Free-air CO₂ enrichment (FACE) reduces the inhibitory effect of soil nitrate on N₂ fixation of Pisum sativum

Abstract

Background and Aims Additional carbohydrate supply resulting from enhanced photosynthesis under predicted future elevated CO₂ is likely to increase symbiotic nitrogen (N) fixation in legumes. This study examined the interactive effects of atmospheric CO₂ and nitrate (NO₃^–) concentration on the growth, nodulation and N fixation of field pea (Pisum sativum) in a semi-arid cropping system.

Methods Field pea was grown for 15 weeks in a Vertosol containing 5, 25, 50 or 90 mg NO₃^–-N kg^–1 under either ambient CO₂ (aCO₂; 390 ppm) or elevated CO₂ (eCO₂; 550 ppm) using free-air CO₂ enrichment (SoilFACE).

Key Results Under aCO₂, field pea biomass was significantly lower at 5 mg NO₃^–-N kg^–1 than at 90 mg NO₃^–-N kg^–1 soil. However, increasing the soil N level significantly reduced nodulation of lateral roots but not the primary root, and nodules were significantly smaller, with 85 % less nodule mass in the 90 NO₃^–-N kg^–1 than in the 5 mg NO₃^–-N kg^–1 treatment, highlighting the inhibitory effects of NO₃^–. Field pea grown under eCO₂ had greater biomass (approx. 30 %) than those grown under aCO₂, and was not affected by N level. Overall, the inhibitory effects of NO₃^– on nodulation and nodule mass appeared to be reduced under eCO₂ compared with aCO₂, although the effects of CO₂ on root growth were not significant.

Conclusions Elevated CO₂ alleviated the inhibitory effect of soil NO₃^– on nodulation and N₂ fixation and is likely to lead to greater total N content of field pea growing under future elevated CO₂ environments.

INTRODUCTION

Elevated atmospheric CO₂ concentrations under future climate scenarios (IPCC, 2001) are predicted to produce major changes in the productivity and sustainability of agricultural cropping systems. It has been widely demonstrated that elevated CO₂ stimulates photosynthesis, reduces evapotranspiration and subsequently increases plant productivity when water and nutrients are not limiting (Ainsworth and Long, 2005; de Graaff et al., 2006; Luo et al., 2006; Leakey et al., 2009). The relative responses of different systems are highly dependent on plant species and soil properties.

Legumes generally produce greater responses to elevated CO₂ than non-legumes (Rogers et al., 2009). Additional carbohydrate supply under eCO₂ resulting from enhanced photosynthesis is likely to increase symbiotic N₂ fixation and fulfil any additional nitrogen (N) requirement (Cabrerizo et al., 2001; West et al., 2005). Field pea (Pisum sativum) is an economically important crop in many semi-arid cropping systems and is both mycorrhizal and a legume, and therefore may have a greater ability to avoid photosynthetic acclimation by diverting carbon (C) to its symbioses (Aranjuelo et al., 2013). The increase in N₂ fixation under eCO₂ could be derived from an increase in the number, size and/or activity of root nodules (Rogers et al., 2009). Lam et al. (2012) reported no change in number or mass of field pea root nodules under elevated CO₂ despite 44–51 % increases in N₂ fixation. In contrast, field pea grown under elevated CO₂ had greater nodule biomass but not a greater number of root nodules (Cabrerizo et al., 2001). However, few studies have examined the effects of elevated CO₂ on field pea.

The response of legumes to elevated CO₂ in Australian agricultural systems could be less predictable than other systems since they are frequently water limited and soil N levels are highly variable. Preceding cereal crops generally have poor N fertilizer use efficiency (Chen et al., 2008), and consequently high soil N levels during legume establishment could limit responses to elevated CO₂. Small amounts of soil N at the start of the growing season can enhance legume production, possibly due to bridging the interval between seed N utilization and N₂ fixation and the lower energy requirement of nitrate (NO₃^–) uptake and assimilation than of N₂ fixation (Jensen, 1986; Streeter and Wong, 1988; Peoples et al., 2012). Alternatively, plants supplied with small amounts of N may establish greater leaf area and photosynthesis during early growth stages that provides a greater capacity for subsequent nodulation and N₂ fixation, although this is not always the case (Peoples et al., 1995).

