Arbuscular mycorrhizal colonization on carbon economy in perennial ryegrass: quantification by 13CO2/12CO2 steady-state labelling and gas exchange

Authors


Author for correspondence: Hans Schnyder Tel: +49 8161 713242 Fax: +4 8161 713243 Email: schnyder@wzw.tum.de

Summary

  • • Effects of the arbuscular mycorrhizal fungus (AMF) Glomus hoi on the carbon economy of perennial ryegrass (Lolium perenne) were investigated by comparing nonmycorrhizal and mycorrhizal plants of the same size, morphology and phosphorus status.
  • • Plants were grown in the presence of CO2 sources with different C isotope composition (δ13C −1 or −44). Relative respiration and gross photosynthesis rates, and belowground allocation of C assimilated during one light period (‘new C’), as well as its contribution to respiration, were quantified by the concerted use of 13CO2/12CO2 steady-state labelling and 13CO2/12CO2 gas-exchange techniques.
  • • AMF (G. hoi) enhanced the relative respiration rate of the root + soil system by 16%, inducing an extra C flow amounting to 3% of daily gross photosynthesis. Total C flow into AMF growth and respiration was estimated at < 8% of daily gross photosynthesis. This was associated with a greater amount of new C allocated belowground and respired in mycorrhizal plants. AMF colonization affected the sources supplying belowground respiration, indicating a greater importance of plant C stores in supplying respiration and/or the participation of storage pools within fungal tissues.
  • • When ontogenetic and nutritional effects were accounted for, AMF increased belowground C costs, which were not compensated by increased photosynthesis rates. Therefore the instantaneous relative growth rate was lower in mycorrhizal plants.

Introduction

As obligate symbionts, arbuscular mycorrhizal fungi (AMF) depend entirely on the supply of carbon substrates from the host plant (Ho & Trappe, 1973), and therefore are an integral part of the C economy of colonized plants. For instance, AMF typically enhance belowground demand for photoassimilates (Snellgrove et al., 1982; Harris et al., 1985; Douds et al., 1988; Jakobsen & Rosendahl, 1990; Eissenstat et al., 1993). In some scenarios, the proportion of C flow to the AMF was large enough to slow down plant growth (Johnson et al., 1997). More generally, the C cost of the fungus appears to be overcompensated by the benefits of the symbiosis, which consist mainly of the improvement of nutrient uptake (principally P, but also NH4, zinc, copper and other micronutrients), and water status (Koide, 1991; Marschner, 1995; Smith & Read, 1997); increases in plant tolerance to various kinds of stress (Smith & Read, 1997); and protection against root pathogenic fungi (Newsham et al., 1995). In the grass species Lolium perenne, we have recently shown that AMF effects on plant morphology (Grimoldi et al., 2005) and leaf growth (Kavanováet al., 2006) were largely mediated by the improvement of the phosphorus nutrition status. Some authors reported that AMF could cause an enhancement of photosynthesis rates not mediated by nutritional effects (Brown & Bethlenfalvay, 1988; Wright et al., 1998a; Miller et al., 2002), which was attributed to the additional sink activity of the fungus. But this effect, which counterbalanced part of the symbiotic costs, seems not to be general (Snellgrove et al., 1982; Douds et al., 1988; Freeden & Terry, 1988; Bass et al., 1989; Pearson & Jakobsen, 1993; Black et al., 2000 and references therein).

The C flow to AMF is equivalent to anything between 4 and 20% of daily gross photosynthesis (Smith & Read, 1997). This variability in C demand of the AMF appears to be related to the particular identity of the partners involved in the symbiosis (Pearson & Jakobsen, 1993; Lerat et al., 2003; Heinemeyer & Fitter, 2004; Munkvold et al., 2004; Smith et al., 2004; Jakobsen et al., 2005). But the actual cause(s) behind it are still unclear, as mechanistic understanding of how plants regulate C partitioning to the fungal partner is far from complete (Fitter, 2005). In part, this is caused by methodological problems in assessing daily gross photosynthesis, and the share and sources of C substrates taken by the AMF. Further, it is of no minor importance that experimental designs are used that disentangle the effects of AMF from those of ontogenetic and nutritional status (Staddon et al., 1999; Heinemeyer & Fitter, 2004; Grimoldi et al., 2005).

The present study investigates the effects of the AMF Glomus hoi on the C economy of perennial ryegrass (L. perenne). The specific aims were: to estimate C flow to the AMF; to analyse the influence of AMF on the amount of recently assimilated C (termed ‘new C’) allocated belowground; and to quantify the importance of new C in supplying respiration. To this end, the C-balance components of nonmycorrhizal and mycorrhizal plants of the same size, morphology and P status were assessed by the concerted use of 13CO2/12CO2 steady-state labelling and 13CO2/12CO2 gas-exchange techniques so as to label all photosynthate assimilated during one light period, and to follow the contribution of new (labelled) and old C (unlabelled) to above- and belowground respiration over the next dark period. To the best of our knowledge, this is the first whole-plant C balance comparing AMF effects in a grass species. We also provide the first quantitative analysis of allocation and respiratory use of new C as affected by mycorrhizal symbiosis.

