Biological inhibition of denitrification (BDI): an early plant strategy for Fallopia × bohemica seedling development

Abstract

Background and Aims

The successful plant Fallopia × bohemica presents interesting capacities for control of the soil nitrogen cycle at the adult stage, termed biological inhibition of denitrification (BDI). The BDI strategy allows the plant, via the production of secondary metabolites (procyanidins), to compete with the denitrifying microbial community and to divert nitrate from the soil for its benefit. In this study, we analysed whether seedlings of F. × bohemica can implement BDI at the seedling stage. We also determined whether soil nitrogen availability influences the implementation of BDI and seedling growth.

Methods

We sowed achenes of F. × bohemica in soils representing a nitrogen gradient (six treatments) and harvested seedlings after 20 or 40 days of growth. The denitrification and related microbial communities (i.e. functional gene abundances of nirK and nirS), soil parameters (nitrate content and humidity) and plant performance (biomass, growth and root morphology) were determined.

Key Results

On soil without addition of nitrogen, BDI was observed after 20 days of growth, whereas a stimulation of denitrification was found after 40 days. The increase of soil N content had few effects on the activity and structure of the soil denitrifying community and on the plant biomasses or the relative growth rates. Correlations between plant and microbial parameters were observed after 20 days of growth, reflecting early and strong chemical interactions between plants and the denitrifying community, which decreased with plant growth after 40 days.

Conclusions

This study shows that an early BDI enhances the efficiency of nitrogen acquisition in the first weeks of growth, allowing for a conservative root strategy after 40 days. This switch to a conservative strategy involved resource storage, an altered allocation to above- and below-ground parts and an investment in fine roots. It now seems clear that this storage strategy starts at a very young age with early establishment of BDI, giving this clonal plant exceptional capacities for storage and multiplication.

INTRODUCTION

Some plants have exceptional capabilities for growth and colonization and fall into the category of exploitative or acquisitive plants (Gross et al., 2007). Owing to their functional traits (high photosynthesis rates, leaf span, leaf N or root N contents), these plants are fast-growing species with high ability to capture nutrient and use them efficiently (Grime, 2001). Among nutrients, nitrogen (N) is one of the major factors that limit plant growth in many ecosystems (Vitousek and Howarth, 1991; LeBauer and Treseder, 2008). Thus, plant growth depends strongly on N cycle processes that drive N transformation and availability in soils (Bothe et al., 2006). The N uptake by plants is mainly in the form of nitrate (NO₃⁻) and ammonium (NH₄⁺), and the competition between roots and microorganisms for these two forms of N can be highly intense (Hodge et al., 2000; Boudsocq et al., 2012; Kuzyakov and Xu, 2013). Microorganisms often out-compete plant roots to acquire this limited resource because of their large area-to-volume ratio and rapid growth rate (Hodge et al., 2000; Kuzyakov and Xu, 2013). However, some plants bypass this competition by taking control of the N cycle in their rhizosphere through the release of secondary metabolites that inhibit N cycle processes [i.e. N mineralization (Dietz et al., 2013), nitrification (Subbarao et al., 2009) and denitrification (Bardon et al., 2014)]. According to plant–soil feedback theories (Ehrenfeld et al., 2005; van der Putten et al., 2013), plants that modify N availability in rhizosphere soil might, in turn, influence their own growth (Subbarao et al., 2009; Bardon et al., 2014, 2017). The biological control of microbial activities of the N cycle in soils by plants is an important strategy that allows nitrogen acquisition and participates in plant growth and performance (Cantarel et al., 2015; Moreau et al., 2015; Pommier et al., 2018).

The plant model Fallopia (Japanese knotweed in the broad sense) presents interesting possibilities for the control of N cycle microorganisms. This model has allowed us, for the first time, to demonstrate the biological inhibition of denitrification (BDI) (Bardon et al., 2014). This inhibition allows the plant, via the production of secondary metabolites (procyanidins), to compete with the denitrifying microbial community of the N cycle and to divert, for its benefit, the nitrate from the microbial pathway (Dassonville et al., 2011; Bardon et al., 2014, 2016a, b, 2017). In Japan, Fallopia spp. influence N cycling in the soil. On Japanese volcanoes, Fallopia japonica is a common perennial pioneer species. In this habitat, N is a major factor limiting plant growth. The concentration of N in soil increases in the course of patch development, accumulating in organic and mineral forms (Hirose and Tateno, 1984; Tateno and Hirose, 1987) that are available for other plants in plant succession. These results show that Fallopia spp. are important primary colonizers and a particularly pertinent plant model in studies of the interactions between plant–soil and N-associated microorganisms.

Biological inhibition of denitrification has been highlighted in established Fallopia spp. in the field (Dassonville et al., 2011) or in controlled greenhouse conditions in plants regenerated from vegetative propagules (rhizome fragments) (Cantarel et al., 2020). For adult plants, BDI leads to an increase in available soil nitrate and a change in plant root traits, stimulating root system growth and their competitive capacity (Bardon et al., 2017). However, plant N demand varies throughout the plant life cycle, associated with changes in dry matter production and light interception (Perthame et al., 2020), which are strongly linked to life and phenological stages (i.e. germination compared with the reproductive stage) and/or soil N availability. For highly productive plants, such as Fallopia spp., previous studies have reported a higher germination rate (Engler et al., 2011; Lamberti-Raverot et al., 2019), higher photosynthetic activity, a greater N uptake and greater N accumulation in seedlings than in adult plants (Agren and Franklin, 2003). These results suggest that seedlings could compete more strongly with soil N microorganisms for N and develop BDI strategies at early growth stages. No study has yet shown the impacts of seedlings on the soil N and microorganism functioning, and even less for the seedlings of BDI plants, such as Fallopia spp.

The aim of this study was to analyse whether seedlings of F. × bohemica, known to develop a BDI strategy at the adult stage, can do so at early life stages and whether the N nutrition influences the establishment of BDI and the growth of seedlings. Currently, the procyanidin concentration is difficult to measure directly in soil, making root procyanidine exudation difficult to quantify (Kraus et al., 2004), notably because procyanidins bind strongly to soil components (Kraal et al., 2009). In our study, BDI is indicated by a significant decrease of denitrification under F. × bohemica planted conditions relative to unplanted conditions, expressed as a percentage (Dassonville et al., 2011, 2011; Cantarel et al., 2020). Through an experiment in controlled microcosm conditions, we tested the following hypotheses: (1) F. × bohemica seedlings develop a BDI strategy as early as 20 or 40 days after emergence; (2) low soil N content triggers the establishment of BDI; and (3) the BDI strategy allows the seedlings to improve their performance with growth advantages, such as higher growth rate and biomass. To test these hypotheses, measurements on the denitrifying microbial community in the soil (denitrification and denitrifying gene abundances) and on plant traits (biomasses and root morphology) were made at early stages of plant development [seedlings at 20 and 40 days of growth after emergence, i.e. stages 12 and 13 respectively, on the Biologische Bundesanstalt, Bundessortenamt und CHemical Industry (BBCH) scale] along a nitrate/ammonium mixture gradient.

