Browse Subject Areas

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

Evolutionary Divergences in Root Exudate Composition among Ecologically-Contrasting Helianthus Species

  • Alan W. Bowsher ,

    Affiliation Department of Plant Biology, University of Georgia, Athens, Georgia, United States of America

  • Rifhat Ali,

    Affiliation Department of Environmental Health Science, University of Georgia, Athens, Georgia, United States of America

  • Scott A. Harding,

    Affiliations Department of Genetics, University of Georgia, Athens, Georgia, United States of America, Warnell School of Forestry and Natural Resources, University of Georgia, Athens, Georgia, United States of America

  • Chung-Jui Tsai,

    Affiliations Department of Genetics, University of Georgia, Athens, Georgia, United States of America, Warnell School of Forestry and Natural Resources, University of Georgia, Athens, Georgia, United States of America

  • Lisa A. Donovan

    Affiliation Department of Plant Biology, University of Georgia, Athens, Georgia, United States of America

Evolutionary Divergences in Root Exudate Composition among Ecologically-Contrasting Helianthus Species

  • Alan W. Bowsher, 
  • Rifhat Ali, 
  • Scott A. Harding, 
  • Chung-Jui Tsai, 
  • Lisa A. Donovan


Plant roots exude numerous metabolites into the soil that influence nutrient availability. Although root exudate composition is hypothesized to be under selection in low fertility soils, few studies have tested this hypothesis in a phylogenetic framework. In this study, we examined root exudates of three pairs of Helianthus species chosen as phylogenetically-independent contrasts with respect to native soil nutrient availability. Under controlled environmental conditions, seedlings were grown to the three-leaf-pair stage, then transferred to either high or low nutrient treatments. After five days of nutrient treatments, we used gas chromatography-mass spectrometry for analysis of root exudates, and detected 37 metabolites across species. When compared in the high nutrient treatment, species native to low nutrient soils exhibited overall higher exudation than their sister species native to high nutrient soils in all three species pairs, providing support for repeated evolutionary shifts in response to native soil fertility. Species native to low nutrient soils and those native to high nutrient soils responded similarly to low nutrient treatments with increased exudation of organic acids (fumaric, citric, malic acids) and glucose, potentially as a mechanism to enhance nutrition acquisition. However, species native to low nutrient soils also responded to low nutrient treatments with a larger decrease in exudation of amino acids than species native to high nutrient soils in all three species pairs. This indicates that species native to low nutrient soils have evolved a unique sensitivity to changes in nutrient availability for some, but not all, root exudates. Overall, these repeated evolutionary divergences between species native to low nutrient soils and those native to high nutrient soils provide evidence for the adaptive value of root exudation, and its plasticity, in contrasting soil environments.


Mineral nutrient availability in soils is considered a key factor influencing plant productivity and species distributions [13]. Although nutrient limitation is common in terrestrial ecosystems, plants have evolved numerous mechanisms to acquire nutrients from soils via their complex root systems. For example, root system morphology, distribution, and architecture play vital roles in exploring the soil for nutrients [4]. In addition to these physical mechanisms for securing nutrient acquisition, plants can chemically influence nutrient availability in soils through root exudation [5,6]. Root exudation, the passive or active release of inorganic ions, volatiles, and primary and secondary metabolites from roots, is a ubiquitous phenomenon in higher plants [7]. Root exudates can increase local nutrient availability in the rhizosphere by influencing soil pH, competing for mineral adsorption sites, chelating mineral nutrients, and solubilizing soil minerals [5,811]. Exudates can also indirectly improve plant nutrient acquisition through interactions with microbes. For example, exudation of flavonoids is a critical component for nodule establishment in N-fixing symbioses, and plays an important role in mycorrhizal interactions [1214]. Root exudates also impact rhizosphere community composition, as well as soil carbon and nutrient cycling [15,16]. Consequently, root exudation is considered to play an important role in plant nutrient acquisition, particularly in low fertility environments [5,17].

Given the potential for root exudation to improve nutrient acquisition by plants, numerous studies have investigated variation among species for root exudate composition [1820]. Many species respond to nutrient deficiency with increased exudation of organic acids and sugars [2125], although this response varies both qualitatively and quantitatively among species [22,23,26]. Increased exudation of sugars and organic acids under nutrient deficiency is often expected to be adaptive, given the influence of these metabolites on soil nutrient availability [9,27,28] and microbial community structure [25,29,30]. Therefore, infertile soils may generally favor species with either inherently high exudation of organic acids and sugars, or a high capacity to increase exudation of these metabolites in response to nutrient deficiency [21,23,31]. However, it has also been suggested that exudation of organic compounds in general should be inherently higher in fast-growing species characteristic of fertile soils, due to their high nutrient demand and capacity to sustain rapid metabolic activity via high rates of nutrient assimilation [32,33]. For example, higher exudation in species native to fertile soils may contribute to the faster rates of nutrient cycling observed in the rhizosphere of such species in common garden studies [33,34]. Studies directly comparing species native to low nutrient soils and those native to high nutrient soils are therefore needed to clarify how root exudate composition (i.e. metabolite identity and relative quantity) has differentiated across environments, and to shed light on the role of root exudation in adaptation to soil fertility [35].oweHoHoasdfoweHoHoasdf

Of the studies that have examined variation among species for root exudate composition in relation to native soil characteristics [18,36,37], few have done so in a phylogenetically-informed framework. Controlling for species’ phylogenetic relatedness is an important consideration in comparative studies in order to remove the potential confounding influence of species’ shared evolutionary histories on inferences of adaptive differentiation [38,39]. Similar trait shifts in response to local environmental conditions in multiple phylogenetically-distinct lineages indicates convergent evolution, and provides evidence for the adaptive significance of that trait [40,41]. Thus, while comparative analyses of root exudate composition across species native to contrasting environments are useful for generating testable hypotheses, studies which incorporate species pairs from multiple lineages are particularly informative for making adaptive inferences.