High levels of soil N, predominantly NO₃^– from fertilizers, are known to inhibit the initiation, development and function of legume symbioses. Nodule number is depressed in the presence of NO₃^–, possibly due to a reduction in root hair growth, signalling molecules (flavonoids) and subsequent expression of nodulation genes in Rhizobium that disrupt the infection process (Waterer and Vessey, 1993; Bollman and Vessey, 2006). Nodule formation and hence nodule number is often reported to be less sensitive to NO₃^– than nodule growth and activity (Streeter and Wong, 1988), although this is not always the case (Waterer and Vessey, 1993). For established nodules, NO₃^– is thought to retard nodule growth and activity by causing preferential diversion of photosynthate to NO₃^– assimilation, by reducing oxygen transport within nodules or via toxic effects once reduced to nitrite (NO₂^–), although these are highly debated (Streeter and Wong, 1988; Arrese-Igor et al., 1997). Moreover, the relative effect of NO₃^– on crop legumes differs widely between species and between cultivars (Cowie et al., 1990; Peoples et al., 2012).

The combined effects of elevated CO₂ and soil N level on the development and function of field pea root nodules is not clear. Previous studies on the mechanisms of NO₃^– inhibition of legumes under artificial and controlled conditions are not likely to reflect the response of plants grown in soil under field conditions. Naudin at al. (2010) showed that N₂ fixation of field pea intercropped with wheat could recover from NO₃^– inhibition at all growth stages. This could be due to temporal reduction of bulk soil N or to spatial depletion of rhizosphere NO₃^– that allows nodulation and N₂ fixation to proceed. Furthermore, greater C flow to the roots under elevated CO₂ could alter the response of field pea to the inhibitory effects of soil NO₃^–. This experiment aimed to quantify the interactive effects of soil NO₃^– concentration and CO₂ concentration on nodulation and N₂ fixation of field pea under field [free-air CO₂ enrichment (FACE)] conditions. We hypothesized that nodule establishment (nodule number), development (nodule mass) and function (nitrogenase activity, N derived from the atmosphere) would be progressively inhibited with increasing NO₃^– concentration, but these effects would be reduced under elevated CO₂ via enhanced N demand due to greater photosynthetic activity and plant biomass accumulation.

MATERIALS AND METHODS

Field site and sampling details

This study was carried out at the SoilFACE facility at the Department of Economic Development, Jobs, Transport and Resources (DEDJTR) Plant Breeding Centre, Horsham (36 °44'57''S, 142 °06'50''E), Victoria, Australia. The SoilFACE system is part of the Australian Grains FACE (AgFACE) facility in the Victorian Wimmera region (Mollah et al., 2009). A detailed description of SoilFACE is given in Butterly et al. (2015). Soil was collected in June 2011 by excavating small areas of a nearby roadside (0–15 cm) following the removal of plant material and debris from the soil surface. Soil was sieved (<4 mm), thoroughly mixed and air-dried. This virgin soil was classified as a Vertosol (Isbell, 1996) or Vertisol (FAO/ISRIC/ISSS, 1998), respectively, and had the following physiochemical properties: total C, 14 mg kg^–1; total N, 0·9 mg kg^–1; organic C, 9·4 mg kg^–1; Colwell P, 6 mg kg^–1; NH₄-N, 3 mg kg^–1; NO₃-N, 5 mg kg^–1; pH (1:5 in 0·01 m CaCl₂), 7·8; and clay, 50 % (Butterly et al., 2015).

Column experiment design and establishment

A column experiment was conducted using longitudinally sectioned PVC columns (10 cm ID × 60 cm high rejoined with silicon and PVC tape and fitted with PVC end caps at the base). Vertosol was mixed 1:1 (w/w) with triple-washed white sand (mean diameter 430 μm) to facilitate root sampling. Each column was filled with 5·5 kg of soil–sand mix containing basal nutrients (mg kg^–1) (KH₂PO₄, 70·4; K₂SO₄, 103; CaCl₂, 186; MgSO₄, 122; MnSO₄, 6; ZnSO₄, 8; CuSO₄, 6; CoCl₂, 0·4; FeCl₃, 0·6; Na₂B₄O₇, 1·6; and Na₂MoO₄, 0·4) at a bulk density of 1·2 g cm^–3. Columns also contained four NO₃-N levels; 5, 25, 50 and 90 mg N kg^–1, with 20, 20, 4 and 4 %, respectively, of the NO₃^– added as Ca(¹⁵NO₃)₂ (20 % atom excess, Shanghai Research Institute of Chemical Industry, China). Each column was adjusted to 80 % field capacity (θg = 0·182 g g^–1) using reverse osmosis (RO) water and allowed to equilibrate.