MATERIALS AND METHODS

Biological material and nitrogen gradient

Fallopia × bohemica is rhizomatous plant of the Polygonaceae family. Fallopia × bohemica was chosen for this experiment as a plant model adopting a BDI strategy (Bardon et al., 2014). The Fallopia complex includes F. japonica, Fallopia sachalinensis, the complex of polyploid F. × bohemica hybrids (Chrtek & Chrtková), originating from the hybridization of F. japonica and F. sachalinensis, backcrosses and F2 and other crosses involving Fallopia baldschuanica (Bailey et al., 2009). Within the complex, F. × bohemica presents the highest genetic and morphological variation (Mandák et al., 2005; Tiébré et al., 2007). Although the spread of F. × bohemica has been considered mainly as clonal, recent sexual spread has been shown in Europe (Rouifed et al., 2011; Lamberti-Raverot et al., 2017, 2019). Sexual propagules consist of a single-seeded achene that is surrounded by a winged fruiting perianth. Fallopia × bohemica produces a large number of fertile achenes each year (Lamberti-Raverot et al., 2019). Achenes of F. × bohemica were harvested from a unique F. japonica individual at the Chambon-Feugerolles site [Loire (42), France, 45°23ʹ49″N, 4°18ʹ33″E] in winter 2021 and were stored at 4 °C until the experimentation.

The soil was harvested at Saint Maurice de Gourdans [Ain (01), France, 45°48ʹ15″N, 5°10ʹ29″E] 5 days before the start of the experiment, sieved to 4 mm and stored at 4 °C, in order to limit any changes in mineral nitrogen levels and microbial communities. This soil was chosen: (1) for its physicochemical characteristics (Table 1), with low mineral N contents allowing the creation of a N gradient; and (2) because Fallopia spp. have never been detected at this site.

Table 1.

Saint Maurice de Gourdans [Ain (01), France, 45°48ʹ15″N, 5°10ʹ29 ″E] soil characteristics at the start of the experiment (termed initial soil conditions): soil nitrate concentration (NO₃⁻), soil ammonium concentration (NH₄⁺), organic matter (OM), total carbon (C), total nitrogen (N), soil texture (clays/silts/sands), total carbonates (Cbt) and cation exchange capacity (CEC).

pH	Moisture	NH4⁺	NO3⁻	Ratio	Soil texture	Clays	Silts	Sands	OM	C	N	Cbt	CEC
	(% for 100 g of dry soil)	(μg N-NH4⁺ g−1 dry soil)	(μg NNO3⁻ g−1 dry soil)	NH4+/NO3−	Jamagne diagram (1967)	(g kg−1)	(g kg−1)	(g kg−1)	(g kg−1)	(g kg−1)	(g kg−1)	(g kg−1)	(mol kg−1)
8.01	16.49	0.279	0.761	0.366	Sandy-clay loam	245	304	451	59.93	34.84	3.88	479	120.89

A gradient was carried out with six N enrichment treatments (Table 2), with respect to the initial ammonium/nitrate coefficient of soil (0.36; see Table 1), by the addition of 0, 4.1, 6.83, 17.08, 27.33 and 68.33 of N (µgN-NH₄⁺+NO₃⁻ g⁻¹ dry soil) to the soil, hereafter termed no add, Add-N1, Add-N2, Add-N3, Add-N4 and Add-N5, respectively. Enrichments were prepared by adding KNO₃ and (NH₄)₂SO₄ powder to 150 mL of distilled water, as shown in Table 2. Each N enrichment was added to 2 kg of soil with a sprayer, in order to moisten the soil, add the solution evenly and blend the soil.

Table 2.

The studied nitrogen gradient: six nitrogen enrichments with ascending values of nitrate and ammonium, with respect to the initial ammonium/nitrate coefficient in the initial soil (0.36; see initial values of soil ammonium and nitrate concentration in Table 1).

Nitrogen treatment	Nitrate added (µg N-NO3− g−1 dry soil)	Ammonium added (µg N-NH4+ g−1 dry soil)	Total N added (µg N-NH4+NO3− g−1 dry soil)
no add	0	0	0
Add-N1	3	1.1	4.1
Add-N2	5	1.83	6.83
Add-N3	12.5	4.58	17.08
Add-N4	20	7.33	27.33
Add-N5	50	18.33	68.33

Experimental design and harvests

The six treatments (one treatment without N and five with N enrichment) were each distributed in 14 pots (7 cm × 7 cm) with 140 g of soil (±1 g) per pot. In total, the experiment had 84 pots. In each 14 pots by treatment, 12 were sown with an achene of F. × bohemica and two remained unplanted (control without plant). After sowing, all pots were randomly placed in a phytotron (Sanyo MLR-351, versatile environmental test chamber) at the ‘Serre et Chambres Climatiques’ platform (FR BioEEnVis UCBLyon1) regulated at 22 °C with 16 h light–8 h dark. Every 2 days, pots were watered with 20 mL of water, their places were randomly changed in the phytotron to avoid position effects and weeds removed.

The plant and soil samples were collected at early vegetative phenological stages, after 20 days (T20) or 40 days (T40) of plant growth after emergence (Fig. 1A), corresponding to seedlings at the 12th and 13th stages, respectively, on the BBCH scale (Hack et al., 1992). Fallopia × bohemica emergence (i.e. 09 on the BBCH scale) occurs 8 days after sowing. Thus, 20 days after this emergence, plants reach the 12th stage, with two expanded leaves, and 40 days after emergence, the 13th stage, with three expanded leaves. Four planted pots and one unplanted pot were harvested by treatment and harvest time, for a total of 30 samples by harvest date. The T20 pots were selected randomly for the T20 harvest. The T20 harvest had no effect on the T40 pot cohort, because each plant was in an individual pot. The four replicates of planted soil by treatment were used as follows: (1) for plants, to measure the plant functional traits; and (2) to collect rhizosphere soil (the soil under root influence), to analyse the microbial activities and the microbial functional gene abundances. The unplanted soils were used only to test the effect of plant presence compared with unplanted soil (as a percentage), which was calculated as follows:

Fig. 1.

