OUP user menu

Carbon dynamics in mycorrhizal symbioses is linked to carbon costs and phosphorus benefits

Pål Axel Olsson , Jannice Rahm , Nasser Aliasgharzad
DOI: http://dx.doi.org/10.1111/j.1574-6941.2009.00833.x 125-131 First published online: 1 April 2010


The nutrient and carbon (C) allocation dynamics in mycorrhizal hyphal networks cause variation in costs and benefits for individual plants and fungi and influence the productivity, diversity and C cycling in ecosystems. We manipulated light and phosphorus (P) availability in a pot experiment with Trifolium subterraneum colonised by the arbuscular mycorrhizal (AM) fungus Glomus intraradices. Stable 13C-labelling was used to trace assimilated CO2 to the mycorrhizal fungus in roots and soil using compound-specific isotope ratio mass spectrometry. We used the neutral lipid fatty acid 16:1ω5 as a signature for AM fungal storage lipids. Both P and shading reduced the AM fungal lipid accumulation in the intraradical mycelium, while only P reduced the amount of lipids in the extraradical mycelium. Recently assimilated plant C was only allocated to the mycorrhizal fungus to a small extent when plant mycorrhizal benefit was reduced by P fertilization, while increasing the plant C cost by shading did not reduce the C flow to the fungus. These results are of importance for our conception of mycorrhizal dynamics during periods of shade in nature.

  • arbuscular mycorrhiza (AM)
  • carbon-13
  • Glomus intraradices
  • mycorrhizal networks
  • shading
  • phosphorus
  • stable isotopes
  • symbiotic costs


The outcome of mycorrhizal symbiosis varies in terms of the benefits to plant hosts (Johnson et al., 1997; Klironomos, 2003). Reduced diversity of mycorrhizal fungi in agricultural fields (Helgason et al., 1998) may result from selection for a few, less beneficial, fungi (Johnson et al., 1992; Johnson, 1993), and fertilization may convert a mutualistic fungus into a parasite (Johnson et al., 1997). Mycorrhizal symbiosis has often been observed to depress plant growth, and this is believed to be particularly important in carbon (C)-limited systems (Hayman, 1974; Peng et al., 1993). The dynamics of mycorrhizal symbiosis may be important for the maintenance of diversity and productivity in plant communities (Bever, 1999; Maherali & Klironomos, 2007; Van der Heijden et al., 2008).

Mycorrhizal fungi influence the flow and retention of C in the rhizosphere by increasing the root allocation, speeding up respiration and retaining C in the extraradical mycelium (Jakobsen & Rosendahl, 1990; Staddon et al., 2003; Olsson & Johnson, 2005). Furthermore, it has been suggested that AM fungal networks may transport C between plants linked to a common mycorrhizal network (Grime et al., 1987), although this is debated, and others have reported that fungal translocation of C to sink plants, stay in the intraradical mycelium (Robinson & Fitter, 1999; Nakano-Hylander & Olsson, 2007; Voets et al., 2008). Still, such translocation may be of importance for reducing the C cost of some mycorrhizal plants in a community.

Arbuscular mycorrhizal (AM) fungi store large amounts of C as triacylglycerides. These compounds are synthesized only within the intraradical mycelium (mycelium inside roots) and are then transported to the extraradical (outside roots) mycelium (Bago et al., 2002). The production of fungal triacylglycerides constitutes a substantial C sink for the host plants (Bago et al., 2002). Triacylglyceride synthesis can be estimated using the major fatty acid in these lipids, 16:1ω5 (Olsson et al., 2005). 13C-labelling followed by compound-specific isotope ratio mass spectrometry (IRMS) can be used to trace the allocation of C specifically to the AM fungus in complex matrices such as soil and roots (Olsson et al., 2005).

