Background and Aims Root phenotypes contribute to environmental adaptation. We hypothesized that root phenotypes of maize (Zea mays L. ssp. mays) landraces reflect their adaptation to edaphic limitations in their native soil environments, and that some root phenotypes may confer broad edaphic adaptation.Methods We phenotyped the roots of maize landraces and used the functional-structural plant/soil model OpenSimRoot_v2 to simulate landraces and their native environments to analyse how root phene states interact with each other and with environment variables to regulate edaphic adaptation.Key Results Landraces from low phosphorus regions have root phenotypes with shallow growth angles and greater nodal root numbers, allowing them to adapt to their native environments by improved topsoil foraging. We used machine learning algorithms to detect the most important phenotypes responsible for adaptation to multiple environments. The most important phene states responsible for stability across environments are large cortical cell size and reduced diameter of roots in nodes 5 and 6. When we dissected the components of root diameter, we observed that large cortical cell size improved growth by 28, 23 and 114 %, while reduced cortical cell file number alone improved shoot growth by 137, 66 and 216 %, under drought, nitrogen and phosphorus stress, respectively. Functional-structural analysis of 96 maize landraces from the Americas, previously phenotyped in mesocosms in the glasshouse, suggested that parsimonious anatomical phenotypes, which reduce the metabolic cost of soil exploration, were the main phenotypes associated with adaptation to multiple environments, while root architectural phenotypes were related to adaptation to specific environments.Conclusions These results indicate that integrated root phenotypes with anatomical phene states that reduce the metabolic cost of soil exploration increase tolerance to edaphic stress across multiple environments and therefore would improve yield stability, regardless of their root architecture.

Roots are essential for plant adaptation to changing environments, yet the role of roots in crop domestication remains unclear. This study examined the evolution of root phenotypes from teosinte to maize, a transition resulting in reduced nodal root number (NRN), multiseriate cortical sclerenchyma (MCS), and increased seminal root number (SRN). We reconstructed the root phenotypes of maize and teosinte, as well as the environments of the Tehuacán Valley over the last 18 000 yr using a combination of ancient DNA, paleobotany, and functional-structural modeling. Our models reveal that increasing Holocene atmospheric CO2 concentrations favored the appearance of reduced NRN and MCS between 12 000 and 8000 yr before present (yrBP), promoting deeper root systems. The advent of irrigation by 6000 yrBP switched nitrogen distribution from topsoil to subsoil domains, a change that increased the utility of reduced NRN and MCS. Comparison of allelic frequencies among ancient samples ranging from 5500 to 500 yrBP suggests that increased SRN may have appeared c. 3500 yrBP, coinciding with a period of increased human population, agricultural intensification, and soil degradation. Our results suggest that root phenotypes that enhance plant performance under nitrogen stress are important for maize adaptation to changing agricultural practices in the Tehuacán Valley.

Background and Aims Root axes with greater penetration ability are often considered to be beneficial in hard soils. We hypothesized that maize root phenotypes with greater plasticity (meaning reduced elongation in response to mechanical impedance, i.e. a 'stop signal') have fitness advantages over phenotypes with reduced plasticity (i.e. unimpeded root elongation) in native (virgin, uncultivated) soils, by reallocating root foraging to softer, presumably wetter, soil domains, and that the value of the stop signal reduced with soil cultivation and crop domestication.Methods We used OpenSimRoot to simulate native and cultivated soils and evaluated maize root phenotypes with varying axial and lateral root penetration ability in water, nitrogen (N) and impedance regimes associated with Neolithic agriculture.Key Results The stop signal was advantageous in native soils but was less beneficial in cultivated, irrigated soils. Reduced root foraging in hard, dry topsoil enabled root growth in deeper domains where water is available, resulting in an improved balance of resource expenditure and acquisition. The value of the stop signal declined during crop domestication with the advent of irrigation, which increased water availability in the topsoil. Soil cultivation reduced N availability, while irrigation increased N leaching, resulting in a shift in the fitness landscape, with greater lateral root length (i.e. reduced plasticity) being advantageous by colocalizing root foraging with N availability. The importance of the stop signal is evident in modern high-input systems in which drought is a limiting factor.Conclusions Our results support the hypotheses that the reduction of lateral root growth by mechanical impedance is adaptive in native soil, but became less adaptive with soil cultivation and irrigation associated with Neolithic agriculture.