Columns were hand sown with field pea (Pisum sativum ‘PBA Twilight’) (six per column) on 27 June 2011 at a depth of 2 cm using uniform germinated seeds. Field pea was inoculated using a commercial Group E peat inoculum (Rhizobium leguminosarum). Columns were placed within one of the eight SoilFACE bunkers treated with either ambient CO₂ (aCO₂; 390 ppm) or elevated CO₂ (eCO₂; 550 ppm) concentration, replicated four times on 1 July 2011. After 3 weeks, field pea seedlings were thinned to two plants per column. Overall, the experiment consisted of a randomized split-plot design with 2 CO₂ concentrations (main plots) × 4 NO₃-N levels (sub-plots) × 4 replicates (32 columns in total). Wheat (Triticum aestivum L. ‘Yitpi’) (eight seeds were sown and seedlings thinned to three per column) was included as a non-legume reference plant in the same design but with only two NO₃-N levels (5 and 90 mg N kg^–1) (eight columns in total).

Column harvest and processing

Columns were removed from SoilFACE on 13 October, after 15 weeks of growth (late pod development; 50–60 % of pods at maximum size) and destructively sampled. Prior to removal, the youngest fully emerged leaf was cut from each plant, weighed and placed in liquid N (two leaves per column). Plant shoots were cut off at the soil surface, washed three times with RO water and dried at 70 °C for 3 d. One half of each PVC column was removed and the soil was sectioned into depths of 0–10 and 10–55 cm. Roots within each sample were carefully removed, washed free of soil and debris with RO water, and stored at 5 °C. The soil was thoroughly mixed, stored at 5 °C overnight, and soil N was extracted the following day on field moist soil as outlined below. The remaining soil samples were air-dried at 25 °C for subsequent analyses. After morphological measurements (within 4 d), root samples were dried at 70 °C for 3 d for subsequent analyses.

Nodulation and leghaemoglobin content

For surface (0–10 cm) roots, nodules on the main root and lateral roots were counted; the nodules were then removed using a scalpel and weighed, before being stored at –20 °C. Nodule mass and number of nodules in the subsoil (10–55 cm) were independently assessed by three individuals using a visual ranking system (1–5; lowest/smallest–greatest/largest).

The leghaemoglobin concentration in nodules collected from 0–10 cm was determined using the method of Riley and Dilworth (1985). Briefly, approx. 300 mg of nodule material was ground in 5 mL of 0·1 m sodium phosphate buffer (pH 6·8; 5 °C) using a pre-cooled mortar and pestle. The macerate was passed through muslin and centrifuged at 500 g for 10 min at 5 °C to remove coarse material. The supernatant was then centrifuged at 4400 g for 25 min at 5 °C, and the leghaemoglobin concentration was determined spectrophotometrically at 555 nm (Cary 50, UV-Visible Spectrophotometer, Agilent, USA) following conversion to pyridine haemochromogen (1:1 supernatant/reagent; containing 0·2 m KOH in 50 % pyridine) according to Paul et al. (1953).

Nitrate reductase activity

Nitrate reductase activity (NRA) in field pea leaves was determined using the approach of Kaiser et al. (2000) but with some modifications. Leaves (approx. 500 mg) were ground with 2 mL of extraction buffer [50 mm potassium phosphate buffer, pH 7·6; 1 mm dithiothreitol (DTT); 10 μm FAD; 10 mm MgCl₂] using a mortar and pestle pre-cooled with liquid N. Extracts were transferred to tubes and centrifuged at 16 000 g for 10 min at 4 °C. Aliquots (100 μL) were added to 900 μL of reaction buffer (50 mm potassium phosphate buffer, pH 7·6; 1 mm DTT; 10 μm FAD; 10 mm MgCl₂; 5 mm KNO₃; 9·2 mm NADH) and incubated for 3 min at 24 °C. The reaction was stopped using 125 μL of 0·5 m zinc acetate, and solutions were centrifuged as previously described. Initial and incubated extracts were mixed 1:1 with colour reagent [1·5 g L^–1 NED (naphthylethylenediamine dihydrochloride); 2·5 g L^–1 sulphanilamide; 1·13 n HCl), allowed to stand for 30 min, and nitrite concentrations were determined colorimetrically at 540 nm (Cary 50, UV-Visible Spectrophotometer, Agilent). The desalting of extracts prior to the assay performed by Kaiser et al. (2000) did not show any benefit and hence was omitted. Furthermore, potassium phosphate buffer was used instead of HEPES-KOH due to greater enzyme activity in preliminary tests (data not shown).