In this controlled environment study, we used gas chromatography-mass spectrometry (GC-MS) to analyze the root exudates of six Helianthus species native to habitats differing in soil fertility. GC-MS is best-suited for detection of small, thermostable compounds of primary metabolism, such as sugars and organic acids. Although secondary metabolites such as sesquiterpene lactones have also been detected in root exudates of cultivated sunflower [42], we focused our study on small, primary metabolites which are expected to comprise the majority of root exudates in higher plants [43], and can impact plant nutrient acquisition by influencing both nutrient availability as well as plant-microbe interactions [9,25,27,2830]. Species were chosen as pairs of phylogenetically-independent contrasts, with each pair including one species native to a relatively low nutrient soil (LNS) and the other native to a relatively high nutrient soil (HNS). Specifically, we asked: (1) Does root exudate composition consistently (i.e. in all three species pairs) differ either quantitatively or qualitatively between species native to LNS and those native to HNS? In addition, responses to changes in nutrient availability (i.e. shifts in root exudate composition between high nutrient-treated and low nutrient-treated plants) may also be adaptive. Therefore, we also asked: (2) Do species native to LNS versus HNS consistently (in all three species pairs) differ in their response to low nutrient treatments? Consistent differences between species native to LNS and their sister species native to HNS in all three species pairs would indicate independent, repeated evolution of root exudate composition (question 1), or its response to low nutrient availability (question 2), providing evidence for its adaptive significance.

Materials and Methods

Study system

The Helianthus (sunflower) genus includes approximately 50 herbaceous species found in diverse habitats across North America [44,45]. As part of a separate study, seeds were collected for 28 diploid Helianthus species, either directly from wild population native sites, or from accessions established at the USDA National Genetic Resources Program [46]. Soil cores were also collected from the location of each seed source and analyzed for soil fertility characteristics at A & L Laboratories (North Chesterfield, VA, USA), and soil N concentrations were assessed at the University of Georgia Stable Isotope Laboratory (Micro-Dumas combusition; NA1500, Carlo Erba Strumentazione, Milan, Italy). The results of these soil fertility analyses have largely been reported previously [46]. Based on the native soil data for each species, we chose three pairs of sister species (six species total) from different clades in the Helianthus phylogeny [47], such that each pair contains one species native to a low nutrient soil (LNS) and the other native to a high nutrient soil (HNS) (Fig 1). Seeds from these six species were all collected directly from wild populations as part of the original seed and soil collections in either 2011 or 2012 at the following locations: H. annuus (Kansas; N39°06’N 96°36’W), H. argophyllus (Texas; 27°38’N 97°13’W), H. petiolaris (Illinois; 41°55’N 90°06’W), H. praecox ssp. runyonii (Texas; 27°39’N 97°18’W), H. grossesseratus (Illinois; 41°38’N 89°32’W), H. microcephalus (South Carolina; 34°15’N 82°39’W). Seeds for H. argophyllus, H. praecox, H. grossesseratus, and H. microcephalus were collected on public roadsides and right-of-ways, and required no permissions for collection. Seeds for H. petiolaris were collected under a permit from the Illinois Nature Preserve Commission, while H. annuus seeds were collected with permission from Konza Prairie Biological Station. None of these collections involved endangered or protected species.

Fig 1. Phylogeny (based on the most recent phylogeny for Helianthus [47]) and native soil fertility of the study species.

Data are the mean (SE) of five replicates per species. Soil fertility values for four of the species have been previously reported [46]. Different letters indicate significant differences (p < 0.05) between the two sister species within a given clade. Species native to a low nutrient soil (relative to its sister species) indicated by (§). (N) nitrogen; (P) phosphorus; (K) potassium; (OM) organic matter.

Plant growth conditions, nutrient treatments, and root exudate collection

Seeds were scarified and placed on moist filter paper in petri dishes in the dark for 24 hours before removing the seed coats with forceps. Germinating seedlings were transferred to a growth chamber (Conviron, Winnipeg, Canada) programmed for a 12-hour 25/20°C day/night cycle with 70% relative humidity. Seedlings were grown hydroponically as described [48]. Briefly, one week after scarification, a small section of rock wool was wrapped around the hypocotyl of each seedling, and seedlings were suspended in individual 125 mL Erlenmeyer flasks wrapped in foil. Flasks were filled with 50% strength modified Hoagland’s complete nutrient solution [49] and were aerated using aquarium pumps. Nutrient solutions and rock wool were replaced every other day. Due to seedling mortality in H. microcephalus and H. praecox, these two species were replaced with a second set of seedlings germinated 4 weeks after the first set. To assess the potential for confounding temporal effects on root exudation, replicate H. petiolaris seedlings were germinated in both sets (n = 3 seedlings per set). No significant temporal effects were observed for the abundance of any metabolite detected in H. petiolaris (t-tests, all p > 0.05); therefore, temporal effects were not considered further.

As root exudate composition is known to vary with plant ontogeny [50], nutrient treatments were initiated when individual plants had reached the same developmental stage to control for differences in growth rates among species. At the emergence of the third true leaf pair, seedling root systems were rinsed with deionized water, then suspended in individual 125 mL Erlenmeyer flasks containing either 50% strength (high) or 1.25% strength (low) modified Hoagland’s complete nutrient solution. For each of the six species, replicate seedlings (n = 5–7 per treatment) were randomly assigned to each nutrient treatment. Nutrient treatment solutions were replaced at two and four days after the initial nutrient treatments. Five days after initiating the nutrient treatments, root exudates were collected as described [48]. Briefly, root systems of intact seedlings were rinsed in deionized water and placed in individual 15 mL glass vials wrapped in foil. Vials were then filled with enough deionized water to cover the root system, and these volumes were recorded. After five hours, the aqueous medium containing root exudates was collected from each vial and adjusted to the original volume. Samples were immediately snap-frozen in liquid nitrogen and stored at -80°C.