Polyploidization played a crucial role in crop domestication and modern agriculture. While increased cell size in polyploids is known to enhance plant biomass and vigor, its impact on soil exploration remains poorly understood. Using wheat as a model, we identify a ploidy-induced belowground domestication syndrome, characterized by (a) increased root cortical cell size reducing root respiration, nitrogen content, and phosphorus content; (b) enlarged metaxylem vessels, increasing axial hydraulic conductance; and (c) blunter root tips, limiting penetration ability in compacted soils. Our empirical and in silico experiments show that reduced root respiration and reduced cellular nutrient content in wheat polyploids improved nutrient use and acquisition efficiency under suboptimal nitrogen and phosphorus availability. These adaptations would have been advantageous in nutrient-depleted agroecosystems of the Pre-Pottery Neolithic B (PPNB) Fertile Crescent, where continuous cultivation depleted soil fertility over time. Functional-structural modeling indicates that larger cortical cells in wheat polyploids increase vacuolar occupancy, reducing root metabolic costs. Enhanced axial hydraulic conducta nce may have improved water transport, an advantage in irrigated PPNB agroecosystems. However, polyploids have blunter root tips, which reduces their penetration ability in compacted soils, making them less suited for native soils with greater mechanical impedance. We propose that root anatomical changes driven by ploidy played an important role in adaptations of wheat domesticates to PPNB agriculture. One-Sentence Summary Polyploidy induced changes in root anatomy may have improved adaptation of wheat to Neolithic agroecosystems. ### Competing Interest Statement The authors have declared no competing interest. US Department of Energy ARPA-E, DE-AR0000821 Howard G. Buffett Foundation, https://ror.org/05v76zz53 US Department of Agriculture National Institute of Food and Agriculture, Hatch Appropriations Projects PEN04732, PENW-2020-03632 and Accession #:7009406

Pearl millet is a key food security grain crop in the world's drylands due to its tolerance to abiotic stresses. However, its yield remains low and is negatively impacted by climate change. Root phenes are potential targets to improve crop productivity and resilience to environmental stress. However, the sheer number of combinations resulting from interactions of multiple phenes is a challenge for empirical research. In silico approaches are a plausible alternative to assess the utility of different phene combinations in varying states over diverse environmental contexts. Here, we developed an implementation of the functional-structural plant/soil model – OpenSimRoot, for pearl millet in typical sub-Sahelian soil and environmental conditions. Root architectural, anatomical, and physiological parameters were measured using a popular pearl millet variety (Souna 3) and implemented in the model. The above-ground biomass and root length density predicted by the model were similar to data from field trials. The utility of different root phenes was then evaluated for improved phosphorus uptake and plant growth in P deficient soils. Doubled root hair length and density, shallower root angle (-15°) and doubled long lateral root density were found to improve plant growth by 76%, 33% and 33% respectively under low P conditions. Moreover, these phenes showed synergism when combined in silico and led to optimal biomass production in low P supply conditions that resulted in a 75% loss of biomass in the reference variety. Our study suggests that these phenotypes could be targeted to improve biomass production in pearl millet and consequently its yield in low-P availability conditions.

Improving nitrogen use efficiency in rice would provide economic and environmental benefits, but little is known about root morphological and anatomical responses to low nitrogen. In this study, two sets of rice genotypes, one set from the RDP1 panel, and one set of recombinant inbred lines, were used to characterize responses to gradual nitrogen depletion by plant uptake and movement of nitrogen to deeper soil strata as a result of leaching, so that more nitrogen was available at depth in aerobic soil mesocosms in a greenhouse. There was significant genetic variation in shoot biomass reductions in response to low nitrogen. The root to shoot biomass ratio was increased by low nitrogen in both sets of genotypes. Relative investment in nodal root number was accentuated with low nitrogen, and shoot biomass was correlated with numbers and lengths of nodal and large lateral roots. There was genetic variation for nodal root number and length in both sets of genotypes. Anatomical responses to low nitrogen were assessed in nodal roots of the RILs, where root cross-sectional area, stele area, and metaxylem vessel number were reduced by low nitrogen, and root diameter was reduced in the RDP1 genotypes. There were significant interactions of nitrogen with genotype for stele area and percent aerenchyma in the RILs. Genetic variation for low nitrogen responses may be useful for selection of rice lines with greater nitrogen acquisition under nitrogen-leaching conditions. ### Competing Interest Statement The authors have declared no competing interest.