Fallopia × bohemica seedlings after 20 and 40 days of growth after emergence. (A) Plant general morphology. (B) Focus on the root system after using the p-dimethylaminocinnamaldehyde method (Prior et al., 2010) to highlight production of soluble proanthocyanidins by the root system (red arrows indicate blue colour conferred by the p-dimethylaminocinnamaldehyde method when proanthocyanidins are produced).

$${\displaystyle \begin{array}{l}{\displaystyle \mathrm{P}\mathrm{l}\mathrm{a}\mathrm{n}\mathrm{t}\mathrm{e}\mathrm{f}\mathrm{f}\mathrm{e}\mathrm{c}\mathrm{t}(\mathrm{\%})=100\times \frac{(X_i-X_{\mathrm{U}\mathrm{N}})}{\left(X_{\mathrm{U}\mathrm{N}}\right)}}\end{array}}$$

where X is the tested variable for a planted treatment, i is the individual sample and X_UN is the value for the tested variable for the unplanted treatment.

For plants, individual above-ground and root parts were collected separately from each pot to measure biomass and root traits. Denitrification activities and the soil nitrate concentration were measured on fresh soil stored at 4 °C until analysis, and a subsample was frozen immediately at −20 °C for DNA extraction and gene abundance analysis.

Plant traits

After sowing, analysis of seedling emergence was conducted every 2 days during the experiment. For each treatment, total survival (as a percentage) was calculated using this equation:

$${\displaystyle \begin{array}{l}{\displaystyle \mathrm{T}\mathrm{o}\mathrm{t}\mathrm{a}\mathrm{l}\mathrm{s}\mathrm{u}\mathrm{r}\mathrm{v}\mathrm{i}\mathrm{v}\mathrm{a}\mathrm{l}(\mathrm{\%})=100\times \frac{(\mathrm{t}\mathrm{o}\mathrm{t}\mathrm{a}\mathrm{l}\mathrm{s}\mathrm{o}\mathrm{w}\mathrm{n}\mathrm{p}\mathrm{l}\mathrm{a}\mathrm{n}\mathrm{t}\mathrm{s}-\mathrm{t}\mathrm{o}\mathrm{t}\mathrm{a}\mathrm{l}\mathrm{d}\mathrm{e}\mathrm{a}\mathrm{d}\mathrm{p}\mathrm{l}\mathrm{a}\mathrm{n}\mathrm{t}\mathrm{s})}{\left(\mathrm{t}\mathrm{o}\mathrm{t}\mathrm{a}\mathrm{l}\mathrm{s}\mathrm{o}\mathrm{w}\mathrm{n}\mathrm{p}\mathrm{l}\mathrm{a}\mathrm{n}\mathrm{t}\mathrm{s}\right)}}\end{array}}$$

Plant growth was evaluated as the mean relative growth rate in height of the above-ground system (RGR_h; Hunt, 1982) of individual seedlings, calculated as:

$${\displaystyle \begin{array}{l}{\displaystyle \mathrm{R}\mathrm{G}{\mathrm{R}}_{\mathrm{h}}(\mathrm{c}\mathrm{m}\mathrm{d}\mathrm{a}{\mathrm{y}}^{-1})=\frac{[\mathrm{l}\mathrm{n}\left(h_2\right)-\mathrm{l}\mathrm{n}\left(h_1\right)]}{\left(t_2/t_1\right)}}\end{array}}$$

where measurements of initial height (h₁) were taken 10 days after emergence (t1) and final height (h₂) at the time of harvest (t2) (i.e. after 20 and 40 days).

Immediately after the two harvests, the fresh above-ground system and the root system were weighed, and the root system was stored in distilled water for a few days at 4 °C until determination of the fresh root morphology by digital scanning. The root system of each seedling was suspended in 1 cm of water in a 20 cm × 25 cm clear acrylic tray and scanned at 600 dpi with a scanner (Epson V850 Pro Perfection). Images were analysed with WinRHIZO™ software (Regent Instruments Inc., Canada) to determine the root length and average root diameter (ARD). After the analysis, the samples (above-ground and root parts) were dried for 72 h at 65 °C and weighed. Specific root length (SRL; in metres per gram) was computed as the ratio of root length to dry matter (DM). Above-ground and root DM content [ADMC and RDMC, respectively (in milligrams per gram)] were computed as above-ground DM or root DM divided by above-ground fresh mass or root fresh mass. Above-ground DM and root DM were used to calculate the plant shoot-to-root ratio (SRR). The root mass ratio (RMR; as a percentage) was calculated as the ratio of whole root system dry mass to total plant dry mass, multiplied by 100 (Roumet et al., 2008).

Physicochemical soil measurements

Fresh soil samples (2.5 g equivalent of dry soil) were placed in plasma vials (150 mL). A 0.01 m of CaCl₂ solution (Houba et al., 2000) was added after consideration of the soil moisture (20 mL). Soil suspensions were sealed with M^® parafilm, incubated at 20 °C and shaken at 140 rpm for 2 h at 10 °C. The suspensions were filtered through a pore size of 0.22 μm. The NO₃⁻ concentration was quantified using a Smartchem 200 photometer (KPM Analytics, AMS Alliance, Villeneuve-la-Garenne, France).

Soil humidity was measured according to the ISO 11465 protocol. Two grams of fresh soil samples were placed in aluminium cups (previously weighed and annotated) and weighed. The samples were then placed in an oven at 105 °C for 24 h, weighed again, and moisture was calculated as the mass of water per 100 g of dry soil as follows:

$${\displaystyle \begin{array}{l}{\displaystyle \mathrm{S}\mathrm{o}\mathrm{i}\mathrm{l}\mathrm{h}\mathrm{u}\mathrm{m}\mathrm{i}\mathrm{d}\mathrm{i}\mathrm{t}\mathrm{y}(\mathrm{\%})=100\times \frac{\left(\mathrm{f}\mathrm{r}\mathrm{e}\mathrm{s}\mathrm{h}\mathrm{s}\mathrm{o}\mathrm{i}\mathrm{l}\mathrm{m}\mathrm{a}\mathrm{s}\mathrm{s}\right)-\left(\mathrm{d}\mathrm{r}\mathrm{y}\mathrm{s}\mathrm{o}\mathrm{i}\mathrm{l}\mathrm{m}\mathrm{a}\mathrm{s}\mathrm{s}\right)}{\left(\mathrm{d}\mathrm{r}\mathrm{y}\mathrm{s}\mathrm{o}\mathrm{i}\mathrm{l}\mathrm{m}\mathrm{a}\mathrm{s}\mathrm{s}\right)}}\end{array}}$$

Fresh and dry soil masses were expressed in grams.