Soil and plant analyses

Soil texture was characterized by determining the particle size distribution using a Laser Particle Size Analyser (Malvern Mastersizer 2000, Worcestershire, UK) following dispersion of soil (approx. 10 g) with 10 mL of 0·164 m Na₆P₅O₁₈ (VWR, Australia) in 800 mL of RO water. Soil N extractions were performed on field moist soil using 25 g of soil (dry weight basis) with 2 m KCl (1:1) and shaking end-over-end for 1 h, centrifuging at 2000 g for 5 min and filtering through Whatman #1 filter papers (Whatman International, UK). Filtered extracts were frozen and later analysed for nitrate (NO₃^–) and ammonium (NH₄⁺) using a QuickChem 8500 Flow Injection Analyser (Lachat Instruments, USA).

Dried plant samples were ground (<2 mm) using a Retsch ZM200 centrifugal mill (Retsch GmbH, Haan, Germany) to reduce sample volume. Sub-samples of both ground plant material and whole soil were then finely ground using a Retsch MM400 mixer mill (Retsch GmbH).

Isotope ratio mass spectrometry (IRMS) (Hydra 20-20; SerCon, Crewe, UK) was used to determine the total C and N concentration and ¹⁵N abundance (atom% ¹⁵N) of plant and soil samples. Total N content (mg N per column) and the amount of N derived from either the atmosphere or the fertilizer were calculated using combined N and ¹⁵N in the whole plant and the entire soil column. The percentage of N derived from the atmosphere (% Ndfa) and fertilizer (% Ndff) was calculated according to McAuliffe et al. (1958) using wheat as the reference (Supplementary Data Table S1). The isotope dilution approach can overestimate % Ndfa as it does not account for N derived from soil. In the current study, the soil–sand mix contained 4 mg of NO₃^–+ NH₄⁺kg^–1 which constitutes an average error of approx. 6·3 %.

Statistical analyses

A two-way analysis of variance (ANOVA) using a split-plot design was used to test the effects of CO₂ concentration (main plots) and NO₃-N level (sub-plots) on plant and soil parameters. Differences between means were tested using the least significance difference (LSD) test at P = 0·05.

RESULTS

Plant growth

Shoot growth of field pea was significantly (P = 0·016) greater under eCO₂ than under aCO₂. Shoot biomass ranged from 11 to 15 g per column under aCO₂ and on average increased by 32 % under eCO₂ (Fig. 1A). Overall, shoot biomass was ten times greater than root biomass. Since the root biomass in the topsoil (0–10 cm) was small (15 % of the total root mass) and was not affected by the treatments, the pooled root biomass for each column was used for analyses. For root biomass, there was an apparent increase under eCO₂; however, this was not significant (P = 0·086). Higher rates of NO₃^–-N significantly (P = 0·038) increased root mass. In particular, 25 mg NO₃^–-N kg^–1 soil had a lower root mass than 90 mg NO₃^–-N kg^–1 (Fig. 1B). No effect of CO₂ or N level on the root:shoot ratio was observed (Fig. 1C). For shoot and root biomass, no significant interactions between CO₂ and NO₃^–-N concentration were observed.

Fig. 1.

Shoot (A) and root (B) biomass and root:shoot ratio (C) of field pea grown for 15 weeks under either an aCO₂ or eCO₂ concentration and with 5, 25, 50 or 90 mg NO₃^–-N kg^–1 soil. Root mass includes nodules. Error bars indicate the s.e.m. (n = 4). Means with the same lower case letter are not significantly (P < 0·05) different.

Nodule number, mass and leghaemoglobin content

Nodulation of field pea was affected by NO₃^–-N more than CO₂ concentration, and no interactions were observed (Table 1). For surface roots (0–10 cm), nodule numbers on primary and lateral roots were not different between aCO₂ and eCO₂. Under aCO₂, both the number and mass of nodules on lateral roots significantly (P < 0·05) decreased with increasing N level; however, at eCO₂, this was only significant for 50 mg NO₃^–-N kg^–1. The mass of field pea nodules on roots in the 0–10 cm layer was reduced by 85 % between 5 and 90 mg NO₃^–-N under aCO₂. While the mass of nodules in the 0–10 cm layer was also reduced under eCO₂, the reduction was much lower in magnitude (72 %) and only significantly lower in the 90 mg NO₃^–-N treatment. Hence, the decrease in nodule mass under eCO₂ was not as sharp when NO₃^–-N was increased compared with aCO₂. The decrease in nodule mass with increasing NO₃^–-N was largely due to a reduction in the size of nodules on the primary root. However, NO₃^–-N and CO₂ concentration did not alter the leghaemoglobin concentration expressed per mass of nodule tissue (mmol g^–1) or normalized to account for nodule size (mmol per 1000 nodules) (Table 1). Due to the volume of roots in the deeper soil layer (10–55 cm), visual ranking was used to assess the number and size of root nodules. Consistent with surface soil, the number and size of nodules in the deeper layer were significantly reduced when the NO₃^–-N concentration was increased (Table 1). In general, nodules in the subsoil were more numerous but much smaller in size than in the topsoil.