We conducted two experiments to investigate how C allocation to intra- and extraradical mycelium is influenced by C and phosphorus (P) availability to the plant host. By increasing the soil P availability the plant benefit of mycorrhizal P uptake is expected to be reduced (Pacovsky et al., 1986) and by reducing the light intensity the C cost of the symbiosis is expected to increase since the mycorrhizal activities will consume a higher proportion of the C-flow into shaded plants (Peng et al., 1993; Johnson et al., 1997). We manipulated light and soil-P in Trifolium subterraneum (clover) colonised by the relatively P-tolerant mycorrhizal fungus Glomus intraradices (Douds & Schenck, 1990). The experiments were performed under conditions that have previously been shown to provide little benefit, and sometimes mycorrhizal-induced growth reduction in host plants (Söderberg et al., 2002). We tested the hypothesis that when plants in established mycorrhizal symbiosis experience decreased mycorrhizal benefit (through P addition) or increased mycorrhizal cost (through shading) the fungal symbiont will experience decreased availability of plant assimilated C.

Materials and methods

P-fertilization (experiment 1)

Clover (T. subterraneum L.) seeds were surface sterilized in 5% calcium hypochlorite for 15 min and pregerminated in aerated sterile water for three days. Seedlings were then planted in 48 pots (two seedlings per pot) with 230 g of a 1 : 1 w/w mixture of washed quartz sand (fraction <0.55 mm). The field soil had been γ-irradiated (10 kGy) in order to eliminate indigenous AM fungi and had properties as described by Gavito & Olsson (2003). Mild radiation has been shown to have little influence on soil P availability (Jakobsen & Andersen, 1982). The field soil had a pH of 6.6, contained 5.8% organic matter and 16 mg kg−1 available P (Olsen-P). Half of the pots were inoculated with a 3 cm2 phytagel disc of G. intraradices propagated in vitro (Schenck & Smith, BEG 87) on Ri T-DNA-transformed carrot root (AM), and the rest were left uninoculated (NM). All pots were inoculated with a bacterial inoculum as described below.

The plants were allowed to grow in a greenhouse at temperatures of approximately 22 °C during the daytime and approximately 18 °C at night-time. The photosynthetic photon flux density was at least 270 μmol m−2 s−1 and the photoperiod 18 h. The light intensity was kept constant through automatic sunshine protection during sunny days. The plants were watered daily with distilled water to 60% water-holding capacity. An overview of the experimental set-ups is given in Table 1. On days 24, 31 and 36, 5 mg of nitrogen (N) was added to each pot using a NH4NO3 solution with a concentration of 28.6 g L−1. At 44 days after sowing, 22.5 mg P was added to 24 pots (P), 12 AM and 12 NM using a KH2PO4 solution with a concentration of 44 g L−1. This corresponds to 90 mg P kg−1 soil. The pots were arranged in a factorial randomized block design consisting of two factors (AM inoculation and P fertilization) with 12 replicates for each treatment.

View this table:

The designs of the two experiments with nonmycorrhizal and AM-inoculated Trifolium subterraneum

TimeExperiment 1Experiment 2
Day 1Planting of seedlings half of pots inoculated with Glomus intraradices
Days 20–36All pots fertilized three times with 5 mg N
Day 37Shading treatment to half of the pots
Days 43–44P amendment to half of the pots
Days 44+3Harvest 1
Days 44+7Harvest 2Harvest
Days 44+14Harvest 2
  • Both experiments had a full factorial design: Experiment 1 with AM inoculation, P fertilization and Harvest time as factors (n=4), and Experiment 2 with AM inoculation, Shading and P fertilization as factors (n=5).

The contents of the pots were harvested 3, 7 and 14 days after P fertilization (four replicates in each harvest). The shoots were dried at 40 °C for 4 days and their dry weight noted; they were then subjected to Kjeldahl digestion with 4 mL concentrated H2SO4 in the presence of a catalytic mixture (K2SO4 and CuSO4× 5H2O), which was boiled at 350 °C for 2 h. Ammonium from the digested samples were then determined using flow injection analysis according to Lima et al.. (1999) and phosphate with a flow injection application of the stannous chloride-molybdate method. The soil was removed from the roots and a sample was taken and frozen at −20 °C for later fatty acid analysis. The roots were washed and weighed and were then divided into two fractions, one for fatty acid analysis, which was frozen at −20 °C, and one for mycorrhizal colonization measurements, which was fixed in 50% ethanol for later staining.