Materials and Methods

Plant material, AMF inoculation and growth conditions

Seeds of perennial ryegrass (Lolium perenne L. cv. Condesa) were surface sterilized for 20 min in NaOCl (6% active chlorine) and sown in pots (diameter 5 cm, depth 35 cm). Pots were filled with quartz sand (0.3–0.8 mm) and fertilized with fine powdered Hyperphos (63 mg P per pot). Half the pots were inoculated with the AMF Glomus hoi BEG 104 (provided by Dr A. Heinemeyer, University of York, UK). For each pot of the AMF treatment, 15 ml of the inoculum, consisting of a mixture of dead fine roots of mycorrhizal Plantago lanceolata and fine sand, was mixed thoroughly with the quartz sand (Grimoldi et al., 2005). Pots with and without AMF inoculum were placed in separate plastic containers to prevent colonization of the nonmycorrhizal plants with AMF.

Plants were grown for 10 wk in two controlled environment chambers (E15, Conviron, Winnipeg, Canada) with 80% relative humidity, 20 : 15°C (day : night) and 525 µmol m−2 s−1 photosynthetic photon flux density at plant height for 16 h d−1. Two containers (–AMF and +AMF) were placed in each chamber. Plants were watered using an automatic irrigation system supplying, to each plant, four times a day, 25 ml of a modified half-strength low-P Hoagland's solution (0.02 mm KH2PO4, 2.5 mm KNO3, 2.5 mm Ca (NO3)2, 1 mm MgSO4, 0.98 mm KCl, 0.5 mm NaCl, 0.125 mm Fe-EDTA, 23 µm H3BO3, 4.5 µm MnSO4, 0.38 µm ZnSO4, 0.16 µm CuSO4, 0.05 µm Na2MoO4). Pots were flushed weekly with distilled water to prevent salt accumulation.

Sampling protocol

All variables were analysed in plants of similar size (1.1–1.5 g C per plant) selected from a large group of plants in order to isolate AMF effects from ontogenetic drift and P status. Immediately after harvest, roots were freed from the soil substrate by washing with tap water. Washing caused the loss of most of the extraradical mycelium of the AMF. A sample of the fresh root material was weighed and used for detection of AMF colonization. Shoots were separated into individual tillers, which were counted and dissected into lamina and sheath. The area of the laminas was measured with a LI-3100 leaf area meter (Li-Cor, Lincoln, NE, USA). Leaf-area ratio (LAR; cm2 g−1 C) was determined as total leaf area divided by plant C mass; specific leaf area (SLA; cm2 g−1 C) as plant leaf area divided by lamina C mass. Samples were frozen in liquid nitrogen, freeze-dried, weighed, ground and stored at −25°C.

Chemical and isotopic analyses and AMF colonization

The concentrations of C and N and 13CO2/12CO2 isotope ratios were determined on aliquots of 0.7 mg dry ground material using an elemental analyser (NA1110, Carlo Erba Instruments, Milan, Italy) interfaced to a continuous-flow isotope-ratio mass spectrometer (IRMS, Delta Plus, Finnigan MAT, Bremen, Germany). All C-isotope data are expressed in the conventional form: δ13C sample = (13C/12C in sample/13C/12C in the VPDB standard −1) × 1000. Phosphorus concentration was determined on 25-mg (DW) aliquots, ashed in a muffle furnace (4 h at 500°C), and then digested in HNO3/HCl. Phosphorus was quantified by phosphovanado-molybdate colorimetry (Hanson, 1950). Water-soluble carbohydrates were analysed as described by Schnyder & de Visser (1999). The content of C in water-soluble carbohydrates was estimated as total hexose units × 0.4.

Mycorrhizal colonization of roots was determined by histological detection of mycorrhizal structures after root staining as described by Grimoldi et al. (2005). Briefly, a sample of fresh material was cleared in KOH (10% w/v) for 10 min at 105°C, acidified in HCl (1% v/v) for 5 min, then stained with Trypan Blue (0.05% w/v; Sigma-Aldrich, Steinheim, Germany) in acid glycerol for 10 min at 105°C. The percentage of root length with AMF colonization was determined in glycerol–gelatine-mounted roots by evaluating 100 random intersections for each plant, using the gridline interception method (Giovannetti & Mosse, 1980).