Root systems were stained with 0.01% (w/v) Toludine Blue O (Carolina Biological Supply Co, Burlington, NC, USA) for three minutes, then rinsed in deionized water to remove excess stain. Root systems were placed in a clear plastic tray in a thin layer of water to minimize overlap and scanned on a desktop scanner at 400 dpi. Images were analyzed using the software WinRHIZO (v. 2002c, Regent Instruments, Quebec, Canada) for calculation of total root length and surface area. Root systems were then dried at 60°C for 72 hours before weighing. Across species and treatments, both root system length and surface area were highly correlated with root mass (r = 0.831, P < 0.0001, n = 75; and r = 0.896; P < 0.0001; n = 75, respectively). Therefore, we report root exudate abundance normalized per unit root mass (described below).

Metabolite analysis

One mL samples of the root exudate collection medium were evaporated to dryness in a vacuum concentrator (CentriVap, Labconco, Kansas City, MO, USA). Samples were then derivatized and analyzed using GC-MS, as described [48,51,52]. Concentrated samples were methoximated with 15 μL methoxyamine hydrochloride (40 mg/mL in pyridine) for 90 minutes at 30°C, then silylated with 30 μL N-methyl-N-(trimethylsilyl)-trifluoroacetamine for 90 minutes at 60°C. Samples (1 μL) were injected by autosampler onto a DB-5MS column (30 m x 0.25 mm ID, and 0.25 μm film, with deactivated guard column) in 1:5 split mode. An Agilent 7890A GC (Agilent Technologies, Wilmington, DE, USA) was used for metabolite separation using the following temperature settings: 80°C hold for 1 minute, ramping of 20°C min-1 to 200°C, then 10°C min-1 to 310°C, followed by a 6.5 minute hold. Detection was with an Agilent 5975C quadrupole MS, with source set at 230°C and mass filter set at 150°C. Spectra were collected in scanning ion mode (m/z 50–500) with Chemstation (Agilent Technologies), and, following spectral deconvolution using AnalyzerPro (SpectralWorks, Runcom, UK), peak identities were assigned using both public (NIST08 [53]; Fiehnlib [54]) and in-house MS libraries constructed using authentic standards. Although the use of public databases and in-house libraries allows determination of putative metabolite identities, strict confirmation requires co-elution with pure authentic standards. Future work should confirm the putative metabolite identities reported in this study; however, given that our study detected common, anticipated primary metabolites, we chose a stringent spectral match threshold to improve confidence in assignment of metabolite identities. Metabolite identities were assigned for peaks with a spectral match ≥ 80% in AnalyzerPro (0.7X forward match percentage + 0.3X reverse match percentage) in at least one species and treatment. Peaks were aligned across samples using Metalab [55] as well as manual curation.

Peak areas of each sample were normalized according to the volume of deionized water in which the sample was collected, as well as according to root system dry mass. We then conducted a principal component analysis (PCA) of log-transformed normalized peak areas for visualization of overall differences among species and nutrient treatments in root exudate composition. Missing values comprised a very small portion (less than 4%) of the dataset compared to the typically 10–20% of missing values reported in GC-MS studies [56,57]; therefore, to generate the full dataset required for PCA, missing values were imputed using the mean of the given species and treatment [5860]. In addition to examining differences in overall root exudate composition by PCA, we also examined whether species native to LNS and their sister species native to HNS differ in exudation of individual metabolites in each of the three separate evolutionary lineages. Therefore, peak areas of each detected metabolite were modeled within each of the three Helianthus clades separately, using two-way ANOVA with species, nutrient treatment, and their interaction as explanatory variables. Data were log-transformed as needed to better approximate ANOVA assumptions of normality and homoscedasticity. Effects were considered significant at p < 0.05. Significant effects in ANOVA were extremely congruent when missing data were excluded and when they were replaced with means as described for the PCA; therefore, we report p-values calculated with the missing values excluded. All statistical analyses were completed in JMP Pro v. 11 (SAS Institute Inc., Cary, NC).


In total, 37 metabolites were identified across species and treatments, including organic and amino acids, sugars, and sugar derivatives (Table 1). In a principal components analysis of root exudate composition across all individuals, the majority of the variation was accounted for by two principal components (Fig 2). The first principal component (PC1), which accounted for 64.3% of the overall variability among samples, revealed a strong influence of phylogeny, as the H. petiolaris-H. praecox clade (shown in triangle symbols) largely clustered separately from the other two clades along the PC1 axis (Figs 2 and 3A). This can be attributed to an overall higher abundance of most metabolites in the H. petiolaris-H. praecox clade than the other four species (Figs 4 and 5), consistent with the positive loadings of all detected metabolites for PC1 (Table 1).

Fig 2. Principal components analysis of abundance of 37 metabolites in root exudates of six Helianthus species.

Each point represents an individual seedling. Species native to low nutrient soils are filled symbols, while those native to high nutrient soils are open symbols. Seedlings in the low nutrient treatment (L) are black symbols and seedlings in the high nutrient treatment (H) are red symbols.

Fig 3. Scores of first (a) and second (b) axes (PC1 and PC2, respectively) of the principal components analysis for root exudate composition of six Helianthus species depicted in Fig 3.

H. annuus (H. ann), H. argophyllus (H. arg), H. petiolaris (H. pet), H. praecox (H. pra), H. grossesseratus (H. gro); H. microcephalus (H. mic). Data are the mean of 5–7 replicates (± SE) for seedlings treated with either low (black bars) or high (red bars) nutrient treatments. Species are arranged by clade, with the species native to low nutrient soils (relative to its sister species) indicated by (§). Significant differences between species or treatments as assessed by t-tests indicated by (*).