Light-manipulation (experiment 2)

Seedlings of T. subterraneum were pregerminated in Petri dishes overnight and planted in 40 pots containing 230 g of a 1 : 1 w/w mixture of washed quartz sand (fraction <0.55 mm) and γ-irradiated field soil with the properties described above. Twenty of the pots were inoculated with 20 g G. intraradices (BEG 87) inoculum and 20 g of the sand:soil mixture was added to the other 20 pots. All pots were provided with a bacterial inoculum as described below.

These plants were also grown in a greenhouse with a daytime temperature of approximately 22 °C and a night-time temperature of approximately 18 °C. The photosynthetic photon flux density and photoperiod were as described above. The plants were watered with distilled water to 60% of water-holding capacity. On days 20, 27 and 30, 5 mg of N was added to each pot using a NH4NO3 solution with a concentration of 28.6 g L−1. Thirty-seven days after sowing, 20 of the 40 pots were shaded with a net to 25% of the photosynthetic photon flux density of the control pots (Table 1).

P was added to 20 pots on day 43 (22.5 mg, corresponding to 90 mg P kg−1 soil) as described above. Ten pots were kept under normal light conditions without extra P (controls); 10 were given extra P; 10 were shaded, and the remaining 10 pots were given extra P and shaded. In half the number of pots of each treatment, the soil had been inoculated with the AM fungus G. intraradices. There were thus five replicates of each treatment. The light saturation point of T. subterraneum is 650 μmol m−2 s−1 (Tester et al., 1986) and the light reduction was expected to cause growth reduction in mycorrhizal plants (Söderberg et al., 2002) and reduced root allocation (Tester et al., 1986).

After 43 days (see Table 1), the plants were transferred to a greenhouse seedling propagation box and were labelled with 13CO2 under normal light conditions (Olsson et al., 2005). The pots were divided into two groups of 20 pots each, both containing at least two replicates of each treatment. The box was then sealed with paraffin to make it airtight. The lid of the box was equipped with a fan, a thermometer and a gas-sampling tube connected to an infrared gas analyzer. The CO2 concentration inside the box was recorded, and 125 mL 13CO2 was injected through a septum in the lid using a gas-tight syringe. The CO2 concentration in the box increased from 100–190 to 430–500 μL L−1 after injection, and the labelling period was 2 h (during the middle of the day, between 10:00 and 14:00 hours), by which time the CO2 concentration was the same as, or below, the initial level. After labelling, the pots were returned to the same conditions as before for 7 days, which was the chase period for the 13C-labelling. This time was chosen due to earlier experiences with the same plant–fungus combination when 13C enrichment peaked at 7 days (Nakano-Hylander & Olsson, 2007, fig. 5a).

On day 50, the contents of the pots were harvested. The shoots were cut, their fresh weights were noted and they were then frozen until further analysis. The soil was removed from the roots and a soil sample was taken and frozen for later fatty acid analysis. The roots were washed and weighed and were then divided into two fractions, one for fatty acid analysis, which was frozen, and one for mycorrhizal colonization measurements, which was processed immediately.

Preparation of bacterial inoculum

Field soil (200 g) was mixed with 1000 mL sterile distilled water and stirred for 1 h. The soil suspension was then passed through filter paper (Munktell no. 3). All pots received 5 mL of the soil filtrate (Van Aarle & Olsson, 2003).

AM fungal root colonization assessment

Roots were washed twice with distilled water, cleared in 10% KOH at 90 °C for 30 min. After rinsing in water, roots were stained at room temperature in 0.1% Trypan blue in lactic acid, glycerol and water (1 : 1 : 1, v/v/v), rinsed in water and destained in 50% glycerol. The magnified intersections method, as described by McGonigle et al.. (1990), was used to determine the total proportion of the root length with AM fungal root colonization and the frequency of AM fungal arbuscules and vesicles.

Lipid analysis

Neutral lipids are storage compounds that may comprise up to 20% of the biomass of hyphae, vesicles and spores of AM fungi (Olsson & Johansen, 2000). The neutral lipid fatty acid (NLFA) 16:1ω5 is an effective signature for AM fungal lipids (Van Aarle & Olsson, 2003) and the C-flow to this NLFA can be used to track the C-flow to the AM fungi in soil and roots (Olsson et al., 2005).