Table 1.

Number, mass and leghaemoglobin content of root nodules in the 0–10 cm layer (Root₁) and visual ranking (1–5; lowest to highest and smallest to largest) of root nodules in the 10–55 cm layer (Root₂) of field pea grown for 15 weeks under either an aCO₂ or eCO₂ concentration and with 5, 25, 50 or 90 mg NO₃^–-N kg^–1 soil

Factor	N rate (mg kg–1)	Primary root1 nodules (per plant)	Lateral root1 nodules (per plant)	Root1 nodule mass (mg per column)	Root1 nodule size (mg per nodule)	Root2 nodule ranking (1–5)		Root1 leghaemoglobin concentration (µmol)
						n	Size	g–1	Per nodule
N rate (mg kg–1)	5	28·0	127·9	512	1·26	3·4	4·0	628	1·66
	25	27·2	103·8	327	1·29	3·1	3·5	575	2·42
	50	27·7	70·4	267	1·01	2·5	2·8	557	2·27
	90	25·8	93·4	108	0·42	1·2	1·5	616	2·77
CO2 level	aCO2	27·1	97·7	272	0·97	2·4	2·9	567	2·5
	eCO2	27·2	100·0	335	1·02	2·8	3·0	621	2·1
N rate		n.s.	*	*	***	**	***	n.s.	n.s.
LSD (P = 0·05)		–	31	236	0·32	0·38	0·73	–	–
CO2 level		n.s.	n.s.	n.s.	n.s.	n.s.	n.s.	n.s.	n.s.
LSD (P = 0·05)		–	–	–	–	–	–	–	–

No CO₂ × N interactions were observed (P > 0·05).

n.s., P > 0·05; *P < 0·05; **P < 0·01; ***P < 0·001.

–, not applicable; LSD, least significant difference.

N concentration and ¹⁵N abundance

The concentration of N in shoot and roots of field pea was not affected by CO₂ or N level (Table 2). Consequently, the C:N ratio of plant material was also not affected by the treatments. However, the N concentration of deeper roots (10–55 cm) was higher than that of surface roots (0–10 cm), resulting in lower C:N ratios. In contrast to overall N concentration, ¹⁵N abundance (atom%) of plant material was significantly affected by CO₂ concentration (P < 0·05) and N level (P < 0·001). For shoot, ¹⁵N abundance was greater under aCO₂ than eCO₂, except for 90 mg NO₃^–-N kg^–1. In general, greater ¹⁵N abundance under aCO₂ was observed for roots, except surface roots (0–10 cm) with 5 mg NO₃^–-N kg^–1 and deep roots (10–55 cm) with 5 and 90 mg NO₃^–-N kg^–1.

Table 2.

N concentration, C:N ratio and ¹⁵N abundance in shoot and root of field pea grown for 15 weeks under either an aCO₂ or eCO₂ concentration and with 5, 25, 50 or 90 mg NO₃^–-N kg^–1 soil

Factor	N rate (mg kg–1)	N concentration (g kg–1)			C:N			15N (atom%)
		Shoot	Root1	Root2	Shoot	Root1	Root2	Shoot	Root1	Root2
N rate (mg kg–1)	5	20·7	24·3	29·3	20·4	16·3	14·0	1·6	1·2	2·5
	25	20·3	24·3	29·5	20·8	16·1	14·1	6·0	4·1	7·4
	50	21·9	22·3	29·3	20·7	16·9	14·1	2·6	2·0	2·9
	90	20·5	24·2	29·2	20·7	16·4	13·6	3·1	2·4	3·3
CO2 level	aCO2	20·8	24·1	30·1	20·1	16·0	13·4	3·8	2·7	4·3
	eCO2	20·9	23·4	28·5	21·3	16·8	14·6	2·9	2·2	3·8
N rate		n.s.	n.s.	n.s.	n.s.	n.s.	n.s.	***	***	***
LSD (P = 0·05)		–	–	–	–	–	–	1·08	0·32	0·63
CO2 level		n.s.	n.s.	n.s.	n.s.	n.s.	n.s.	*	*	*
LSD (P = 0·05)		–	–	–	–	–	–	0·72	0·30	0·30

Roots from 0–10 cm (root₁) and 10–55 cm (root₂) were analysed separately.

No CO₂ × N interactions were observed (P > 0·05).

n.s., P > 0·05; *P < 0·05; ***P < 0·001.