Fig 4. Normalized peak areas of organic acids, amino acids, and derivatives detected under low (black bars) and high (red bars) nutrient treatments in six Helianthus species.

Data are the mean of 3–7 replicates (± SE) for which those peaks were present. Species are arranged by clade, with the species native to low nutrient soils (relative to its sister species) indicated by (§). γ-guanidobutyric acid (GBA); α-keto-glutaric acid (KGA). Species abbreviations as in Fig 3.

Fig 5. Normalized peak areas of sugars and sugar derivatives detected under low (black bars) and high (red bars) nutrient treatments in six Helianthus species.

Data are the mean of 3–7 replicates (± SE) for which those peaks were present. Species are arranged by clade, with the species native to low nutrient soils (relative to its sister species) indicated by (§).Glycerol-1-phosphate (G-1-P); galacturonic acid (GalA); 2-O-Glycerol-galactopyranoside (GG). Species abbreviations as in Fig 3.

Table 1. Loading scores (± SE) of the detected compounds on the first and second principal components axes (PC1 and PC2, respectively) depicted in Fig 2.

The clear influence of phylogeny on root exudate composition seen in the principal component analysis (Fig 2) reinforced the need to assess the dataset using phylogenetically-independent contrasts (i.e. comparing root exudate composition within each LNS-HNS species pair) to control for species’ evolutionary relationships. In the high nutrient treatment, the mean PC1 score of each species native to LNS was significantly higher than its sister species native to HNS, although this difference was marginally significant between H. annuus and H. argophyllus (p = 0.07; Figs 2 and 3A). This is consistent with the general observation of higher abundance of individual metabolites in species native to LNS than their sister species native to HNS under high nutrient supply (Figs 4 and 5). The results for the high nutrient treatment differed from those for the low nutrient treatment: in the low nutrient treatment, sister species tended to cluster together (Fig 2), and analysis of PC1 scores confirmed that species native to LNS did not significantly differ from their HNS sister species (Fig 3A).

The second principal component (PC2), which accounted for an additional 9.8% of the variance, primarily revealed the influence of nutrient treatments, as indicated by the clear separation between seedlings in the low nutrient treatment (black symbols) and high nutrient treatment (red symbols) on this axis (Figs 2 and 3B). All species exhibited a significantly higher PC2 score under the low nutrient treatment, relative to the high nutrient treatment (Figs 2 and 3B). Organic acids, hexoses, and amino acids were the main drivers of this nutrient response based on PC2 loadings (Table 1). Regardless of species’ native site fertility, low nutrient treatment generally resulted in higher exudation of organic acids (fumaric acid, citric acid, and malic acid) and hexoses (e.g. glucose), and lower exudation of amino acids (glutamic acid, aspartic acid; Tables 1 and 2; Figs 4 and 5) than in the high nutrient treatment. However, significant species by nutrient treatment interaction terms in the ANOVA for 22 of the 37 detected metabolites, including sugars, organic acids, and amino acids, indicated that species native to LNS and those native to HNS responded differently to changes in nutrient supply for some, but not all, metabolites (Table 2). All of those interaction terms were due to relatively larger increases (or smaller decreases in the case of amino acids) in response to low nutrient supply in species native to HNS than their sister species native to LNS (Table 2; Figs 4 and 5). No significant species by nutrient supply interactions were shared across all three clades, indicating that for any given metabolite, species of HNS did not consistently differ from species of LNS in the response to changes in nutrient supply (Table 2; Figs 4 and 5).

Table 2. Significant (p < 0.05) p-values of the effects of species (Sp), nutrient treatment (N), and their interaction as assessed by two-way ANOVA on the abundance of root exudates within three different Helianthus clades.


In this controlled hydroponic study, we detected large variation in root exudate composition among six wild Helianthus species chosen as phylogenetically-independent contrasts with respect to native soil fertility. Although root exudation has been shown to differ quantitatively between hydroponically-grown plants and plants grown in a solid soil substrate [6163], differences among genotypes and treatments are widely accepted to be maintained in both hydroponics and soil [18,23,25]. Thus, our objective was to demonstrate relative quantitative differences in root exudation among species and nutrient treatments that are likely maintained in soil, rather than absolute quantities of metabolites which may have little relevance for soil conditions.

Focusing first on the high nutrient treatment, species native to LNS exhibited higher exudation than their sister species native to HNS in all three clades, providing evidence for repeated evolutionary shifts that may be of adaptive significance. Higher overall exudation of sugars, amino acids, and organic acids in LNS species may represent an adaptive strategy for initiating beneficial root-microbe associations, or for accessing relatively immobile nutrients such as phosphorus in the native habitats of these species [64,65]. However, in contrast to the differences between LNS and HNS species within each clade in the high nutrient treatment, no such differences were detected in the low nutrient treatment. Based on similar PCA clustering, no shifts in exudate composition between species native to LNS and their sister taxa native to HNS were observed within any of the three clades, indicating that there is no evidence for consistent differentiation among species under low nutrient supply.