Soil microbe measurements

Denitrification enzyme activity (DEA) was measured as N₂O production using a method adapted from Dassonville et al. (2011). Five grams (equivalent dry weight) of fresh soil were placed in a flask hermetically sealed with a rubber stopper. In each flask, air was removed and replaced with a He–C₂H₂ mixture (90:10, v/v) to create anoxic conditions and inhibit N₂O reductase. A nutritive solution containing glucose (0.5 mg of C-glucose per gram of dry soil), glutamic acid (0.5 mg of C-glutamic acid per gram of dried soil), potassium nitrate (50 µg of N-KNO₃ per gram of dry soil) and distilled water was added through the rubber stopper using a syringe to ensure non-limiting amounts of carbon and NO₃⁻ for denitrification activity. The amount of N₂O during incubation at 28 °C was measured each hour for 6 h. The slope of the linear regression was used to estimate anaerobic respiration (denitrification) as the N₂O produced (in micrograms of N-N₂O per gram of dry soil per hour). Gases (CO₂ and N₂O) were measured with a gas chromatograph coupled to a micro-catharometer detector (µGC-R990; SRA instruments, Marcy L’Etoile, France). To highlight the BDI strategy developed by F. × bohemica, the effect of the plant on DEA is used (see plant effect calculation in the subsection ‘Experimental design and harvests’). Biological inhibition of denitrification was observed when denitrification under the soil with F. × bohemica was significantly decreased compared with the soil lacking F. × bohemica (i.e. with a significant negative plant effect percentage).

Microbial DNA was extracted from each soil sample using the NucleoSpin^® Soil Kit (Macherey-Nagel GmbH & Co., Düren, Germany), following the manufacturer’s protocol. DNA concentrations were determined with the QuBit dsDNA BR Assay Kit (Invitrogen, Carlsbad, CA, USA) according to the manufacturer’s instructions. Extracted DNA was stored at −20 °C until use. The abundance of the nitrite reductase genes nirS and nirK, as proxies for the abundance of denitrifying bacteria, were quantified based on the number of gene copies present per gram of dry soil. Quantification was performed by quantitative real-time PCR, using iTaq™ Universal SYBR^® Green Supermix (Bio-Rad) as the fluorescent dye on a CFX Connect Real-Time PCR Detection System (Bio-Rad). The standard curves for nirK and nirS quantitative real-time PCR were generated by amplifying 10-fold dilutions of a linearized plasmid. Melting curve analysis confirmed the specificity of amplification, and gene amplification efficiencies were in the range 85–99 %. The R² values were always >0.95. Possible inhibition of PCR was tested in advance, and appropriate dilutions were chosen.

Statistical analyses

The data were subjected to JUMP PRO v.17.0 (SAS Institute Inc., Cary, NC, USA) and were analysed using two-way ANOVA (harvested dates, N enrichment treatments and the interaction), t-test and Pearson correlation analysis, as appropriate. The normality of residuals was tested by Shapiro’s test and homogeneity of variance by Levene’s test. When necessary, data for microbial activities were transformed. Significance levels and other data formats are as illustrated in the figures and table legends.

RESULTS

Effects of plant presence on denitrification and the microbial denitrifying community

In the treatment with no addition of nitrogen (termed ‘no add’ hereafter), we observed significant effects of plant presence on denitrification (DEA; Fig. 2) depending on the plant growth time, with a negative effect of plant presence on DEA after 20 days of plant growth (T20: −9.16 % on average) and a positive effect after 40 days (T40: 15 % on average). For the microbial denitrifying community in the no add treatment, we observed contrasting effects of plant presence depending on the microbial population considered (i.e. nirK or nirS gene abundance). A negative effect of plant presence was found at each growth time for nirK abundance (−16.77 % for T20 and −22.9 % for T40; Fig. 3A) and for nirS abundance only at T40 (no effect at T20 and an average of −20.57 % at T40; Fig. 3B).

Fig. 2.

Effects of Fallopia × bohemica on denitrifying enzyme activity (DEA) after 20 days (T20, white circles) and 40 days (T40, grey circles) of growth after emergence, as a function of nitrogen enrichments (Add-N1 < Add-N2 < Add-N3 < Add-N4 < Add-N5), expressed as a percentage change from unplanted soil. Means and s.e. are presented per treatment and date (n = 4). Negative values indicate a negative effect of plant presence on DEA (plant microbial activity values are lower than unplanted soil values), and vice versa for positive values. Multiple one-sample t-tests were conducted to discover differences between plants and unplanted soil (μ = 0), and the level of significance is indicated as follows: *P < 0.05, **P < 0.01 and ***P < 0.001.

Fig. 3.

Effects of Fallopia × bohemica on nirK (A) and nirS (B) gene abundances after 20 days (T20, white circles) and 40 days (T40, grey circles) of growth after emergence, as a function of nitrogen enrichments (Add-N1 < Add-N2 < Add-N3 < Add-N4 < Add-N5), expressed as a percentage change from unplanted soil. Means and s.e. are presented per treatment and date (n = 4). Negative values indicate a negative effect of plant presence on denitrifying gene abundances (gene abundance values under plants are lower than unplanted soil values), and vice versa for positive values. Multiple one-sample t-tests were conducted to discover differences between plants and unplanted soil (μ = 0), and the level of significance is indicated as follows: *P < 0.05, **P < 0.01 and ***P < 0.001.

By comparing unplanted and planted pots, we observed that plant presence had contrasting effects on DEA depending on the N enrichment treatment and plant growth time (significant N enrichment treatment and time interaction; F_1,5 = 6.62, P < 0.001; Fig. 2). Throughout all N enrichments, the plant effect on DEA increased with plant growth time, with an average plant effect on DEA of −9.98 % at T20 and +6.36 % at T40 (F_1,1 = 16.81, P < 0.001). For T20, a negative effect of plant presence on DEA was found for the three strongest N enrichments (average of −19.47, −9.92 and −18.75 % for Add-N3, Add-N4 and Add-N5, respectively). This negative effect was negated by the addition of the first and second enrichments of N (average −2 % for Add-N1 and −0.47 % for Add-N2, with no difference between unplanted and planted soil; Fig. 2). For T40, the stimulation of DEA associated with the plant presence was reduced 2-fold when the soil was amended to the strongest N addition (average −10 % in Add-N5; Fig. 2) and increased with the third N addition (average of +28 % in Add-N3). At T40, Add-N1, Add-N2 and Add-N3 showed no difference on DEA between planted and unplanted soil.