13CO2/12CO2-labelling procedure

The 13CO2/12CO2-labelling system described by Schnyder et al. (2003) was used. During the whole experiment, one growth chamber received CO2 from a mineral source (δ13C −1), the other from an industrial source (δ13C −44; both CO2 sources from Linde AG, Höllriegelskreuth, Germany). The system allowed independent control of CO2 concentration (360 µl l−1) and isotopic composition (δ13C +3 or −40 at the outlet of the growth chambers). As an example, the isotopic compositions (δ13C; ) of shoot biomass of nonmycorrhizal plants continuously grown in the same chamber (end members) were −22.2 ± 0.3 and −64.6 ± 0.2, respectively. After 10 wk growth, C assimilation was quantified by labelling the plants with the different CO2 isotopic sources. For this, plants were swapped between chambers (mineral → industrial CO2 and vice versa) shortly before the lights went on, and labelled during one entire light period (16 h) as described by Lattanzi et al. (2005). Labelling for one light period caused a shift in the δ13C of total plant biomass of nonmycorrhizal plants of approx. 4, which was equivalent to approx. 10% new C in total biomass. The δ-shift brought about by labelling was virtually the same in the different chambers, and approximately 10 times larger than the variability among plants.

Respiratory 13CO2/12CO2 exchange system and measurement

The respiration measurement system has been described in detail by Lötscher et al. (2004); Klumpp et al. (2005). The system allowed simultaneous measurement of dark respiration rate and isotopic composition of the respired CO2 of shoot and root + soil compartments for four individual plants. Respiration measurements were performed at dark-period temperature (15°C). Respiration rates of unlabelled and labelled plants were measured during the dark period following the labelling light period, allowing quantification of the contribution of ‘new C’ (labelled) and ‘old C’ (unlabelled) to the total respiration rate (see below). Pots were removed from chambers at the beginning of the dark period, flushed with 0.5 l distilled water, and rinsed with CO2-free nutrient solution, removing all CO2 from the root + soil compartment. System tests demonstrated that up to 1 h was required to purge the system from all extraneous CO2 (Klumpp et al., 2005). Dark respiration rates were measured every 38 min for 5–6 h, after which plants were harvested. This procedure was performed during four subsequent days; every day one unlabelled and one labelled individual plant from each treatment (–AMF and +AMF) were included. Thus labelled target plants (–AMF and +AMF) were measured in pairs, and a daily whole C balance was constructed for each individual plant (see below).

Gas-exchange parameters and C balance

Carbon isotope data (δ13C) were used to calculate the fraction of new C incorporated into biomass and respired during a complete diurnal cycle. The fraction of new C in biomass (fB new) was calculated following Schnyder & de Visser (1999):

fB new = (δP − δPO)/(δPL − δPO)(Eqn 1)

where δP is the δ of a given sample from a labelled plant; and δPL and δPO are the δ of the same tissue collected from plants grown continuously in the labelling chamber (to which the plant was transferred for labelling during one light period), and in the chamber of origin (in which the plant was grown before the transfer). In the same way, the fraction of new C in the respired CO2 (fR new) was calculated as:

fR new = (δR − δRO)/(δRL − δRO)(Eqn 2)

where δR is the δ of the respired CO2 of a plant after one labelling light period; and δRL and δRO are the δ of the respired CO2 of plants continuously grown in the labelling chamber and the chamber of origin, respectively.

13C discrimination (Δ13C) provides information on the relative importance of stomatal conductance and photosynthetic capacity in limiting photosynthesis (Farquhar et al., 1989). Nonlabelled plants from each chamber were used to determine Δ13C as (Farquhar et al., 1989):

Δ13C () = [(δCO2 − δBiomass)/(1000 + δBiomass)] × 1000(Eqn 3)

Respiration rate for the whole day (16 h light, 8 h dark) was calculated as:

R(mg C d−1 per plant) = 16Rlight + 8Rdark(Eqn 4)

where Rlight and Rdark are the hourly respiration rates (shoot and root + soil) during the light and dark period, respectively. Respiration of root + soil (including AMF) in the light period (Rlight R) was estimated from Rdark of root + soil as:

image(Eqn 5)

where TL and TD are the temperatures of the light and dark period, respectively. Shoot respiration in light (Rlight S) was calculated accordingly, but values were corrected because of the inhibition of dark respiration by light observed at leaf (Krömer, 1995 and references therein) and canopy level (Schnyder et al., 2003). In the present experimental conditions, light caused an approx. 30% inhibition of shoot respiration in perennial ryegrass (C. Piel and co-workers, unpublished data). Thus:

image(Eqn 6)

A Q10 = 2 for both shoot and root + soil respiration was used to account for short-term effects of the warmer conditions in the light than in the dark period (Amthor, 1989; Lötscher et al., 2004; Lötscher & Gayler, 2005).