In addition to inherent differences in root exudate composition expressed under a single level of nutrient supply, species responses to nutrient deficiency may also be adaptive. For example, increased exudation of organic acids in response to low nutrient supply has been detected in numerous species [25,27,37,43,48], and has been hypothesized as a mechanism to release mineral nutrients into soil solution through ligand exchange and/or dissolution of soil minerals [9,27,28], while increased exudation of sugars in response to nutrient deficiency may lead to shifts in microbial community structure in the rhizosphere [25,29,30]. Most of the Helianthus species generally responded to low nutrient treatments with higher exudation of organic acids (fumaric acid, citric acid, and malic acid) and hexoses (e.g. glucose) than under high nutrient supply (with the exception of H. microcephalus). The magnitude of this response was similar among species native to LNS and their sister species native to HNS in both of the annual clades, suggesting that the LNS species are not adapted to infertile soils through a unique capacity to increase organic acid and/or sugar exudation in response to nutrient deficiency. Rather, our results suggest that increased exudation of these metabolites is a generally conserved response to nutrient deficiency in Helianthus, potentially as a mechanism to improve nutrient acquisition. The only exception was the LNS species H. microcephalus, which shifted towards lower exudation of nearly all detected metabolites in the low nutrient treatment relative to the high, suggesting that increased exudation of the primary metabolites detected in this study in response to nutrient deficiency is not a mechanism by which this species is adapted to infertile soils.

Although Helianthus species generally responded to nutrient deficiency with increased exudation of several organic acids as well as glucose, species native to LNS and their sister species native to HNS differed in their response to nutrient deficiency for a number of other metabolites. For 15 of the 22 metabolites with a significant species by nutrient treatment interaction term in ANOVA, species native to HNS had a larger increase in abundance in response to the low nutrient treatment than their sister species native to LNS. These differential nutrient responses were largely in the H. grossesseratus-H. microcephalus clade, due to H. microcephalus responding to low nutrients with lower exudation of nearly all metabolites detected in this study. The remaining seven metabolites for which species native to LNS and those native to HNS responded differently to nutrient deficiency (i.e. the remaining seven metabolites with a significant species by nutrient interaction term in ANOVA) were all amino acids. In all three Helianthus clades, species native to LNS exhibited a larger decrease in amino acid exudation in response to low nutrient treatments than their sister taxa native to HNS. Decreased exudation of amino acids is a common response to nutrient (particularly nitrogen) deficiency [25,48,66,67], and has been suggested to result from lower amino acid content in roots of nitrogen-deficient plants [25,68]. However, both amino acid influx and efflux are influenced by metabolic inhibitors, suggesting that amino acid exudation results from the combined effects of passive diffusion and active regulation by the plant [69]. For example, amino acid uptake transporters in roots are up-regulated under nitrogen-deficiency [70], resulting in the higher amino acid uptake capacity seen in plants grown under low nitrogen supply [71] and allowing re-uptake of amino acids that were previously exuded into the rhizosphere. The large reductions in net amino acid exudation under nutrient limitation detected in species native to LNS may result from selection to either minimize efflux, maximize influx, or both, for a greater capacity to conserve nitrogen-rich metabolites in LNS.

Taken together, our findings indicate that species native to LNS exhibit higher overall root exudation than species native to HNS under high nutrient supply, and that species native to contrasting soil fertilities exhibit a unique sensitivity to nutrient deficiency for some, but not all, metabolites. Given the unpredictability of nutrient pulses in LNS, there may be selection in LNS for species to maintain generally high exudation rates in order to capitalize during brief periods of high nutrient availability. Such a strategy may help plants of LNS to exploit temporary nutrient pulses, and to prime the rhizosphere for future situations of low nutrient availability by stimulating microbial nutrient cycling [72,73]. Alternatively, high exudation of carbon-rich compounds in LNS may lead to nutrient immobilization, reducing soil nutrient concentrations below that which can be tolerated by fast-growing species with high nutrient requirements [73,74]. Either mechanism is likely to improve plant fitness in LNS, resulting in selection for high exudation rates. In contrast, the lower root exudation in species native to HNS under high nutrient supply may suggest that these species have adapted to their environment by mechanisms other than root exudation. Species native to chronically HNS may rely on their generally fast root growth rates to a greater extent than root exudation in order to rapidly exploit nutrients and gain an edge over competitors in high nutrient conditions [75,76]. In response to nutrient deficiency, however, the increased exudation of organic acids and hexoses detected for HNS species (as with LNS species) may be a means for chemically or biologically (i.e. microbial mining) increasing nutrient acquisition [73,77]. Given that root tissues are expected to be longer-lived in species native to LNS, the rapid root growth strategy is likely not viable in LNS, even in nutrient-rich soil patches, as rapid production of new roots may be too expensive to maintain as those patches are depleted [7880]. Species native to such LNS, rather, are generally expected to grow slowly, but maintain long-lived root systems which can tolerate extended periods of low nutrient availability [7879]. For plants native to LNS, root exudation may therefore represent a more cost-effective strategy for maximizing nutrient acquisition than allocating resources for increased root length production.


We would like to thank the Academic Editor and two anonymous reviewers for their helpful comments which improved this manuscript, S. Bowsher for her assistance with the hydroponics system, B. Nyamdari for assistance with GC-MS, C. Mason, D. Grise, and M. Ungerer for help with locating and collecting seeds and soil cores, and the Donovan lab group for helpful comments on early versions of this manuscript.

Author Contributions

Conceived and designed the experiments: AWB SAH CJT LAD. Performed the experiments: AWB RA. Analyzed the data: AWB. Contributed reagents/materials/analysis tools: AWB SAH CJT LAD. Wrote the paper: AWB RA SAH CJT LAD.