For the effect of plant presence on nirK abundance, we observed significant changes only with the N enrichment treatments (F_1,5 = 12.49, P < 0.0001). The plant effect on nirK abundance was reduced in Add-N4 at T20 (−27.32 %) and increased in Add-N5 (48.55 and 21.7 % for T20 and T40, respectively; Fig. 3A). As for the effect of plant presence on nirK abundance, we did not observe significant differences between plant growth times for the plant presence effect on nirS abundance. We observed that plant presence had contrasting effects on nirS abundance depending on the N enrichment treatment and plant growth time (significant N enrichment treatment and time interaction; F_1,5 = 13.41, P < 0.0001; Fig. 3B). For T20, a negative effect of plant presence on nirS abundance was found for the Add-N4 and Add-N5 treatments (average of −18.07 and −43.87 %, respectively) and a positive effect for the Add-N2 treatment (17.54 %; Fig. 3B). For T40, a negative effect of plant presence on nirS abundance was found for the Add-N2 treatments (average of −24.83 %) and a positive effect for the Add-N5 treatment (18.05 %; Fig. 3B).

Plants also influenced the relationships between DEA and microbial communities, such as nirS abundance or the concentration of nitrate in soil (Fig. 4). When considering all N treatments together, DEA was positively correlated with the effect of plant presence on nirS abundance after 20 days of plant growth but not after 40 days (R² = 0.39, F_1,22 = 14.03, P = 0.0011; Fig. 4A). For T20 and T40, the effect of plant presence on DEA was negatively correlated with the effect of plant presence on soil nitrate concentration (R² = 0.48, F_1,22 = 20.05, P = 0.0002 for T20; and R² = 0.23, F_1,22 = 6.79, P = 0.016 for T40; Fig. 4B).

Fig. 4.

Correlation between the effects of Fallopia × bohemica compared with unplanted soil after 20 days (T20, white) and 40 days (T40, grey) of growth after emergence for all N treatment values. (A) Plant effect on denitrification (DEA) vs. plant effect on nirS gene abundance. (B) Plant effect on DEA vs. plant effect on concentration of soil nitrate. When a correlation is significant, the R² and P-value are indicated; otherwise, NS (non-significant) is displayed.

In conclusion, a reduction of DEA was observed after T20 and a stimulation after T40 in the no add treatment. The increase of soil N content with an ammonium nitrate gradient showed few and contrasting plant effects on the activity and abundance of the soil denitrifying community compared with the no add treatment.

Comparison between plant growth and N enrichment treatments for above-ground traits and root traits

Total survival of F. × bohemica seeds was variable depending on the N enrichment treatment, with 76.9 % for no add, 100 % for Add-N1 and Add-N2, 92.3 % for Add-N3 and Add-N4, and 84.6 % for Add-N5 (data not shown).

For above-ground traits, we observed significant changes with plant growth time (F_1,1 = 121.6, P < 0.0001 for above-ground biomass; Supplementary Data Fig. S1A), with an increase of 228.7 % for above-ground biomass at T40 compared with T20. No effect of N enrichment treatment was found. No effect of plant growth time and N enrichment were found on RGR_h (F_1,11 = 1.31, P = 0.26; Supplementary Data Fig. S2). For ADMC, we observed changes depending on the N enrichment treatment and plant growth time (significant N enrichment treatment and time interaction; F_1,5 = 3.40, P = 0.0128; Fig. 5A). Plant growth time showed a significant positive effect on ADMC, with increased ADMC at T40 compared with T20 for each N enrichment treatment (F_1,1 = 196.3, P < 0.0001) except for Add-N1, which shown no difference between T20 and T40 (Fig. 5A).

Fig. 5.

Effects of nitrogen enrichment treatments and time on plant dry matter content [A, above-ground (ADMC) and B, root (RDMC)] after 20 days (T20, white) and 40 days (T40, grey) of growth after emergence. Means and s.e. are presented (n = 4). Different letters indicate significant differences between the interaction of nitrogen enrichment treatments and time (P < 0.05) based on Tukey’s HSD tests.

For root traits, time had a significant effect on root biomass (F_1,1 = 133.9, P < 0.0001), with an increase of 405.6 % on average after 40 days of growth compared with 20 days (Supplementary Data Fig. S1B). The same results were found for the SRR, with a decrease between T20 and T40 (F_1,1 = 44.33, P < 0.0001; Supplementary Data Fig. S1C). For RMR, the same results were found, with no effect of N treatment and an increase of 10 % on average of the biomass allocation towards root biomass than above-ground biomass after 40 days (F_1,1 = 62.15, P < 0.0001; Fig. 6).

Fig. 6.

Effects of nitrogen enrichment treatments and time on root mass ratio (RMR; as a percentage) after 20 days (T20, white) and 40 days (T40, grey) of growth after germination. Means and s.e. are presented (n = 4). Different letters indicate significant differences between the interaction of nitrogen enrichment treatments and time (P < 0.05) based on Tukey’s HSD tests.

Root dry matter content had contrasting responses depending on the N enrichment treatment and plant growth time (significant N enrichment treatment and time interaction; F_1,5 = 10.47, P < 0.0001; Fig. 4B). An increase of RDMC was found between T20 and T40 for no add, Add-N3, Add-N4 and Add-N5. At T40, Add-N3 and Add-N4 had significantly higher values of RDMC compared with Add-N1 and Add-N2 (Fig. 4B). No N enrichment or plant growth effects was found in SRL (Supplementary Data Fig. S3). The ARD was significantly higher at T20 than at T40 (F_1,1 = 20.93, P < 0.0001; Supplementary Fig. S3). The gain in RDMC between T20 and T40 [measured as RDMC gain (in milligrams per gram) = values of RDMC at T40 minus values of RDMC at T20] was significantly higher in treatments where plants developed a BDI strategy at T20 (χ² = 11.76, P = 0.0006; Fig. 7). The BDI treatments were grouped according to the results shown in Fig. 2 at T20, with BDI in the no add, Add-N3, Add-N4 and Add-N5 treatments and no BDI in the Add-N1 and Add-N2 treatments.

Fig. 7.

Comparison between biological inhibition of denitrification (BDI) and no BDI for the root dry matter content gain between 20 and 40 days of growth. The BDI and no BDI treatments are defined by Fig. 2 for T20. The BDI treatment is defined as treatments with significant decreases in DEA (no add, Add-N3, Add-N4 and Add-N5 treatments), and the no BDI treatments include the Add-N1 and Add-N2 treatments. All values are presented according to their N treatment (no add, Add-N1, Add-N2, Add-N3, Add-N4 and Add-N5); each symbol is explained in the key. Means and s.e. are presented (dark dots with error bars on the BDI treatment, n = 16; and the no BDI treatment, n = 8).

