum, simulations suggested that restricting maximum transpiration would increase transpiration efficiency and sustain physiological activities and yield (Sinclair 2005). Large variability for this transpiration restriction phenotype was observed in sorghum and in pearl millet (Kholová et al. 2010; Reddy et al. 2017; Tharanya et al. 2018; Karthika et al. 2019). In pearl millet, transpiration restriction was recently linked with root and shoot aquaporins expression although their precise function in this mechanism remains largely unknown (Reddy et al. 2017; for review see Shekoofa & Sinclair 2018; Tharanya et al. 2018). Improving transpiration efficiency (biomass/water transpired) over the entire crop cycle, possibly by restricting maximum transpiration at high VPD through reduction in root xylem size and possibly modulation in aquaporin functions could well conserve soil water for the critical reproduction and grain filling stage (Kholová et al. 2010; Vadez, Kholová, Yadav & Hash 2013; Vadez 2014).
Root hairs are well-known to improve P acquisition from the soil by increasing the absorption area of the root system (Lynch 2019). In sorghum and pearl millet, characterization of root hair density and length, genotypic variation, as well as their role in P uptake, remains limited. Recent study of the genetic architecture of phosphorus efficiency in sorghum showed colocalization between QTL for P acquisition efficiency, grain yield, surface area and root diameter (Bernadino et al., 2019). Interestingly, one QTL located on chromosome 3 was in close physical proximity with the sorghum homolog of rice serine/threonine kinase OsPSTOL1 , which was previously found to enhance early root growth and grain yield in rice under low-P (Gamuyao et al. 2012; Bernardino et al. 2019). Increasing exploration of shallow soil by increasing root length density or by increasing root hair length and density might be beneficial for P capture. In addition, benefits from more root hairs of sorghum could come from their contribution to the synthesis and exudation of sorgoleone that have demonstrated roles in growth inhibition of weeds (Netzly & Butler 1986; Pan et al. 2018).
There are clear trade-offs between root phenes beneficial for drought and low-P tolerance. In environments where P scarcity is always a constraint but drought is often intermittent, increased top-soil root hair length and density combined with drought-inducible plasticity in deeper root growth possibly through reduced top-soil root carbon cost (more RCA for instance) could co-optimize drought and low-P tolerance (Fig. 2B). Drought-related root plastic response has been described inSetaria italica (a close relative of sorghum and pearl millet), where an interruption in crown root growth under drought was observed (Sebastianet al. 2016). This plastic response appeared to be conserved in sorghum and pearl millet where nodal root length was significantly reduced when grown in split-pot system where the seminal root grew in moist soil while the crown roots grew in dry soil (Rostamza, Richards & Watt 2013). These observations suggest that root plasticity in response to drought exists, and could be exploited to improve drought-tolerance. Intercropping systems where root systems of neighbouring plants (sorghum/pearl millet and cowpea/groundnut for instance) could have complementary interactions in terms of water and nutrient availability without competing with each other may also help reducing these trade-offs (Brooker et al. 2015).
Beyond root architectural and anatomical phenes, targeting the rhizosphere could be another future avenue to improve dryland cereals performance (de la Fuente Cantó et al. 2020). The rhizosphere is the volume of soil around the root under the influence of the root system, i.e. whose physico-chemical and biological properties are modified by the root, which, in turn, impacts plant nutrition, development and physiology (York, Carminati, Mooney, Ritz & Bennett 2016; de la Fuente Cantó et al. 2020). One potential rhizosphere target phene in sorghum and pearl millet would be rhizosheath formation, i.e. the aggregation of soil particles around the roots (Ndour, Heulin, Achouak, Laplaze & Cournac 2020). This fraction of the soil firmly attached to the roots corresponds to the most biologically active fraction of the rhizosphere. First described in desert grasses, rhizosheath formation has since been reported in many cereal crops including sorghum and pearl millet (Duell & Peacock 1985; Brown, George, Neugebauer & White 2017; Ndour et al.2017b). A positive impact on water and mineral nutrition was reported for several plants in laboratory conditions and could be in part explained by improved contact between the soil and the root surface (Ndour et al. 2020). Phenotyping for rhizosheath size is high throughput and this phene is largely under plant genetic control and large variability exists in the germplasm in pearl millet thus making it a potential target for breeding (Ndour et al. 2021). However, further work is needed to demonstrate the impact of a larger rhizosheath on dryland cereals in field conditions.