–, not applicable; LSD, least significant difference

Total N content and N derived from fertilizer

Total N content was significantly (P = 0·011) greater under eCO₂; however, NO₃^–-N concentration was less (Fig. 2A), and was only different between 25 and 90 mg NO₃^–-N kg^–1. On average, total N content was 294 mg N per column under aCO₂, and was increased by 23–54 % under eCO₂. Fertilizer recovery was generally high (55 and 74 %) since inorganic N (NO_x^–+ NH₄⁺) (45·7 mg N per column) was relatively low compared with the amount of fertilizer added (495 mg N per column in the 90 mg kg^–1 treatment) (data not shown). The percentage of N derived from fertilizer (% Ndff) showed a significant (P = 0·002) CO₂ × N interaction (Fig. 2B). Specifically, field pea plants grown under aCO₂ derived a higher proportion of N from fertilizer and the maximal increase was 61 % for the 50 mg NO₃^–-N kg^–1 treatment.

Fig. 2.

Total N content (A) and percentage of N derived from the fertilizer (B) for field pea grown for 15 weeks under either an aCO₂ or eCO₂ concentration and 5, 25, 50 or 90 mg NO₃^–-N kg^–1 soil. Bars indicate the s.e.m. (n = 4). Means with the same lower case letter are not significantly (P < 0·05) different.

N₂ fixation

Consistent with reduced fertilizer utilization, the percentage of N derived from the atmosphere (% Ndfa) was significantly (P = 0·011) greater for field pea grown under eCO₂ than under aCO₂ (Fig. 3). An increase in Ndfa of 10 and 139 % under eCO₂ was observed for 5 and 90 mg NO₃^–-N kg^–1, respectively. In contrast, a greater NO₃^– concentration strongly (P < 0·001) reduced Ndfa, indicating that N fixation was severely inhibited. At 90 mg NO₃^–-N kg^–1, the mean Ndfa was five times lower than that measured at 5 mg. Thus, the relative effect of eCO₂ was greater at the higher N rate than the lower one, despite a much lower Ndfa overall.

Fig. 3.

Percentage of N derived from the atmosphere for field pea grown for 15 weeks under either an aCO₂ or eCO₂ concentration and 5 or 90 mg NO₃^–-N kg^–1 soil. Bars indicate the s.e.m. (n = 4). Means with the same lower case letter are not significantly (P < 0·05) different.

Nitrate reductase activity

Nitrate reductase activity was determined in the youngest fully expanded leaves of field pea. There was an apparent increase in NRA between 5 and 50 mg NO₃^–-N kg^–1 and then a sharp drop in NRA at 90 mg NO₃^–-N kg^–1. However, the data obtained for NRA was highly variable, and consequently no effects of NO₃^– and CO₂ concentration were observed (P > 0·05) (data not presented).

Soil inorganic N

Inorganic N (NO_x^–+ NH₄⁺) in the soil was pooled since NH₄⁺ concentration was low (<4 % of NO₃^–) most probably due to the neutral pH. The inorganic N concentration in the soil was significantly (P < 0·001) affected by the N rate (Fig. 4). The inorganic N concentration increased with increasing rate of NO₃^–; however, except for the 90 mg NO₃^–-N kg^–1 treatment, all other N rates were not significantly different (P > 0·05). In particular, the 90 mg NO₃^–-N kg^–1 treatment had approx. 310 mg inorganic N per column, which was between six and 43 times more than the other treatments. While the inorganic N concentration of soil appeared to be lower under eCO₂, this was not significant (P > 0·05).

Fig. 4.

Inorganic N (NO_x^–+ NH₄⁺) in soil following harvest of field pea grown for 15 weeks under either an aCO₂ or eCO₂ concentration and 5, 25, 50 and 90 mg NO₃^–-N kg^–1 soil. Bars indicate the s.e.m. (n = 4). Means with the same lower case letter are not significantly (P < 0·05) different.