  1. 1. Coudun C, Gégout J-C, Piedallu C, Rameau J- C. Soil nutritional factors improve models of plant species distribution: an illustration with Acer campestre (L.) in France. Journal of Biogeography. 2006;33: 1750–1763.
  2. 2. John R, Dalling JW, Harms KE, Yavitt JB, Stallard RF, Mirabello M, et al. Soil nutrients influence spatial distributions of tropical tree species. Proceedings of the National Academy of Sciences of the United States of America. 2007;104: 864–869. pmid:17215353
  3. 3. Condit R, Engelbrecht BMJ, Pino D, Pérez R, Turner BL. Species distributions in response to individual soil nutrients and seasonal drought across a community of tropical trees. Proceedings of the National Academy of Sciences of the United States of America. 2013;11: 5064–5068.
  4. 4. Lynch JP. Root architecture and plant productivity. Plant Physiology. 1995;109: 7–13. pmid:12228579
  5. 5. Marschner H, Römheld V, Horst WJ, Martin P. Root-induced changes in the rhizosphere: importance for the mineral nutrition of plants. Journal of Plant Nutrition and Soil Science. 1986;149: 441–456.
  6. 6. Hinsinger P. Bioavailability of soil inorganic P in the rhizosphere as affected by root-induced chemical changes: a review. Plant and Soil. 2001;237: 173–195.
  7. 7. Jones DL, Nguyen C, Finlay RD. Carbon flow in the rhizosphere: carbon trading at the soil-root interface. Plant and Soil. 2009;321: 5–33.
  8. 8. Moghimi A, Lewis DG, Oades JM. Release of phosphate from calcium phosphates by rhizosphere products. Soil Biology and Biochemistry. 1978;10: 277–281.
  9. 9. Gerke J, Römer W, Jungk A. The excretion of citric and malic acid by proteoid roots of Lupinus albus L.; effects on soil solution concentrations of phosphate, iron, and alumimum in the proteoid rhizosphere in samples of an oxisol and a luvisol. Journal of Plant Nutrition and Soil Science. 1994;157: 289–294.
  10. 10. Jones DL, Darrah PR. Role of root derived organic acids in the mobilization of nutrients from the rhizosphere. Plant and Soil. 1994;166: 247–257.
  11. 11. Veneklaas DJ, Stevens J, Cawthray GR, Turner S, Grigg AM, Lambers H. Chickpea and white lupin rhizosphere carboxylates vary with soil properties and enhance phosphorus uptake. Plant and Soil. 2003;248: 187–197.
  12. 12. Akiyama K, Matsuzaki K, Hayashi H. Plant sesquiterpenes induce hyphal branching in arbuscular mycorrhizal fungi. Nature Letters. 2005;435: 824–827.
  13. 13. López-Ráez JA, Charnikhova T, Gómez-Roldán V, Radoslava M, Kohlen W, De Vos R, et al. Tomato strigolactones are derived from carotenoids and their biosynthesis is promoted by phosphate starvation. New Phytologist. 2008;178: 863–874. pmid:18346111
  14. 14. Cesco S, Neumann G, Tomasi N, Pinton R, Weisskopf L. Release of plant-borne flavonoids into the rhizosphere and their role in plant nutrition. Plant and Soil. 2010;329: 1–25.
  15. 15. Neumann G. Root exudates and nutrient cycling. In: Marschner P, Rengel Z, editors. Nutrient cycling in Ecosystems, Soil Biology, Vol. 10. Berlin, Heidelberg: Springer; 2007. pp. 123–157.
  16. 16. Henry S, Texier S, Hallet S, Bru D, Dambreville C, Chéneby D, et al. Disentangling the rhizosphere effect on nitrate reducers and denitrifiers: insight into the role of root exudates. Environmental Microbiology. 2008;10: 3082–3092. pmid:18393993
  17. 17. Dakora FD, Phillips DA. Root exudates as mediators of mineral acquisition in low-nutrient environments. Plant and Soil. 2002;245: 35–47.
  18. 18. Tyler G, Strom L. Differing organic acid exudation pattern explains calcifuge and acidifuge behavior of plants. Annals of Botany. 1995;75: 75–78. pmid:21247915
  19. 19. Grayston SJ, Vaughan D, Jones D. Rhizosphere carbon flow in trees, in comparison with annual plants: the importance of root exudation and its impact on microbial activity and nutrient availability. Applied Soil Ecology. 1996;5: 29–56.
  20. 20. Pearse SJ, Veneklaas EJ, Cawthray G, Bolland MDA, Lambers H. Carboxylate composition of root exudates does not relate consistently to a crop species’ ability to use phosphorus from aluminum, iron, or calcium phosphate sources. New Phytologist. 2007;173: 181–190. pmid:17176404
  21. 21. Johnson JF, Vance CP, Allan DL. Phosphorus deficiency in Lupinus albus. Altered lateral root development and enhanced expression of phophoenolpyruvate carboxylase. Plant Physiology. 1996;112: 31–41. pmid:8819319
  22. 22. Jones DL. Organic acids in the rhizosphere–a critical review. Plant and Soil. 1998;205: 25–44.
  23. 23. Neumann G, Römheld V. Root excretion of carboxylic acids and protons in phosphorus-deficient plants. Plant and Soil. 1999;211: 121–130.
  24. 24. Roelofs RFR, Rengel Z, Cawthray CG, Dixon KW, Lambers H. Exudation of carboxylates in Australian Proteaceae: chemical composition. Plant Cell and Environment. 2001;24: 891–903.
  25. 25. Carvalhais LC, Dennis PG, Fedoseyenko D, Hajirezael M-R, Borriss R, von Wiren N. Root exudation of sugars, amino acids, and organic acids by maize as affected by nitrogen, phosphorus, potassium, and iron deficiency. Journal of Plant Nutrition and Soil Science. 2011;174: 3–11.
  26. 26. Wouterlood M, Cawthray GR, Scanlon T, Lambers H, Veneklaas EJ. Carboxylate concentrations in the rhizosphere of lateral roots of chickpea (Cicer arietinum) increase during plant development, but are not correlated with phosphorus status of soil or plants. New Phytologist. 2004;162: 745–753.