The key results are that the plant traits ADMC, RDMC, RMR, SRL, ARD and RGR_h of F. × bohemica showed no change when comparing the N-amended soils with the soil without N addition after 20 and 40 days of plant growth. The root biomass was found to be higher than above-ground biomass at T40 (see RMR and SRR).

Correlations between plant traits and the denitrifying microbial community after 20 and 40 days of plant growth

Regardless of the plant growth time, correlations in plant traits were found for SRL, which was negatively correlated with ARD and with RDMC (for SRL/RDMC: r = −0.56 and P < 0.01 for both T20 and T40; and for SRL/ARD: r = −0.66 and P < 0.001 for T20 and r = −0.47 and P < 0.05 for T40; Table 3A).

Table 3.

Principal pairwise comparisons of Pearson’s correlation among plant traits, microbial traits and soil parameters after 20 days (T20) and 40 days (T40) of growth of Fallopia × bohemica. Abbreviations: ADMC, above-ground dry matter content; ARD, average root diameter; NO₃⁻, nitrate; RDMC, root dry matter content; RGR_h, relative growth rate in length.

	T20		T40
(A) Plant trait correlations		r		r
	RDMC–SRL	−0.56**	RDMC–SRL	−0.56**
	ARD–SRL	−0.66***	ARD–SRL	−0.47*
	ADMC–SRL	−0.09	ADMC–SRL	−0.66***
(B) Plant and microbial trait correlations
	ADMC–%DEA	−0.42*	ADMC–%DEA	0.28
	ADMC–nirS	−0.49*	ARD–nirS	−0.48*
	RDMC–nirK	−0.41*	RDMC–nirK	0.17
(C) Plant trait and soil parameter correlations
	RGRl–NO3−	−0.40*	RGRh–NO3−	−0.01
	ARD–Humidity	−0.41*	ADMC–Humidity	−0.43*

Pearson correlation coefficient (r). Significance levels: *P < 0.05, **P < 0.01 and ***P < 0.001.

For T20, another correlation between plants traits was found, with positive links between SRR and SRL (Table 3A). Moreover, correlations between plant traits and microbial traits were found, with negative relationships between ADMC and two microbial traits (%DEA: r = −0.42, P < 0.05; and nirS abundance: r = −0.49, P < 0.05) and between RDMC and nirK abundance (r = −0.49, P < 0.05; Table 3B). Plant traits were also correlated with some soil parameters, such as RGR_h with soil nitrate content and ARD with soil humidity (r = −0.40 and P < 0.05 and r = −0.41 and P < 0.05, respectively; Table 3C).

For T40, another correlation between plant traits was found between ADMC and SRL (r = −0.66, P < 0.001; Table 3A). There was a correlation between plant and microbial traits, with ARD being negatively correlated with nirS abundances (r = −0.48, P < 0.05; Table 3B), and a correlation between plant traits and soil parameters, with ADMC being negatively correlated with soil humidity (r = −0.43, P < 0.05; Table 3C).

In conclusion, negative correlations were found between RDMC and SRL and between SRL and ARD, as commonly observed in plant ecology. Other correlations were found between plant traits and microbial parameters, mainly at the younger stage (T20).

DISCUSSION

Recently, Bardon and co-workers identified a novel plant strategy to improve soil NO₃⁻ acquisition by plants (Bardon et al., 2014, Bardon et al., 2018). This strategy, called BDI (Bardon et al., 2014), was able to inhibit bacterial denitrification by exudation of procyanidin by the roots (Bardon et al., 2016a, b). Subsequent studies demonstrated the potential for the Fallopia spp. complex to inhibit denitrification over a significant range of environmental conditions (i.e. in field conditions: Dassonville et al., 2011; or in mesocosm conditions: Cantarel et al., 2020), but only in plants regenerated from vegetative propagules (rhizome fragments). For the first time, we focused on: (1) the establishment of BDI at early growth stages (20 and 40 days after emergence); (2) the impacts of soil nitrogen conditions (i.e. by N enrichments); and (3) the influence on plant growth.

For soil without added nitrogen, a reduction of soil DEA was found at T20 in soil containing F. × bohemica compared with unplanted soil, whereas a stimulation of the DEA was found at T40 (Fig. 2). Many studies have reported that the denitrifying activity and the structure of soil heterotrophic denitrifying communities can be affected by the presence of plants (Philippot et al., 2007). Generally, studies have found a stimulatory effect of plants on DEA (Henry et al., 2008; Guyonnet et al., 2017) induced by root-released carbon exudates, called the rhizosphere effect (Bais et al., 2006). Exudation is a cost for the plant, and young seedlings exude ~30–40 % of the carbon fixed by photosynthesis (Whipps, 1990). Root exudation is influenced by the root morphology, with greater exudation by the fine roots (Meier et al., 2020). In our study, the reduction in DEA found after 20 days of F. × bohemica growth suggests an early implementation of BDI that is later counteracted, after 40 days of growth, by the potential increase of a rhizosphere effect.

In the case of inhibition of the N cycle by plants (i.e. biological inhibition of nitrification), production of the nitrification inhibitor was reported for Sorghum varieties at early seedling growth stages (i.e. 7 days; Czarnota et al., 2003), suggesting that inhibitory plant metabolites could to be produced early during plant growth. Using the p-dimethylaminocinnamaldehyde method (an analytical colorimetric method that quantifies specifically the total content of soluble proanthocyanidins; Prior et al., 2010), the 20- and 40-day-old F. × bohemica roots revealed a strong blue coloration, synonymous with the presence of significant amounts of proanthocyanidins (red arrows in Fig. 1B). These findings suggest that seedlings produce and exudate this complex molecule even after 40 days of growth, suggesting that BDI persists after 20 days. In the early life stage of F. × bohemica, the BDI strategy appears to be highly dynamic, with early establishment after the 12th stage of the BBCH scale, but also depending on implemented rate of photosynthesis and the linked root exudation.