AbstractRoot metaxylems are phenotypically diverse structures whose function is particularly important under drought stress. Significant research has dissected the genetic machinery underlying metaxylem phenotypes in dicots, but that of monocots are relatively underexplored. In maize (Zea mays), a robust pipeline integrated a genome‐wide association study (GWAS) of root metaxylem phenes under well‐watered and water‐stress conditions with a gene co‐expression network to prioritize the strongest gene candidates. We identified 244 candidate genes by GWAS, of which 103 reside in gene co‐expression modules most relevant to xylem development. Several candidate genes may be involved in biosynthetic processes related to the cell wall, hormone signaling, oxidative stress responses, and drought responses. Of those, six gene candidates were detected in multiple root metaxylem phenes in both well‐watered and water‐stress conditions. We posit that candidate genes that are more essential to network function based on gene co‐expression (i.e., hubs or bottlenecks) should be prioritized and classify 33 essential genes for further investigation. Our study demonstrates a new strategy for identifying promising gene candidates and presents several gene candidates that may enhance our understanding of vascular development and responses to drought in cereals.

Functional traits are fundamental for understanding and predicting crop responses to abiotic stress and yield improvement. Root functional traits are key determinants of carbon allocation and water transport efficiency. However, there are few studies integrating the effects of physiological and anatomical phenotypes in roots on yield. We performed a global scale data analysis to quantify the ecological effects of root functional traits on yield under abiotic stress. A field study was also conducted to test the relationships between yield and root anatomical traits such as cortex area and xylem area in two semi-arid regions in China. The search of literature on the relationships between yield and root phenotypes was fairly consistent with the results in the field. Meta-analysis showed specific root respiration was negatively related to yield under stress. Critical thresholds of whole root respiration (RR = 0.823) and xylem area (RR = 0.912) were observed for better yield under stress. Hydraulic conductance was positively correlated with xylem area, stele diameter and vessel number under stress. In the field study, there were inverse relationships between yield and cortex area and xylem area was positively related to wheat yield at both sites. Our study suggests that there are inverse relationships between yield and cortex area and respiration. Increased yield was associated with decreased whole root respiration with potentially smaller cortical tissue when whole respiration under stress was reduced by less than 17.7

### Competing Interest StatementThe authors have declared no competing interest.

SUMMARYIt has been hypothesized that vacuolar occupancy in mature root cortical parenchyma cells regulates root metabolic cost and thereby plant fitness under conditions of drought, suboptimal nutrient availability, and increased soil mechanical impedance. However, the mechanistic role of vacuoles in reducing root metabolic cost was unproven. Here we provide evidence to support this hypothesis. We first show that root cortical cell size is determined by both cortical cell diameter and cell length. Significant genotypic variation for both cortical cell diameter (~1.1‐ to 1.5‐fold) and cortical cell length (~ 1.3‐ to 7‐fold) was observed in maize and wheat. GWAS and QTL analyses indicate cortical cell diameter and length are heritable and under independent genetic control. We identify candidate genes for both phenes. Empirical results from isophenic lines contrasting for cortical cell diameter and length show that increased cell size, due to either diameter or length, is associated with reduced root respiration, nitrogen content, and phosphorus content. RootSlice, a functional‐structural model of root anatomy, predicts that an increased vacuolar: cytoplasmic ratio per unit cortical volume causes reduced root respiration and nutrient content. Ultrastructural imaging of cortical parenchyma cells with varying cortical diameter and cortical cell length confirms the in silico predictions and shows that an increase in cell size is correlated with increased vacuolar volume and reduced cytoplasmic volume. Vacuolar occupancy and its relationship with cell size merits further investigation as a phene for improving crop adaptation to edaphic stress.