  27. 27. Gardner WK, Barber DA, Parbery DG. The acquisition of phosphorus by Lupinus albus L. III. The probable mechanism by which phosphorus movement in the soil/root interface is enhanced. Plant and Soil. 1983;70: 19–32.
  28. 28. Dinkelaker B, Römheld V, Marschner H. Citric acid excretion and precipitation of calcium citrate in the rhizosphere of white lupin (Lupinus albus L.). Plant Cell and Environment. 1989;12: 285–292.
  29. 29. Badri DV, Weir TL, van der Lelie D, Vivanco JM. Rhizosphere chemical dialogues: plant-microbe interactions. Current Opinion in Biotechnology. 2009;20: 642–650. pmid:19875278
  30. 30. Shi S, Richardson AE, O’Callaghan M, DeAngelis KM, Jones EE, Stewart A, et al. Effects of selected root exudate components on soil bacterial communities. FEMS Microbiology Ecology. 2011;77: 600–610. pmid:21658090
  31. 31. Lambers H, Clements JC, Nelson MN. How a phosphorus-acquisition strategy based on carboxylate exudation powers the success and agronomic potential of lupines (Lupinus, Fabaceae). American Journal of Botany. 2013;100: 263–288. pmid:23347972
  32. 32. De Deyn GB, Cornellissen JHC, Bardgett RD. Plant functional traits and soil carbon sequesyration in contrasting biomes. Ecology Letters. 2008;11: 516–531. pmid:18279352
  33. 33. Orwin KH, Buckland SM, Johnson D, Turner BL, Smart S, Oakley S, et al. Linkages of plant traits to soil properties and the functioning of temperate grassland. Journal of Ecology. 2010;98: 1074–1083.
  34. 34. van der Krift TAJ, Gioacchini P, Kuikman PJ, Berendse F. Effects of high and low fertility plant species on dead root decomposition and nitrogen mineralization. Soil Biology and Biochemistry 2001;33: 2115–2124.
  35. 35. Rengel Z. Genetic control of root exudation. Plant and Soil. 2002;245: 59–70.
  36. 36. Strom L. Root exudation of organic acids: importance to nutrient availability and the calcifuge and calcicole behavior of plants. Oikos. 1997;80: 459–466.
  37. 37. Denton MD, Veneklaas EJ, Lambers H. Does phenotypic plasticity in carboxylate exudation differ among rare and widespread Banksia species (Proteaceae)? New Phytologist. 2007;173: 592–599. pmid:17244054
  38. 38. Felsenstein J. Phylogenies and the comparative method. The American Naturalist. 1985;125: 1–15.
  39. 39. Harvey PH, Purvis A. Comparative methods for explaining adaptations. Nature. 1991;351: 619–624. pmid:2052088
  40. 40. Nicotra AB, Babicka N, Westoby M. Seedling root anatomy and morphology: an examination of ecological differentiation with rainfall using phylogenetically independent contrasts. Oecologia. 2002;130: 136–145.
  41. 41. Pineda-García F, Paz H, Tinoco-Ojanguren C. Morphological and physiological differentiation of seedlings between dry and wet habitats in a tropical dry forest. Plant Cell and Environment. 2011;34: 1536–1547.
  42. 42. Raupp FM, Spring O. New sesquiterpene lactones from sunflower root exudate as germination stimulants for Orobanche cumana. Journal of Agricultural and Food Chemistry. 2013;61: 10481–10487. pmid:24117219
  43. 43. Neumann G, Römheld V. The release of root exudates as affected by the plant’s physiological status. In: Pinton R, Varanini Z, Nannipieri P, editors. The Rhizosphere: Biochemistry and Organic Substances at the Soil-Plant Interface. New York: Dekker; 2001. pp. 23–72.
  44. 44. Heiser CBJ, Smith DM, Clevenger SB, Martin WCJ. The North American sunflowers: Helianthus. Memoirs of the Torrey Botanical Club. 1969;22: 1–218.
  45. 45. Kane NC, Burke JM, Marek L, Seiler G, Vear F, Baute G, et al. Sunflower genetic, genomic, and ecological resources. Molecular Ecology Resources. 2013;13: 10–20. pmid:23039950
  46. 46. Mason CM, Donovan LA. Evolution of the leaf economics spectrum in herbs: Evidence from environmental divergences in leaf physiology across Helianthus (Asteraceae). Evolution. 2015;69: 2705–2720. pmid:26339995
  47. 47. Stephens JD, Rogers WL, Mason CM, Donovan LA, Malmberg RL. Species tree estimation of diploid Helianthus (Asteraceae) using target enrichment. American Journal of Botany. 2015;102: 910–920. pmid:26101417
  48. 48. Bowsher AW, Ali R, Harding SA, Tsai C-J, Donovan LA. Analysis of wild sunflower (Helianthus annuus L.) root exudates using gas chromatography-mass spectrometry. Journal of Plant Nutrition and Soil Science. 2015;178: 776–786.
  49. 49. Epstein E, Bloom AJ. Mineral nutrition of plants: principles and perspectives. Sunderland: Sinauer Associates; 2005.
  50. 50. Aulakh MS, Wassmann R, Bueno C, Kreuzwieser J, Rennenberg H. Characterization of root exudates at different growth stages of ten rice (Oryza sativa L.) cultivars. Plant Biology. 2001;3: 139–148.
  51. 51. Jeong ML, Jians H, Chen H-S, Tsai CJ, Harding SA. Metabolite profiling of the sink-to-source transition in developing leaves of quaking aspen. Plant Physiology. 2004;136: 3364–3375. pmid:15448196
  52. 52. Frost CJ, Nyamdari B, Tsai CJ, Harding SA. The tonoplast-localized sucrose transporter in Populus (PtaSUT4) regulates whole-plant water relations, responses to water stress, and photosynthesis. PLoS ONE. 2012;7:e44467. pmid:22952983