The nirS-type denitrifier community known to be more closely associated with plants was found to be more impacted by the rhizosphere effect than the nirK-type community (Guo et al., 2013; Kastl et al., 2015; Hou et al., 2018). In our study, nirK and nirS abundances were generally reduced in the presence of F. × bohemica in the no add treatment for all dates (Fig. 3). Only for the nirS-type community at T20 was no change found between planted and unplanted soil (Fig. 2B). The implementation of BDI highlighted at T20 by the reduction of DEA seems to be linked more to the nirK-type community than to the nirS-type community in our sandy clay loam soil. Galland et al. (2019, 2020) have previously found that procyanidins (condensed tannin exuded by roots permitting BDI) have similar effects on nirK-type and nirS-type communities on sandy clay loam or loamy sand soils. Our results suggest more complex interactions between denitrifying communities and root exudation (Philippot et al., 2007) than addition of procyanidins alone. The reduction of nirK and nirS abundances at T40 highlights a potential persistence of BDI after 40 days of plant growth, perhaps linked to a dampening of BDI compared with T20 and/or concealment of BDI by an increase of root-released carbon exudates (i.e. rhizosphere effect). Several studies have indicated that the nirK-type community colonizes habitats with high contents of easily degradable carbon (Ollivier et al., 2011; Kastl et al., 2015). This could be taken as an indication of changes in root exudation by F. × bohemica during its seedling growth, with simple carbohydrates being exuded at T20 and more complex ones at T40.

Generally, enrichment of soil with nitrate-based fertilizers is known to promote microbial denitrification, with an increase in the abundance of denitrifier communities, which is especially true for the nirK-type community (Hallin et al., 2009; Ye et al., 2022). However, in our study, increasing the soil N content with an ammonium nitrate gradient (Table 2) caused few and contrasting effects of plants on the activity and abundance of the soil denitrifying community compared with the no add treatment. The contrasting effects observed between N enrichments and the abundance of denitrification genes (nirK and nirS) suggest complex interactions between the N dose added and the denitrifying communities. For instance, the N level in soil could impact different microbial populations harbouring the nirK or nirS gene, leading to these effects (Yang et al. 2017). After 20 days of plant growth, BDI was found only for the highest N enrichments (Add-N3, Add-N4 and Add-N5; Fig. 2). Cantarel et al. (2020) have shown similar findings, with the strongest BDI on a nitrate-rich soil compared with a nitrate-poor soil. In contrast to the no add treatment, DEA stimulation was not found after 40 days of plant growth, except for the Add-N3 treatment. In our study, carbon exudation of F. × bohemica roots would not stimulate the denitrifying microbial community sufficiently to see a rhizosphere effect. These results suggest that this rhizosphere effect is countered either by an active BDI at T40 or by the persistence of the procyanidin effects of the BDI developed at T20, leading to an effect of plants on DEA close to zero (Fig. 2). Regarding inhibition of nitrification, the effects of plant inhibitory metabolites are known to be stable in soils, often lasting ≤30 days (Tesfamariam et al., 2014), and this might also be the case for effects of procyanidins. Recent results have shown that the effects of procyanidins on denitrifying activity can persist for 50 days with the same intensity and decrease only after 76 days (Galland et al., 2020). Moreover, negative correlations were found between the effect of plants on DEA and soil nitrate concentration after 20 and 40 days of plant growth (Fig. 4B), with an increase of soil nitrate content when plants developed a BDI strategy reducing DEA. Given that the negative correlation is also significant after 40 days of plant growth, this result implies a persistence of the BDI strategy implemented at T20, concealed by a rhizosphere effect at T40.

For denitrifying communities, N enrichments and plant effects seem difficult to disentangle. The N enrichment has different effects on nirS-type and nirK-type communities depending on the N added and the plant growth (Fig. 3). Application of procyanidins is known to reduce the abundance of nirS-type and nirK-type communities (Galland et al., 2019; Ye et al., 2022), and their inhibitory efficacy seems greater with ammonium nitrate-type fertilizers (Ye et al., 2022). Our results reflect not only the effects of procyanidins, but also all the mechanisms by which plants shape rhizosphere denitrifying communities. The concentrations of key factors regulating denitrifying communities in the rhizosphere, such as labile carbon, nitrate and oxygen, are drastically different among plant growth stages (Philippot et al., 2007). However, these changes in the structure of microbial communities effected by plants remain unclear. Interestingly, for Add-N5 treatment at T20, when BDI was strong (Fig. 2), the plants also seemed to inhibit the nirS-type community (Fig. 3A), consistent with the responsiveness of the nirS-type community to the rhizosphere effect found in previous studies (Guo et al., 2013; Kastl et al., 2015; Hou et al., 2018).

Soil N is of particular importance in plant growth, development and nutrition because it regulates biochemical and physiological functions of plants, such as the rate of photosynthesis (Vitousek and Farrington, 1997). Surprisingly, the above-ground and root dry biomasses, RMR and the RGR_h of F. × bohemica showed no changes when comparing the N-enriched soils with the soil without added N after 20 and 40 days of plant growth. Usually, studies conclude that plants, especially plants with high nutrient uptake and high growth, such as knotweeds, are favoured by increased N availability and respond rapidly to this (Lowe et al., 2003). When plants are exposed to increased N availability, they benefit from accelerated nutrient uptake (James and Richards, 2006). Here, our plant growth results seem to reflect no changes in the N uptake status of F. × bohemica seedlings, suggesting a unique ability of F. × bohemica to grow regardless of the soil N status. We found that the ADMC and RDMC at T20 were similar regardless of N levels (Fig. 5). Dry matter content is a key functional trait commonly associated with plant growth strategies (Westoby, 1998). Higher values indicate higher lignin content and a higher proportion of xylem for roots and stems (Hummel et al., 2007) and are often associated with a resource conservation strategy in response to resource limitation (Roumet et al., 2008). We can assume that there are no N limitations at this stage of seedling development and that these changes reflect a change in plant storage status. After 40 days of growth, we found that ADMC and RDMC decreased with the lowest addition of nitrogen (Add-N1; Fig. 4) compared to the no add treatment and the other N enrichments (Add-N2 and Add-N5 for ADMC; and Add-N3 and Add-N4 for RDMC). Our results on the enrichment treatments do not agree, because the highest values of RDMC and ADMC are similar for the lowest N level (no add) and the highest (Add-N5), suggesting that the dry matter content does not vary according to N resource limitation. The Fallopia × bohemica seedlings have an exceptional ability to grow regardless soil N status. However, ADMC and RDMC values generally increased between T20 and T40 (except for Add-N1 for ADMC and Add-N1 and Add-N2 for RDMC). This could mean that seedlings respond at T40 with a shift in resource use strategy towards increased resource conservation in their organs. Fallopia × bohemica is known for its high resource storage capacity in the below-ground system (Price et al., 2001), with establishment of a rhizome after 5 months of growth on average. It is possible that this below-ground storage starts as early as 40 days (Fig. 6). Monitoring the changes in resource allocation in plant organs with total carbon and nitrogen measurements would be particularly interesting to confirm our hypothesis.