Integrated root phenotypes contribute to environmental adaptation and yield stability. We used the functional-structural plant/soil model OpenSimRoot_v2 to reconstruct the root phenotypes and environments of eight maize landraces to understand the phenotypic and environmental factors associated with broad adaptation. We found that accessions from low phosphorus regions have root phenotypes with shallow growth angles and greater nodal root numbers, allowing them to adapt to their native environments by improved topsoil foraging. We used machine learning algorithms to detect the most important phenotypes responsible for adaptation to multiple environments. The most important phene states responsible for stability across environments are large cortical cell size and reduced diameter of roots in nodes 5 and 6. When we dissected the components of root diameter, we observed that large cortical cell size improved growth by 28%, 23 % and 114%, while reduced cortical cell file number alone improved shoot growth by 137%, 66% and 216%, under drought, nitrogen and phosphorus stress, respectively. Functional-structural analysis of 96 maize landraces from the Americas, previously phenotyped in mesocosms in the greenhouse, suggested that parsimonious anatomical phenotypes, which reduce the metabolic cost of soil exploration, are the main phenotypes associated with adaptation to multiple environments, while root architectural traits were related to adaptation to specific environments. Our results indicate that integrated phenotypes with root anatomical phenes that reduce the metabolic cost of soil exploration will increase tolerance to stress across multiple environments and therefore improve yield stability, regardless of their root architecture. ### Competing Interest Statement The authors have declared no competing interest. * OSRv2 : OpenSimRoot_v2 BF : lateral branching frequency CCS : cortical cell size Dia : diameter AN : root angle N# : node number

Purpose Quantifying the stability of individual plants or their contribution to soil reinforcement against erosion or landslides requires an understanding of the tensile properties of their roots. This work developed a new analytical model to understand the tensile stress–strain behaviour of a single root axis, which is the first to incorporating root anatomical features, in order to reduce the existing uncertainty in predictions. Methods The root was modelled as a linear elastic stele connected to a surrounding linear elastic cortex by means of a linear elastic stele–cortex interface. By solving for force equilibrium, an analytical solution for the full tensile stress–strain behaviour — including any intermediate brittle failures of the stele, cortex and/or interface — was obtained. This model was compared to tensile tests and laser ablation tomography for maize roots. Results The new modelling approach demonstrated that the root tensile strength is fully determined by the strength of the stele alone, which was an order of magnitude larger than that of the cortex while also 3–4 times stiffer. The reduction in root stiffness beyond the yield point was linked to continuing fracturing of the cortex and debonding along the stele–cortex interface. A larger proportion of the variation in experimentally measured biomechanical characteristics could be explained compared to root diameter power-law fitting methods typically applied in the literature. Conclusion Stele and cortex biomechanical properties are substantially different, affecting the tensile behaviour of plant roots. Accounting for these anatomical traits increased the accuracy root biomechanical properties from tensile tests.

We describe how increased root cortical parenchyma wall width (CPW) can improve tolerance to drought stress in maize by reducing the metabolic costs of soil exploration. Significant variation (1.0 to 5.0 µm) for CPW was observed within maize germplasm. The functional-structural model RootSlice predicts that increasing CPW from 2 to 4 µm is associated with ca. 15% reduction in root cortical cytoplasmic volume, respiration rate, and nitrogen content. Analysis of genotypes with contrasting CPW grown with and without water stress in the field confirms that increased CPW is correlated with ca. 32 to 42% decrease in root respiration. Under water stress in the field, increased CPW is correlated with 125% increased stomatal conductance, 325% increased leaf CO 2 assimilation rate, 73 to 78% increased shoot biomass, and 92 to 108% increased grain yield. CPW was correlated with leaf mesophyll midrib parenchyma wall width, indicating pleiotropy. GWAS analysis identified candidate genes underlying CPW. OpenSimRoot modeling predicts that a reduction in root respiration due to increased CPW would also benefit maize growth under suboptimal nitrogen, which requires empirical testing. We propose CPW as a new phene that has utility under edaphic stress meriting further investigation. Significance Statement Suboptimal water availability is a primary constraint for global crop production that is intensifying due to climate change. The metabolic cost of soil exploration is a critical factor in plant performance under suboptimal water availability. This study highlights how increased root cortical parenchyma wall width (CPW) reduces root metabolic cost and improves crop adaptation to water deficit. Modeling results also indicate that increased CPW would be beneficial under suboptimal nitrogen availability. Therefore, CPW is a promising target for breeding crops with improved water and nitrogen use efficiency.