Roots were freeze-dried and then ball-milled in stainless steel beakers. The lipids from roots (30 mg dry mass) and soil (3 g wet weight) were extracted and quantified as described by Olsson et al.. (1997). Briefly, samples were vortexed (1 min) in a one-phase mixture of citrate buffer, methanol and chloroform (0.8 : 2 : 1, v/v/v, pH 4.0). The lipids were fractioned into neutral lipids, glycolipids and phospholipids on prepacked silica columns (100 mg sorbent mass, Varian Medical Systems, Palo Alto, CA) by eluting with chloroform, acetone and methanol, respectively. The fatty acid residues of the lipids were then subjected to mild alkaline methanolysis, to transform the fatty acids in the neutral lipids and phospholipids into free fatty acid methyl esters. Identification and quantification were performed using gas chromatography (Hewlett Packard 5890, 50 m HP5 capillary column, Palo Alto, CA, and H2 as carrier gas) and measuring the retention times in comparison with that of an internal standard (fatty acid methyl ester 19:0).

13 CO2 enrichment in crude tissue samples and fatty acids

Approximately 200 μg ball-milled freeze-dried shoot material was enclosed in tin capsules and analysed by continuous-flow IRMS using an ANCA-NT-20-20 Stable Isotope Analyser interfaced to a solid/liquid preparation module (PDZ Europa Scientific Instruments, Crewe, UK). The 13C/12C ratios of the CO2 in the combusted samples (total C) were determined with 0.01% precision. The 13C abundance in the NLFA 16:1ω5 was determined using the isotope analyser interfaced to a gas chromatograph (6890, Hewlett-Packard), which was equipped with a 50-m column (HP-5, Agilent, Palo Alto, CA) using He as carrier gas. 13C enrichment (atom %13C excess) was calculated as described by Olsson et al.. (2005).


Neither P amendment nor AM fungal inoculation influenced shoot biomass significantly in the first experiment (Table 2). Shoot biomass increased with time, but there was only a weak tendency for a positive effect of P fertilization in the nonmycorrhizal plants. Shoot P concentration was increased by mycorrhizal inoculation of plants to which no P had been added, but not in plants fertilized with P (Table 2). N concentrations in plants were not influenced by AM colonization, but increased with P fertilization. The positive effect of P fertilization on N concentration was only found for nonmycorrhizal plants, as seen from a significant AM × P interaction (Table 2). We found that P fertilization resulted in lower amount of NLFA 16:1ω5 in roots within only 3 days, and in the soil mycelium within 7 days, compared with the control (Fig. 1). After 14 days the effect of P fertilization was strongest on the NLFA 16:1ω5 in soil, reflecting the extraradical mycelium. P addition stopped the increase in vesicle colonization rate between 7 and 14 days after fertilization (Fig. 1), but it did not influence total AM fungal colonization (Table 3) or arbuscule formation. The increase in vesicle colonization occurred later than the increase in NLFA 16:1ω5 in the roots.

View this table:

Shoot biomass, P and N concentrations of Trifolium subterraneum growing with (AM) or without (NM) mycorrhizal inoculation

TimeDry shoot舲mass (mg)P concentration舲(mg g−1)N concentration舲(mg g−1)
C10.53 ± 0.0310.12 ± 0.0112.5 ± 0.12
20.63 ± 0.0500.11 ± 0.0082.2 ± 0.13
30.79 ± 0.0570.12 ± 0.0091.9 ± 0.08
P10.52 ± 0.0120.26 ± 0.0103.1 ± 0.07
20.65 ± 0.0370.33 ± 0.0152.7 ± 0.13
30.86 ± 0.0210.35 ± 0.0292.2 ± 0.10
C10.54 ± 0.0210.24 ± 0.0132.9 ± 0.15
20.67 ± 0.0160.24 ± 0.0152.2 ± 0.12
30.82 ± 0.0160.24 ± 0.0202.1 ± 0.05
P10.58 ± 0.0110.27 ± 0.0132.7 ± 0.08
20.66 ± 0.0210.31 ± 0.0152.3 ± 0.09
30.80 ± 0.0220.35 ± 0.0202.1 ± 0.05
PNSP<0.001P< 0.01
TP< 0.001P<0.001P< 0.001
AM × PNSP< 0.001P< 0.001
P × TNSP<0.001NS
  • Half of the pots had been P fertilized, while the other half was not fertilized (C) and they were harvested at three different times (harvest times 1–3; see Table 1). Values are means (± SE) and the level of significance was calculated with a three-factor anova.