Relative respiration rates (RRR; mg C g−1 C d−1) indicate respiration rate per unit C mass of shoot or root (including intraradical mycelium). The respiration rate of new (labelled) C during the dark period was calculated as:

Rdark new (mg C h−1 per plant) = fR new × Rdark(Eqn 7)

For the light period, the evolution of fR new was assumed to be linear, starting from zero at the beginning of the labelling period and reaching the value found during the subsequent dark period by the end of the light period (Lötscher & Gayler, 2005). Thus:

Rlight new = 0.5 × fR new × Rlight(Eqn 8)

Gross photosynthesis rate (PG; mg C d−1 per plant) was calculated for each plant as the mass of labelled C in the plant at the end of the dark period (Cnew) plus labelled C respired during the whole day:

PG = Cnew + Rlight new + Rdark new(Eqn 9)

Relative gross photosynthesis rate (RPR; mg C g−1 C d−1) is PG per unit of total plant C mass (excluding extraradical mycelium). The instantaneous relative growth rate (RGRi; mg C g−1 C d−1) was calculated as:

RGRi = (PG − Rlight − Rdark)/plant C mass(Eqn 10)

Photosynthetic nutrient-use efficiencies for phosphorus (PPUE; g C g−1 P d−1) and nitrogen (PNUE; g C g−1 N d−1) were calculated as PG divided by shoot nutrient mass.

Estimation of C flow to extraradical mycelium

Although the respiration rate of extraradical mycelium has been measured successfully (Heinemeyer et al., 2006), direct measurements of whole-AMF respiration are (so far) not possible. However, there is good evidence that the surplus of RRR of mycorrhizal relative to nonmycorrhizal root systems is attributable to (maintenance and growth) respiration of the AMF (Baas et al., 1989; Peng et al., 1993; Nielsen et al., 1998). Thus the contribution of the AMF to the relative respiration rate of the root (RRRAMF) can be estimated as:

RRRAMF = RRRNMR − RRRMR(Eqn 11)

where NMR and MR refer to nonmycorrhizal and mycorrhizal root systems, respectively. But respiration rate is also closely related to tissue N content (Amthor, 1989), which differed slightly, but significantly, between root systems of mycorrhizal and nonmycorrhizal plants. We estimated RRRAMF from RRRNMR and RRRMR of root systems normalized for N content, thus:

RRRAMF = RRRMR − (RRRNMR × [NMR/NNMR])(Eqn 12)

where N refers to nitrogen concentration (mg N g−1 C) in the root system.

An estimate of the maximum total daily C flow to the AMF (FAMF, mg C d−1) was derived from the respiration rate of the AMF (RAMF, where RAMF = RRRAMF × CR, and CR is the total C in the root system) and a carbon-use efficiency (CUE) of the AMF of 0.6. CUEAMF is defined as production of fungal biomass (ΔCAMF, mg C plant−1) per unit of total C flow to the AMF. Thus:

CUEAMF = ΔCAMF/FAMF(Eqn 13)
with FAMF = ΔCAMF + RAMF(Eqn 14)

Substitution and rearranging of the above equations yields

FAMF = RAMF/(1 − CUEAMF)(Eqn 15)

showing that C flow to AMF can be obtained from respiration rate and knowledge of CUE. CUE (a synonymous term is ‘apparent growth efficiency’) is not a constant, but depends on maintenance respiration, relative growth rate and the efficiency of biosynthesis (Penning de Vries, 1975; Amthor, 1989). A maximum CUE of 0.6 (g biomass C per g substrate C utilized) has been found for many aerobic heterotrophs growing on a variety of substrates (Payne, 1970). Where growth is nil and all imported C is used in (maintenance) respiration, CUEAMF = 0.

The fraction of new C in the C flow to the AMF was assumed to be equal to that in respiratory CO2 of the nonmycorrhizal plants. Thus:

FAMF new = fR new × FAMF(Eqn 16)

Statistical analyses

All data were examined for normality. Nutritional status and morphological variables were compared using t-tests. Physiological measurements were performed during four subsequent days: each day one unlabelled and one labelled individual plant from each treatment were included. Thus all physiological parameters were analysed by t-test for paired samples, as all data for the daily whole-C balance were derived from nonmycorrhizal and mycorrhizal paired plants corresponding to the same day of measurement. Variables that involved percentages were arcsine square-root transformed before analysis. Statistical analyses were performed using the package statistica ver. 6.0 (Stat Soft, Tulsa, OK, USA).

Results

AMF colonization, plant size and morphological traits

Plants inoculated with G. hoi averaged 32 ± 3% root length colonized. Noninoculated plants remained nonmycorrhizal. All plants were carefully selected to have a similar size. Thus there was no difference in plant biomass (dfe = 14; P = 0.71) and total number of tillers per plant (dfe = 14; P = 0.20) between nonmycorrhizal and mycorrhizal plants (Table 1). Also, there was no significant difference in plant morphological traits: leaf area ratio, root-to-shoot ratio and specific leaf area (Table 1; dfe = 14; P > 0.46 for all variables).