The plant economics spectrum postulates that above-ground traits and root traits determine plant trait syndromes for resource acquisition and/or conservation (Wright et al., 2004; Reich, 2014). Few studies have focused on plant seedlings when analysing these trait syndromes (Kramer-Walter et al., 2016; Shen et al., 2022). We examined the relationships between above-ground traits and root traits of F. × bohemica seedlings, N-associated microbial traits and soil parameters. Two known and consistent trait syndromes were found for both harvest dates of the F. × bohemica seedlings (Table 3). As in previous studies on fine roots (Prieto et al., 2015; Roumet et al., 2016; Valverde-Barrantes et al., 2017), we report a negative correlation between RDMC and SRL, and between SRL and ARD. These results confirm the trait syndrome whereby plants (here, F. × bohemica) invest either in long and fine roots (high SRL) associated with higher nutrient uptake, exudation and greater resource acquisition, or in thick and dense roots associated with longer root life spans and greater resource conservation (Eissenstat et al., 2000). In an experiment with F. × bohemica growing with procyanidins, Bardon et al. (2017) observed a significant increase in lateral root proliferation, associated with an increase in SRL and a decrease in RDMC, known to explain an increase in lateral root meristematic activity (Malamy, 2005). These results might be related to the BDI activity observed in our study at T20 in F. × bohemica seedlings. Few relationships were found between seedling growth (RGR_h or biomass) and root traits. These results are coherent with other studies. Shen et al. (2022) found that the variation in seedling growth could not be captured completely by root traits. Recent studies have explained this lack of correlation, which could be attributable to the more multidimensional aspect of root traits developed in a more complex environment (coupling soil, water, nutrient and microbial parameters), hence a single acquisition–conservation axis cannot adequately capture all trait syndromes and the correlation with plant performance traits (Kramer-Walter et al., 2016; Shen et al., 2022). In our study, we found only a negative correlation between SRL and ADMC at T40. This result could be explained by the shift found between T20 and T40 in plant resource allocation in favour of root biomass (Fig. 6, for RMR; Supplementary Data Fig. S1, for SRR), suggesting an increase of root foraging capacity by an increase of fine roots (high SRL) concomitantly with a reduction in resource allocation to the above-ground part (low ADMC).

Interestingly, relationships between plants and microbial parameters were found preferentially for the young F. × bohemica stage (i.e. at T20; Table 3). There was a negative correlation of ADMC with the plant effect on DEA, suggesting either that BDI is implemented (i.e. for the lowest plant effect values) and the resource storage in above-ground part increases (i.e. the highest values of ADMC) or that the seedlings with higher storage capability have enough energy to implement BDI. This correlation is supported by the negative correlation between RGR_h and nitrate soil level (Table 3). If BDI is implemented at T20, this represents a cost for seedlings, reducing their ability to grow, but at the same time BDI increases the available soil nitrate (as already shown by Dassonville et al., 2011; Bardon et al., 2014). In line with the establishment of BDI, the dry matter contents, ADMC and RDMC, were negatively correlated with nirS-type or nirK-type communities, respectively. These correlations observed between plant and microbial traits at T20 seem to reflect early and strong chemical interactions between the plants and the denitrifying communities, which decreased with plant growth and the increase of the rhizosphere effect.

Conclusion

The BDI strategy seems to be a highly dynamic plant process that is specific to the soil conditions (i.e. N level, microbial communities, etc.). All these results advocate for: (1) an early implementation of BDI, particularly after 20 days of growth for the no-add and the highest N treatments; and (2) a unique plant growth capability of F. × bohemica, regardless of soil N levels. This early BDI leads to efficient acquisition of N resources, with a latent period between the establishment of BDI (T20) and effective storage in plant organs (T40). Moreover, F. × bohemica appeared to change their allocation strategy to a root conservative strategy after 40 days, involving resource storage (higher ADMC and RDMC). We found an alteration of allocation to above-ground and below-ground parts (shift to root allocation apparent in RMR) and investment in fine roots (decrease of the ARD), regardless of the N treatment. In the future, we must study the evolution of this gain in RDMC, linked to the early BDI implementation, for plant performance. Monitoring root morphology in the first life stages of F. × bohemica could confirm this root storage hypothesis with the development of particular root morphology (i.e. a strong and coarse root system) or rhizome establishment. Fallopia × bohemica is a rhizomatous species, and the development of its seedlings must involve, after formation of the root system, the formation of the rhizome (generally after 5 months of growth). It now seems clear that this storage strategy starts very young with early establishment of BDI, allowing this clonal plant exceptional storage, multiplication and competitive capacities. Given that Fallopia spp. genotypes are known to be highly variable, it would be interesting to validate these conclusions on a large panel of Fallopia × bohemica genotypes.

SUPPLEMENTARY DATA

Supplementary data are available at Annals of Botany online and consist of the following.

Figure 1: effects of nitrogen enrichment treatments and time on plant dry biomasses (A, shoot biomass; and B, root biomass) and on the shoot-to-root ratio (SRR; C) after 20 days (T20, white) and 40 days (T40, grey) of growth after emergence. Means and s.e. are presented (n = 4). Different letters indicate significant differences (P < 0.05) based on Tukey’s HSD tests.

Figure 2: effects of nitrogen enrichment treatments and time on relative growth rate in height (RGR_h) of individual seedlings after 20 days (T20, white) and 40 days (T40, grey) of growth after emergence. Means and s.e. are presented (n = 4).

Figure 3: effects of nitrogen enrichment treatments and time on average root diameter (ARD; A) and specific root length (SRL; B) after 20 and 40 days of growth after emergence. Means and s.e. are presented (n = 4). Different letters indicate significant differences (P < 0.05) based on Tukey’s HSD tests.

FUNDING

This work was supported by the French National Programme CNRS EC2CO-MICROBIOME 2022-2023 with the MiRe-A project.

AUTHOR CONTRIBUTIONS

A. A. M. Cantarel, M. Simonin and F. Piola contributed to the study conception, design and funding acquisition. Material preparation, experiment monitoring and data collection were performed by A. Signoret, J. Gervaix, P. Defour, A. Delort, C. Creuzé des Châtelliers, C. Beligon, C. Béraud, C. Boisselet, C. Marais, C. Lobreau, E. Louvez and E. Lacroix. Data analyses were performed by A. Signoret, A. A. M. Cantarel, M. Simonin and F. Piola. The first draft of the manuscript was written by A. A. M. Cantarel and F. Piola, and all authors commented on previous versions of the manuscript. All authors read and approved the final version of the manuscript.

Data Availability

All data that support the findings of this study are available from the corresponding author, Amélie A. M. Cantarel, upon reasonable request.