  53. 53. Babushok VI, Linstrom PJ, Reed JJ, Zenkevich IG, Brown RL, Mallard , et al. Development of a database of gas chromatographic retention properties of organic compounds. Journal of Chromatography. 2007;1157: 414–421. pmid:17543315
  54. 54. Kind T, Wohlgemuth G, Lee DY, Lu Y, Palazoglu M, Shahbaz S, Feihn O. FiehnLib: mass spectral and retention index libraries for metabolomics based on quadrupole and time-of-flight gas chromatography/mass spectrometry. Analytical Chemistry. 2009;81: 10038–10048. pmid:19928838
  55. 55. Xue L- J, Guo W, Yuan Y, Anino EO, Nyamdari B, Wilson MC, et al. Constitutively elevated salicylic acid levels alter photosynthesis and oxidative state but not growth in transgenic Populus. Plant Cell. 2013;7: 2714–30.
  56. 56. Xia J, Psychogios N, Young N, Wishart DS. MetaboAnalyst: A web server for metabolomic data analysis and interpretation. Nucleic Acids Research. 2009;37: W652–W660. pmid:19429898
  57. 57. Gromski PS, Xu Y, Kotze HL, Correa E, Ellis DI, Armitage EG, et al. Influence of missing values substitutes on multivariate analysis of metabolomics data. Metabolites. 2014;4: 433–452. pmid:24957035
  58. 58. Dodge Y. Analysis of Experiments with Missing Data. New York: Wiley and Sons; 1985.
  59. 59. Shum H- Y, Ikeuchi K, Reddy R. Principal component analysis with missing data and its application to polyhedral object modeling. IEEE Transactions on Pattern Analysis and Machine Intelligence. 1995;17: 854–867.
  60. 60. Dray S, Josse J. Principal component analysis with missing values: a comparative survey of methods. Plant Ecology. 2015;216: 657–667.
  61. 61. Barber DA, Gunn KB. The effect of mechanical forces on the exudation of organic substances by the roots of cereal plants grown under sterile conditions. New Phytologist. 1974;73: 39–45.
  62. 62. Boeuf-Tremblay V, Plantureux S, Guckert A. Influence of mechanical impedance on root exudation of maize seedlings at two developmental stages. Plant and Soil. 1995;172: 279–287.
  63. 63. Groleau-Renaud V, Plantureux S, Guckert A. Influence of plant morphology on root exudation of maize subjected to mechanical impedance in hydroponic conditions. Plant and Soil. 1998;201: 231–239.
  64. 64. Lambers H, Poorter H. Inherent variation in growth rate between higher plants: a search for physiological causes and ecological consequences. Advances in Ecological Research. 1992;34: 187–261.
  65. 65. Moe LA. Amino acids in the rhizosphere: from plants to microbes. American Journal of Botany. 2013;100: 1692–1705. pmid:23956051
  66. 66. Bowen GD. Nutrient status effects on loss of amides and amino acids from pine roots. Plant and Soil. 1969;1: 139–142.
  67. 67. Haase S, Neumann G, Kania A, Kuzyakov Y, Römheld V, Kandeler E. Elevation of atmospheric CO2 and N-nutritional status modify nodulation, nodule-carbon supply, and root exudation of Phaseolus vulgaris L. Soil Biology and Biochemistry. 2007;39: 2208–2221.
  68. 68. von Wirén N, Lauter FR, Ninnemann O, Gillissen B, Walch-Liu P, Engels C, et al. Differential regulation of three functional ammonium transporter genes by nitrogen in root hairs and by light in leaves of tomato. The Plant Journal. 2000;21: 167–175. pmid:10743657
  69. 69. Lesuffleur F, Cliquet J-B. Characterisation of root amino acid exudation in white clover (Trifolium repens L.). Plant and Soil. 2010;333: 191–201.
  70. 70. Hirner A, Ladwig F, Stransky H, Okumoto S, Keinath M, Harms A, et al. Arabidopsis LHT1 is a high-affinity transporter for cellular amino acid uptake in both root epidermis and leaf mesophyll. Plant Cell. 2006;18: 1931–1946. pmid:16816136
  71. 71. Persson J, Näsholm T. Regulation of amino acid uptake in conifers by exogenous and endogenous nitrogen. Planta. 2002;215: 639–644. pmid:12172847
  72. 72. Kuzyakov Y, Cheng W. Photosynthesis controls of rhizosphere respiration and organic matter decomposition. Soil Biology & Biochemistry. 2001;33: 1915–1925.
  73. 73. Dijkstra FA, Carrillo Y, Pendall E, Morgan JA. Rhizosphere priming: a nutrient perspective. Frontiers in Microbiology. 2013;4: 1–8.
  74. 74. Meier CL, Keyserling K, Bowman WD. Fine root inputs to soil reduce growth of a neighbouring plant via distinct mechanisms dependent on root carbon chemistry. Journal of Ecology. 2009;97: 941–949.
  75. 75. Grime JP. Plant Strategies and Vegetation Processes. Chichester: Wiley and Sons; 1979.
  76. 76. Craine JM, Dybzinski R. Mechanisms of plant competition for nutrients, water, and light. Functional Ecology. 2013;27: 833–840.
  77. 77. Craine JM, Morrow C, Fierer N. Microbial nitrogen limitation increases decomposition. Ecology. 2007;88: 2105–2113. pmid:17824441
  78. 78. Grime JP. The role of plasticity in exploiting environmental heterogeneity. In: Caldwell MM, Pearcy RW, editors. Exploitation of Environmental Heterogeneity by Plants: Ecophysiological Processes Above- and Belowground. San Diego, Academic Press; 1994. pp. 1–19.
  79. 79. Hutchings MJ, de Kroon H. Foraging in plants: the role of morphological plasticity in resource acquisition. Advances in Ecological Research. 1994;25: 159–238.
  80. 80. Fransen B, de Kroon H, Berendse F. Root morphological plasticity and nutrient acquisition of perennial grass species from habitats of different nutrient availability. Oecologia. 1998;115: 351–358.