Table 1.  Morphological traits of nonmycorrhizal (–AMF) and mycorrhizal (+AMF) perennial ryegrass (Lolium perenne) plants grown for 10 wk at low soluble phosphorus supply
Variable–AMF+AMF
  1. Differences were not significant (P > 0.05). Values are means ± SE (n = 8).

Tillers per plant14.7 ± 1.112.8 ± 0.8
Plant biomass (g C)1.32 ± 0.071.36 ± 0.09
Leaf area ratio (cm2 g−1 C) 161 ± 7 154 ± 5
Root : shoot ratio (g C g−1 C)0.32 ± 0.020.34 ± 0.01
Specific leaf area (cm2 g−1 C) 300 ± 15 293 ± 11

Respiration and C balance

Relative respiration rates of the root + soil system were 16% greater in mycorrhizal than in nonmycorrhizal plants (Fig. 1; dfe = 6; P < 0.001 for all measurements). No differences were found in RRR of the shoot compartment (Fig. 1; dfe = 6; P > 0.23).

Figure 1.

Relative respiration rate (RRR) during one dark period (8 h, 15°C) in the shoot (circles) and root + soil compartment (triangles) of perennial ryegrass (Lolium perenne). Closed symbols, nonmycorrhizal plants; open symbols, mycorrhizal plants. Values are means ± SE (n = 4).

Relative gross photosynthesis rate tended to be higher in nonmycorrhizal compared with mycorrhizal plants (Fig. 2; dfe = 6; P = 0.09). This was, in a small part, caused by the loss of new C in the extraradical mycelium that was not recovered when harvesting the plants (see Materials and Methods). The amount of new C flow to extraradical mycelium tissues was estimated to be 1.3 ± 0.1 mg C g−1 C d−1. Accounting for this amount reduced the difference in RPR between mycorrhizal and nonmycorrhizal plants to 18% (Fig. 2), which was not significant (dfe = 6; P > 0.16).

Figure 2.

Daily carbon balance of perennial ryegrass (Lolium perenne). Plants were harvested at the end of the dark period following the 16-h light labelling period. Respiration rates were measured during the dark period (8 h, 15°C), and estimated for the preceding light period (16 h, 20°C). Each value of relative gross photosynthesis rate (RPR), relative respiration rate (RRR) and instantaneous relative growth rate (RGRi) and estimated C cost of the production of arbuscular mycorrhizal fungal (AMF) biomass correspond to the same individual plant. Closed bars, nonmycorrhizal plants; open bars, mycorrhizal plants; grey bars, estimation of the C cost of the production of AMF biomass. Values are means ± SE of labelled plants (n = 4). Significant differences: *, P < 0.05.

Instantaneous relative growth rate (RGRi; mg C g−1 C d−1) was lower in mycorrhizal than in nonmycorrhizal plants (Fig. 2; dfe = 6; P < 0.05), because of the reduced relative gross photosynthesis rate (−18%), higher root respiration (+3%), and C flow to growth of extraradical mycelium tissues (estimated up to approx. 5% of plant C) in mycorrhizal plants. For this calculation, the amount of C flow (new and old) to extraradical mycelium tissues was estimated to be 4.1 ± 0.4 mg C g−1 C d−1 (Fig. 2).

Allocation of new C

Most of the new C in both treatments was allocated to shoots (Fig. 3), and the proportion tended to be higher in nonmycorrhizal compared with mycorrhizal plants (Fig. 3; dfe = 6; P < 0.08). But no difference (dfe = 6; P = 0.54) was found in the amount of new C allocated to shoot respiration (Fig. 3). The amount of new C allocated belowground was 15% higher (dfe = 6; P < 0.05) in mycorrhizal than nonmycorrhizal plants (Fig. 3). This was caused by an increased amount of new C respired belowground, and, probably, flow of new C to growth of extraradical mycelium. The amount of new C retained in mycorrhizal roots was slightly, but not significantly, higher compared with nonmycorrhizal plants (Fig. 3, dfe = 6; P = 0.21).

Figure 3.

Allocation of new carbon in perennial ryegrass (Lolium perenne). Shoot and root allocation are expressed per unit of organ C mass. Biomass and respiration data and estimated C cost of the production of arbuscular mycorrhizal fungal (AMF) biomass correspond to the same individual plant. Plants were harvested at the end of the dark period following the 16-h light labelling period. Respiration rates were measured during the dark period (8 h, 15°C), and estimated for the preceding light period (16 h, 20°C). Hatched bars, new C in end-of-day biomass; open bars, daily new C respired; grey bar, estimation of the C cost of the production of AMF biomass. Values are means ± SE of labelled plants (n = 4).