DISCUSSION

Effects of CO₂ and N level on plant growth and nodulation

Shoot and root biomass of field pea grown under aCO₂ was lower at 25 mg N kg^–1 than at 90 mg N kg^–1 (Fig. 1). The relative difference between these treatments was probably a function of contrasting N source and C allocation by plants. Moderate (approx. 25 mg N kg^–1) levels of soil NO₃^– are known to inhibit legume symbiosis and nodule function (Streeter and Wong, 1988; Cowie et al., 1990; Naudin et al., 2011), resulting in insufficient N acquisition to meet the plant’s requirement. At high (approx. 90 mg N kg^–1) levels of soil NO₃^–, field pea derived less N from N₂ fixation (Fig. 3) but still maintained greater biomass due to greater availability of soil N. Photosynthates are known to be preferentially diverted to nodule morphogenesis rather than growth, particularly during the early growth stages (Voisin et al., 2002a). Therefore, lower biomass of the 25 mg N kg^–1 than the 90 mg N kg^–1 treatment could be attributed to either lower levels of photosynthate C or greater allocation of C to the production and maintenance of root nodules. This phenomenon has been used to explain greater biomass production, particularly of roots, in NO₃^–-fed field pea plants (Aranjuelo et al., 2013). In the current study, the N level did not alter the root:shoot ratio, which is similar to the findings other studies (Voisin et al., 2003a; Aranjuelo et al., 2013), and the low root biomass (root:shoot 0·6–0·7) is consistent for this species (Nuruzzaman et al., 2006) (Fig. 1C).

The inhibitory effects of NO₃^– on nodule formation and development observed in the current study are consistent with previous reports (Tang et al., 1999; Bollman and Vessey, 2006; Aranjuelo et al., 2013). Although the numbers of nodules on the primary roots were the same for all treatments, nodulation of lateral roots and total nodule mass in the topsoil (0–10 cm) were significantly reduced as the NO₃^– level was increased (Table 1). The reduction in nodule number on the lateral roots but not the primary root may be due to spatial and temporal depletion of soil NO₃^–. It is likely that primary root nodules of high N treatments were formed but their development was delayed until later stages of growth once the NO₃^– concentration in the rhizosphere soil was sufficiently depleted. The reduction in nodule mass with increased NO₃^– level, predominantly due to lower primary root nodule mass, supports the idea of an inhibition of nodule growth and development. In contrast, lateral roots that are much finer and constantly grow into new areas of soil may not have been able to deplete the soil NO₃^– concentration sufficiently to allow nodulation and/or nodule function. Reductions in the number and size of root nodules with increasing NO₃^– level in the 10–55 cm soil layer were consistent with field pea nodulation in the topsoil (0–10 cm) (Table 1).

Field pea plants grown under eCO₂ had greater biomass (approx. 32 %) than those grown under aCO₂ and were not affected by the soil NO₃^– level (Fig. 1). Therefore, field pea grown under eCO₂ achieved the same biomass irrespective of N source. While eCO₂ appeared to alleviate the inhibitory effects of NO₃^– on nodule number and mass (as indicated by a greater % Ndfa, discussed later), these effects were not significant (Table 1). Changes in nodulation and nodule morphology were expected under eCO₂, particularly for the 25 mg N kg^–1 treatment that showed the largest gain in biomass production. Other studies have shown increased mass and activity of field pea nodules under eCO₂ (Cabrerizo et al., 2001; Aranjuelo et al., 2014). Enhanced nodule function is likely to have contributed to productivity gains under eCO₂, although the concentration of leghaemoglobin in nodules, used here as an indicator of nodule activity, did not support this (Table 1). However, nodule activity is known to decrease during reproductive growth due to greater allocation of assimilates to grain (Voisin et al., 2002b). Hence, the leghaemoglobin concentrations observed may be a poor indicator of actual nodule activity during the growing season. Furthermore, primary root nodules dominated the nodule biomass, and the greater activity of more numerous but smaller lateral root nodules under eCO₂ may have been masked. Importantly, the parameters used to assess treatment effects on field pea nodules were not related to overall nodule function (Ndfa), as discussed later.

Enhanced productivity of field pea under eCO₂ was probably facilitated by greater availability of photosynthates, particularly to root nodules. Using a ¹³CO₂ pulse-labelling approach, Voisin et al. (2003b) showed that C allocated to field pea roots increased with atmospheric CO₂ concentration due to enhanced net photosynthesis. Similarly, field pea nodules are a major C sink and are generally substrate limited under aCO₂ conditions (Aranjuelo et al., 2013). Furthermore, the additional N demand of field pea grown under eCO₂ could have resulted in a more rapid depletion of soil N and subsequently allowed the establishment of N₂ fixation earlier than plants grown under aCO₂. Since field pea has a relatively small root system and excretes less carbohydrates and enzymes compared with other crop legumes (Nuruzzaman et al., 2006), the contribution of microbial N immobilization to reductions in rhizosphere NO₃^– concentration are expected to be small compared with plant N demand. Nodulation rapidly commences once the NO₃^– concentration falls below a critical level (Voisin et al., 2002b), but this recovery is dependent on C availability to nodules (Naudin et al., 2011). Evidence suggests that C supply to field pea roots is independent of the N source and extent of nodulation; however, nodulated roots have lower C use efficiency due to greater respiration losses (Voisin et al., 2003a, b). In the current study, the relative increase in biomass of field pea grown under eCO₂ was greater for 25 mg N kg^–1 than for 90 mg N kg^–1, and this could be due to the lower C use efficiency of well-nodulated plants in the lower N aCO₂ treatment. Alternatively, photosynthetic acclimation or downregulation of photosynthesis could have occurred in the high NO₃^– treatment due to lower sink strength for photosynthate C (Gavito et al., 2000; Aranjuelo et al., 2013).