Deeper rooted crops are an avenue to increase plant water and nitrogen uptake under limiting conditions and increase long-term soil carbon storage. Measuring rooting depth, however, is challenging due to the destructive, laborious, or imprecise methods that are currently available. Here, we present LEADER (Leaf Element Accumulation from DEep Roots) as a method to estimate in-field root depth of maize plants. We use both X-Ray fluorescence spectroscopy (XRF) and ICP-OES (Inductively Coupled Plasma Optical Emission spectroscopy) to measure leaf elemental content and relate this to metrics of root depth. Principal components of leaf elemental content correlate with measures of root length in four genotypes (R 2 = 0.8 for total root length), and we use linear discriminant analysis to classify plants as having different metrics related to root depth across four field sites in the United States. We can correctly classify the plots with the longest root length at depth with high accuracy (accuracy greater than 0.6) at two of our field sites (Hancock, WI and Rock Spring, PA). We also use strontium (Sr) as a tracer element in both greenhouse and field studies, showing that elemental accumulation of Sr in leaf tissue can be measured with XRF and can estimate root depth. We propose the adoption of LEADER as a tool for measuring root depth in different plant species and soils. LEADER is faster and easier than any other methods that currently exist and could allow for extensive study and understanding of deep rooting.

Excavation of plants from soil readily reveals that the shape of root systems is quite complex. The elegant drawings of Kutschera (e.g., Fig. 7.1; Kutschera and Lichtenegger 1992), collected painstakingly over many years of fieldwork, provide beautiful illustrations of the complexity and diversity of the spatial configuration of root systems, or ‘root architecture’. The possible importance of this complexity and diversity has been considered in the scientific literature for over a century (e.g., Anonymous 1887; Cannon 1911; Weaver 1926). Speculation has primarily focused on the role of root architecture in the acquisition of water (e.g., Pages et al. 1997) and nutrients (reviewed below), and in support of the shoot (Coutts 1983; Ennos and Fitter 1992; Bailey et al. 2002), although other important functions include storage, signaling processes (Aiken and Smucker 1996), disease and insect resistance, and belowground symbioses (e.g., Hetrick 1991). Although the focus of this chapter is the role of root architecture in nutrient acquisition, root architecture can best be understood as an aggregate of traits that co-optimize plant adaptation to several current and potential constraints. By analogy, the functional importance of shoot architecture has been fruitfully analyzed as co-optimizing solutions to the needs for structural support, light capture, and propagule dispersal (Niklas 1997). For roots, architectural patterns may co-optimize the acquisition of immobile soil resources such as phosphorus, the acquisition of mobile soil resources such as water, and support of the shoot. It must also be considered that roots exist in a hostile environment in which any given root may be lost through abiotic or biotic stress, so redundancy may be a virtue in root architecture. The term ‘root architecture’ is used to refer to the spatial configuration of a root system, excluding anatomical or morphological features of individual roots such as root hairs, but including assemblages of connected roots or entire root systems (Fitter 1991a; Lynch 1995). Root architecture is dynamic and environmentally plastic. It results from an aggregate of specific physio-

The root is a well-studied example of cell specialisation, yet little is known about the metabolism that supports the transport functions and growth of different root cell types. To address this, we used computational modelling to study metabolism in the elongation zone of a maize lateral root. A functional-structural model captured the cell-anatomical features of the root and modelled how they changed as the root elongated. From these data, we derived constraints for a flux balance analysis model that predicted metabolic fluxes of the 11 concentric rings of cells in the root. We discovered a distinct metabolic flux pattern in the cortical cell rings, endodermis and pericycle (but absent in the epidermis) that involved a high rate of glycolysis and production of the fermentation end-products lactate and ethanol. This aerobic fermentation was confirmed experimentally by metabolite analysis. The use of fermentation in the model was not obligatory but was the most efficient way to meet the specific demands for energy, reducing power and carbon skeletons of expanding cells. Cytosolic acidification was avoided in the fermentative mode due to the substantial consumption of protons by lipid synthesis. These results expand our understanding of fermentative metabolism beyond that of hypoxic niches and suggest that fermentation could play an important role in the metabolism of aerobic tissues.