Fraction of new C in respired CO2

The fraction of new C in respired CO2 in the shoot and root + soil compartment was constant during the dark period in both AMF treatments (Fig. 4). The presence of AMF did not affect this parameter in respiration of the shoot compartment (Fig. 4; P = 0.56). In contrast, the fraction of new C in respiration was lower in mycorrhizal roots compared with nonmycorrhizal ones (Fig. 4; P < 0.05).

Figure 4.

Fraction of new carbon in respired CO2 during one dark period (8 h, 15°C) in shoot (circles) and root + soil compartments (triangles) of perennial ryegrass (Lolium perenne). Closed symbols, nonmycorrhizal plants; open symbols, mycorrhizal plants. Values are means ± SE of labelled plants (n = 4).

Plant nutrition and photosynthetic nutrient-use efficiency

The presence of AMF had no statistically detectable effect (dfe = 14; P > 0.05) on plant P content, expressed as shoot and root P : C ratios (w/w; Table 2). Nitrogen concentrations in the shoot and root were 14 and 12%, respectively, higher in nonmycorrhizal compared with mycorrhizal plants (w/w; Table 2; dfe = 14; P < 0.05). The N : P ratio of the total biomass was 14% higher in nonmycorrhizal compared with mycorrhizal plants. Concentrations expressed per unit structural C (when C in water-soluble carbohydrates was subtracted from total C) showed the same relationships (data not shown). Mycorrhizal plants had a higher concentration of water-soluble carbohydrates (mg C g−1 C) in leaf sheaths (–AMF, 204 ± 8 vs +AMF, 265 ± 11; dfe = 14; P < 0.001). But there were no differences in water-soluble carbohydrate concentration in roots (–AMF: 91 ± 9 vs. +AMF: 102 ± 10; dfe = 14; P = 0.45). Photosynthetic nutrient-use efficiencies for phosphorus (PPUE; g C g−1 shoot P d−1) and nitrogen (PNUE; g C g−1 shoot N d−1) were slightly higher (+7 to +12%) in nonmycorrhizal plants, but again these differences were not statistically significant (Table 2; dfe = 6; P > 0.35 for both cases). 13C discrimination (Δ13C; ), as expressed in total plant biomass, was identical in the two treatments (–AMF, 25.5 ± 0.1 vs +AMF, 25.5 ± 0.2; dfe = 14; P = 0.99).

Table 2.  Nutritional status and photosynthetic nutrient-use efficiency of nonmycorrhizal (–AMF) and mycorrhizal (+AMF) perennial ryegrass (Lolium perenne) plants
Variable–AMF+AMF
  1. Nutrient analyses were performed on eight plants per treatment. Photosynthetic nutrient-use efficiencies for phosphorus (PPUE) and nitrogen (PNUE) were calculated for each labelled plant as gross photosynthesis rate (PG) per unit of shoot nutrient (n = 4). All values are means ± SE. Significant difference: *, P < 0.05.

Shoot P (mg P g−1 C)2.12 ± 0.162.01 ± 0.10
Root P (mg P g−1 C)3.08 ± 0.373.38 ± 0.22
Shoot N (mg N g−1 C)66.3 ± 2.258.3 ± 2.0*
Root N (mg N g−1 C)58.2 ± 1.752.2 ± 0.9*
PPUE (g C g−1 shoot P d−1)46.2 ± 6.243.2 ± 2.3
PNUE (g C g−1 shoot N d−1)1.73 ± 0.171.55 ± 0.12

Discussion

The demand for C substrates of mycorrhizal roots

Gas-exchange measurements revealed an increased consumption of assimilates in belowground respiration in mycorrhizal plants. When compared with nonmycorrhizal plants of the same size and P status, the presence of the AMF G. hoi enhanced relative respiration rate of the root + soil system of perennial ryegrass by 16% (Fig. 1). This is not surprising, as ‘extraradical mycelium + colonized roots’ often have enhanced relative respiration rates compared with nonmycorrhizal roots (Baas et al., 1989; Rygiewicz & Andersen, 1994), which has usually been attributed to the respiratory activity of the fungus. Root respiration rate reflects diverse processes such as nutrient uptake, tissue biosynthesis and maintenance of structures. But none of these seemed a plausible cause of the difference observed in the present study. The N status of mycorrhizal plants was lower than that of nonmycorrhizal plants (Table 2), indicating that C costs associated with protein turnover and N uptake and assimilation were smaller than in nonmycorrhizal plants. Further, the amount of new C allocated belowground and retained in root biomass was not affected by the presence of AMF (Fig. 3), indicating that root growth rates were similar. These relationships all support the notion that the enhanced consumption of photoassimilates in belowground respiration of mycorrhizal plants was related to the maintenance and growth respiration of the fungal partner.