Effects of CO₂ and N level on legume performance

The current study highlighted that the total N content of field pea grown under eCO₂ was enhanced (Fig. 2A). As expected, the N concentration and C:N ratio of shoots and roots were not affected by either treatment (Table 2) and, subsequently, the total N content was a function of plant biomass. Generally, the N concentration of legumes is maintained under eCO₂ unless environmental or nutritional limitations to N₂ fixation, such as phosphorus, exist (Jin et al., 2012; Lam et al., 2012). Greater N content was primarily achieved by N₂ fixation since N fertilizer recovery (data not shown) and the soil N concentration at harvest (Fig. 4) did not differ between the CO₂ treatments. The % Ndfa values for field pea grown at the low N level (74 % aCO₂, 81 % eCO₂) were consistent with other studies (average ≥80 %) (Voisin et al., 2002b; Hauggaard-Nielsen et al., 2010) but higher than commonly observed in southern Australia (average 68 %) (Unkovich et al., 1997). Accordingly, the amount of N₂ fixed under aCO₂ and eCO₂ ranged from 263 to 376 mg N per column at the low N level and from 49 to 111 mg N per column for the high N level, respectively. However, it is unclear whether specific increases in N₂ fixation efficiency occurred since this process was not directly quantified. The N concentration did not affect the rate of N₂ fixation of field pea but simply delayed the onset of N₂ fixation (Voisin et al., 2002b). Similarly, greater Ndfa of white clover under eCO₂ was due to greater nodule mass rather than specific nodule activity (Zanetti et al., 1998). However, NO₃^– added after legume establishment can reduce both nodule biomass and specific activity (Naudin et al., 2011).

Importantly, this study showed that Ndfa was greater under eCO₂ irrespective of N level. In the current study, Ndfa decreased and was inversely proportional to Ndff as the N level increased (Fig. 3). This substitution of N₂ fixation with soil N has been well established (Streeter and Wong, 1988; Naudin et al., 2011). In fact, Voisin et al. (2002b) suggest that Ndfa can be predicted across the growing season from initial mineral N at sowing. Results obtained in the current study conform with the relationship between N level and Ndfa reported by Jensen (1986) (data not shown). In contrast to other studies, N₂ fixation was less inhibited at the highest N level (equivalent to 630 kg N ha^–1) and may be due to the fact that fertilizer was mixed to a greater depth (uniformly through the 55 cm soil columns). Greater inhibition of N₂ fixation by inorganic N in other studies could also to be due to high existing soil N pools (Lam et al., 2012) or the use of NH₄NO₃^– as the N fertilizer source (Voisin et al., 2002b). NH₄⁺ has been shown to reverse the inhibitory effects of NO₃^– in artificial growth media (Bollman and Vessey, 2006). However, in soil and when added at the start of the growing season in the absence of living roots, NH₄⁺ is likely to be converted to NO₃^–, thereby underestimating the actual NO₃- concentration.

Conclusions

Using FACE combined with a ¹⁵N isotope dilution approach, this study showed for the first time that eCO₂ reduced the inhibitory effects of soil NO₃^– on N₂ fixation of field pea. Under eCO₂, the N content of field pea was enhanced via greater biomass production and higher % Ndfa, although the physiological (number, size and activity of nodules) differences were not evident at the time of sampling. It is possible that eCO₂ increased the critical concentration at which N₂ fixation of field pea was inhibited and/or field pea was more efficient at switching between fertilizer N and N₂ fixation under eCO₂ due to more rapid depletion of the rhizosphere NO₃^– concentration. The current results indicate that field pea may perform well in semi-arid agricultural systems under future CO₂ concentrations irrespective of soil N status, and subsequent gains in N input via enhanced N₂ fixation will be important for maintaining the N fertility of cropping systems.

SUPPLEMENTARY DATA

Supplementary data are available online at www.aob.oxfordjournals.org and consist of Table S1: biomass, N concentration, C:N ratio and ¹⁵N abundance in shoot and root of wheat grown under either aCO₂ or eCO₂ concentrations and with 5 or 90 mg NO₃^–-N kg^–1 soil.