Root architectural phenotypes are promising targets for crop breeding, but root architectural effects on microbial associations in agricultural fields are not well understood. Architecture determines the location of microbial associations within root systems, which, when integrated with soil vertical gradients, determines the functions and the metabolic capability of rhizosphere microbial communities. We argue that variation in root architecture in crops has important implications for root exudation, microbial recruitment and function, and the decomposition and fate of root tissues and exudates. Recent research has shown that the root microbiome changes along root axes and among root classes, that root tips have a unique microbiome, and that root exudates change within the root system depending on soil physicochemical conditions. Although fresh exudates are produced in larger amounts in root tips, the rhizosphere of mature root segments also plays a role in influencing soil vertical gradients. We argue that more research is needed to understand specific root phenotypes that structure microbial associations and discuss candidate root phenotypes that may determine the location of microbial hotspots within root systems with relevance to agricultural systems.

Data and code for figure generation for LEADER (Leaf Element Accumulation from Deep Roots): a nondestructive phenotyping platform to estimate rooting depth in the field

ABSTRACT The root anatomical phenotype root cortical aerenchyma (RCA) decreases the metabolic cost of soil exploration and improves plant growth under drought and low soil fertility. RCA may also change the microenvironment of rhizosphere microorganisms by increasing oxygen availability or by reducing carbon rhizodeposition. We tested the hypothesis that plants with contrasting expression of RCA have different rhizosphere prokaryotic communities. Maize inbreds were grown in two field sites, Limpopo Province, South Africa and Pennsylvania, USA, and their rhizosphere soil sampled at flowering. High- and low-nitrogen fertilization was imposed as separate treatments in the experiment in South Africa. The rhizosphere microbial composition of plants with contrasting RCA was characterized by metabarcoding of the 16S rRNA genes. Geographic location was the most important factor related to the composition of rhizosphere microbial communities. In the site in South Africa, RCA explained greater percent of variance (9%) in the composition of microbial communities than genotype (7%). Although other root anatomical and architectural phenotypes were studied as possible cofactors affecting the microbial composition, RCA was among the best significant explanatory variables for the South African site although it was neutral in the Pennsylvania site. High-RCA rhizospheres significantly enriched OTUs of the families Burkholderiaceae (in South Africa) and Bacillaceae (in USA), compared to low-RCA plants, and OTUs of the families Beijerinckiaceae and Sphingomonadaceae were enriched at the two nitrogen levels in high RCA plants in South Africa. Our results are consistent with the hypothesis that RCA is an important factor for rhizosphere microbial communities, especially under suboptimal nitrogen conditions.

Background Suboptimal nitrogen availability is a primary constraint for crop production in low-input agroecosystems, while nitrogen fertilization is a primary contributor to the energy, economic, and environmental costs of crop production in high-input agroecosystems. In this article we consider avenues to develop crops with improved nitrogen capture and reduced requirement for nitrogen fertilizer. Scope Intraspecific variation for an array of root phenotypes has been associated with improved nitrogen capture in cereal crops, including architectural phenotypes that colocalize root foraging with nitrogen availability in the soil; anatomical phenotypes that reduce the metabolic costs of soil exploration, improve penetration of hard soil, and exploit the rhizosphere; subcellular phenotypes that reduce the nitrogen requirement of plant tissue; molecular phenotypes exhibiting optimized nitrate uptake kinetics; and rhizosphere phenotypes that optimize associations with the rhizosphere microbiome. For each of these topics we provide examples of root phenotypes which merit attention as potential selection targets for crop improvement. Several cross-cutting issues are addressed including the importance of soil hydrology and impedance, phenotypic plasticity, integrated phenotypes, in silico modeling, and breeding strategies using high throughput phenotyping for co-optimization of multiple phenes. Conclusions Substantial phenotypic variation exists in crop germplasm for an array of root phenotypes that improve nitrogen capture. Although this topic merits greater research attention than it currently receives, we have adequate understanding and tools to develop crops with improved nitrogen capture. Root phenotypes are underutilized yet attractive breeding targets for the development of the nitrogen efficient crops urgently needed in global agriculture.