In the present study, C flow to AMF accounted for 3% (estimated assuming a CUEAMF = 0) to 8% (estimated using a maximum CUE of 0.60 of the AMF) of daily gross photosynthesis. Recently, Heinemeyer et al. (2006) reported an estimation of hyphal biomass production of Glomus mosseae, together with the first direct measurement of respiration rate of AMF external mycelium. From these values, it is possible to estimate a CUE of approx. 0.28 at low temperature and approx. 0.51 at higher temperature. Additionally, Peng et al. (1993) presented construction costs of citrus roots colonized by the AMF Glomus intraradices. From these values, assuming that 20% of the root system biomass corresponds to intraradical hyphae (as suggested by the authors), it is possible to estimate a CUE of AMF intraradical hyphae of approx. 0.41 (Fig. 5 in Peng et al., 1993). For our analysis, a CUE of 0.3 and 0.45 would yield a C flow to the AMF equivalent to 4.8 and 6%, respectively, of daily gross photosynthesis. To the best of our knowledge, this is the first estimation of C flow to a mycorrhizal partner in a grass species, and it is at the low end of the range of published data. From our results, while the AMF (G. hoi) did affect the plant C balance, it represented a relatively small C cost for perennial ryegrass plants. This is because, first, the contribution of AMF to the relative respiration rate of the root system was low compared with other species (16%); and second, belowground respiration was a relatively minor component of the plant C balance, possibly because of the low fraction of root biomass (25%) and the optimal water and nutrient (other than P) status. In undisturbed plants of perennial ryegrass, the additional C drain of the AMF G. hoi appears to be easily overridden by the beneficial nutritional effects provided by the mycorrhizal symbiosis (Black et al., 2000). In this species, the C-sink demand of AMF might be more significant in scenarios of C stress and higher root-to-shoot ratios, as suggested following events of intense defoliation (Johnson et al., 1997). Whether this situation turns to the opposite when defoliation drastically decreases assimilation rate and increases the relevance of belowground respiration is currently under investigation.

The C flow from host to AMF has been reported to vary from 4 to 20% of total photoassimilates (Smith & Read, 1997). This wide range is in good agreement with the high variability reported in the increase in relative respiration rates induced by AMF colonization: white clover with AMF field inoculum (+18%, Wright et al., 1998b); cucumber with Glomus caledonium (+32%, Pearson & Jakobsen, 1993); citrus with G. intraradices (+37%, Peng et al., 1993); common bean with G. intraradices (+70%, Nielsen et al., 1998), and Plantago major with Glomus fasciculatum (+80%, Baas et al., 1989). The actual causes behind the large variability in observed C demand of AMF are still unclear. On one hand, evidence is accumulating of high functional diversity in the effect of AMF in relation to the identity of both partners (Pearson & Jakobsen, 1993; Lerat et al., 2003; Heinemeyer & Fitter, 2004; Munkvold et al., 2004; Smith et al., 2004; Jakobsen et al., 2005), suggesting that particular associations can be more (or less) C costly for the plant. In this sense, perennial grasses are generally less colonized by AMF and are considered less dependent on mycorrhiza than legumes and other grassland species (Schweiger et al., 1995; Hartnett & Wilson, 2002). This is because they usually possess a highly branched root architecture with very long root hairs, which seems to render less likely benefits from increases in P uptake rate by mycorrhizal symbiosis (Jakobsen et al., 2005). In principle, our results agree with the general view that AMF are of relatively small importance for grass species (Hartnett & Wilson, 2002). Moreover, C flow to the AMF, a sink for C substrates, may be influenced by the relative activity, hierarchy and developmental stage of the different sinks within the plant (Farrar & Jones, 2000) and by other root symbionts (Brown & Bethlenfalvay, 1988). In this context, the (scarce) available estimations of turnover rate (Staddon et al., 2003) suggest that the short half-life of fungal tissues could be a major factor controlling C drain. In an experiment in which plants were exposed to ‘fossil’ (14C-dead) CO2, those authors found a half-life of hyphae < 4 d, suggesting that the relative rate of fungal biomass production must be large to maintain the AMF. Why the turnover is so fast is unknown, and is perhaps the key for understanding the variability in C demand of the different plant–fungus combinations (Staddon & Fitter, 1998). Further experiments should assess this issue in controlled and natural conditions.

Consequences for plant growth

The overall effect of AMF on growth of the host is believed to depend on the balance between the benefits of increased nutrient uptake and the C cost of the fungus (Koide, 1991; Eissenstat et al., 1993; Johnson et al., 1997). It has been reported that AMF colonization could promote C-assimilation rate even in the absence of an effect on plant nutritional status (Brown & Bethlenfalvay, 1988; Wright et al., 1998a; Staddon et al., 1999; Miller et al., 2002). This was suggested to occur via the easing of a (hypothetical) sink limitation of photosynthesis. In this study, when nonmycorrhizal and mycorrhizal perennial ryegrass plants of similar size, morphology and P status were compared, AMF colonization did not enhance the rate of gross photosynthesis. This is in agreement with Black et al. (2000), who showed that increases in photosynthesis rates in cucumber plants were strictly limited to situations where the AMF improved plant P status.

The lower instantaneous relative growth rate of mycorrhizal plants appeared to be only partly caused by C flow to the AMF. Approximately half the difference in relative growth rate was the consequence of an 18% lower gross photosynthesis rate in mycorrhizal plants. This could not be ascribed to a specific effect of AMF on stomatal conductance or photosynthetic capacity: 13C discrimination was high, and virtually identical in the two treatments, suggesting that the relationship between stomatal conductance and photosynthetic capacity was very similar in the two treatments, and that photosynthesis was mainly limited by photosynthetic capacity in both (Farquhar et al., 1989). The difference was not caused by either light capture, because plants had the same leaf area (Table 1); or differences in P concentration, which were very similar in the two treatments. However, shoot N concentration was 14% lower in mycorrhizal plants, which compares favourably with the lower gross photosynthesis rate. The observation of a decreased N uptake was unexpected (Table 2), as full nitrate-N supply was applied to both treatments, but it might be related to the fact that AMF seem unable to deliver nitrate-N to the host plant (Tanaka & Yano, 2005). Such AMF effects on N economy are not unequivocal, particularly in perennial grasses (Miller et al., 2002; Reynolds et al., 2005), and certainly require further study. In any case, no evidence was apparent for beneficial AMF effects on gross photosynthesis rate per unit nutrient (P or N) (Table 2), confirming former results (Grimoldi et al., 2005; Kavanováet al., 2006) indicating that AMF effects on perennial ryegrass are largely dependent on changes in nutritional status.

Allocation of new C and sources supplying belowground respiration

The total amount of new C allocated belowground was increased by the presence of AMF in otherwise very similar perennial ryegrass plants (Fig. 3). This result, indicating AMF as a competitive sink, is even more remarkable in the light of the somewhat lower gross photosynthesis rate of mycorrhizal plants. None of this extra amount of new C allocated belowground was, in fact, retained in root biomass. This result agrees with observations of Wright et al. (1998b). The AMF increased partitioning of new C to respiration within the root system, as suggested previously for mycorrhizal barley by interpretation of enzyme analyses (Müller et al., 1999). These results highlight the relevance of AMF as a C sink, indicating that AMF activity affected the distribution of newly available C within the plant.

Recent studies found that AMF provides a rapid pathway of C flux from plants back to the atmosphere (Johnson et al., 2002; Staddon et al., 2003; Heinemeyer et al., 2006), which led Johnson et al. (2002) to suggest that AMF primarily use current plant photoassimilates. Our work adds some new information to this point. Despite the higher allocation of new C belowground and to respiration, the fraction of new C in respired CO2 was actually lower in mycorrhizal roots (Fig. 4). Thus there was a differential use of new and old C sources in respiration, but in the opposite direction to that suggested by Johnson et al. (2002). Although the total flux of new C respired belowground was higher in mycorrhizal plants (because of the much higher amount of total respiration), the lower proportion of new C in respired CO2 indicates either a greater importance of plant C stores in supplying the respiration process, and/or the existence (and involvement in respiration) of storage pools within AMF tissues. Note that this finding is not directly related to the turnover rate of the whole fungus biomass, as it refers only to the importance of storage pools in supplying respiration. Evidence for kinetically different C pools supplying respiration has been reported previously in plants (Farrar & Jones, 2000; Schnyder et al., 2003). The higher contribution of a longer-term pool in mycorrhizal roots may result from the use of lipids (a major component of fungus metabolism) as a respiratory substrate.

Conclusions

The AMF (G. hoi) colonization of otherwise very similar perennial ryegrass plants enhanced the relative respiration rate of the root + soil system by 16%, and C flow to the AMF represented 3% to a maximum 8% of daily gross photosynthesis. This was closely associated with a higher amount of new C allocated belowground and respired in mycorrhizal plants. AMF colonization affected the sources supplying belowground respiration, indicating a greater importance of plant C stores in supplying respiration and/or the participation of storage pools within fungal tissues.

Acknowledgements

We especially thank Wolfgang Feneis for assistance with gas-exchange measurements, and Anja Schmidt and Monika Breitsamter for chemical analyses. Angela Ernst-Schwärzli, Erna Eschenbach, Melitta Sternkopf and all members of the Lehrstuhl für Grünlandlehre (Technische Universität München) provided invaluable assistance. Comments from Christoph Lehmeier, Katja Klumpp and Alastair Fitter helped in early stages of this work. We also thank Andreas Heinemeyer (University of York, UK) for kindly providing AMF inoculum. This study was supported by the Deutsche Forschungsgemeinschaft (SFB 607).

Ancillary

Advertisement