Development of vesicles and the AM fungal signature, NLFA 16:1ω5, following P fertilization of 44-day-old Trifolium subterraneum plants inoculated with Glomus intraradices. The background of NLFA 16:1ω5 in nonmycorrhizal roots was 0.020 nmol mg−1 and in soil with nonmycorrhizal plants 0.044 nmol g−1. P had a significant effect on NLFA 16:1ω5 in roots and soil (P<0.001, two-way anova), while harvest time had no effect. Shading and P together had no effect on the 16:1ω5 concentration. Neither P nor time alone had any effect on vesicle incidence, but a significant P × time interaction was seen (P<0.01).

View this table:

Development of AM fungal root colonisation following P fertilization of Trifolium subterraneum colonised by the mycorrhizal fungus Glomus intraradices

Experiment 1Experiment 2
TreatmentTime舲(days)AM舲colonization (%)TreatmentAM舲colonization (%)
C344 ± 3.4C77 ± 3.1
754 ± 2.7P52 ± 6.2
1459 ± 6.7S42 ± 4.7
P352 ± 6.3PS44 ± 5.4
759 ± 3.6
1451 ± 5.8
anovaPNSPP< 0.05
TNSSP< 0.001
P × SNSP × SP< 0.05
  • In Experiment 1 (n=4) we tested the effect of P fertilization (P) with time (T) after amendment and in Experiment 2 (n=5) the effects of P fertilization and shading (S). Means (± SE) are given as well as the P values following two-way anova.

In the second experiment, mycorrhizal colonization significantly reduced the growth of clover, (Fig. 2) and shading significantly reduced clover shoot weight, regardless of P treatment. The addition of P did not influence the growth rate significantly. A significant AM × P interaction was found; P addition stimulated nonmycorrhizal plants, but not mycorrhizal plants.


Effects of mycorrhizal inoculation, P fertilization (P) and shading (S) on the dry weight of Trifolium subterraneum shoots. Controls (C) were neither shaded nor fertilized. Three-way anova revealed significant effects of the AM fungus (P<0.001) and shading (P<0.001), and a significant AM × P effect (P<0.05). P fertilization alone had no significant effect, and no other significant interactions were observed.

Mycorrhizal colonization of roots ranged from 77% in the control plants (normal light conditions) to 42% in the shaded plants (Table 3). P reduced the colonization rate, but only in plants under normal light (a significant shade × P interaction). Mycorrhizal colonization of the clover plants increased the AM fungal NLFA 16:1ω5 from 0.009 nmol mg−1 to between 1 and 9 nmol mg−1, compared with nonmycorrhizal controls. Mycorrhizal colonization, as indicated by the NLFA 16:1ω5, was reduced by both shading and P fertilization (Fig. 3a). The lipid accumulation in the soil mycelium was strongly reduced by P fertilization, while shading did not influence the accumulation of fungal lipids (Fig. 3b).


Influence of P fertilization (P) and shading (S) on the amount of NLFA 16:1ω5 in roots (a) and soil (b) (values from nonmycorrhizal treatments were subtracted), and the 13C enrichment in 16 : 1ω5 in roots (c) and soil (d). Controls (C) were neither shaded nor fertilized. The 13C enrichment was calculated as the %13C above the background level after a 6-day chase period. (a) P<0.001 for P fertilization, <0.001 for shading and <0.01 for the P × shading interaction; (b) P<0.05 for P fertilization; neither shading nor interactions had any significant effect; (c) P<0.001 for shading, neither P nor interactions had any significant effect; (d) no significant effects.

There was no effect of AM inoculation on 13C enrichment in shoots or roots of T. subterraneum. The 13C enrichment in shoots was increased by shading, and particularly so in P fertilized plants (Table 4). Shading also increased the 13C enrichment in roots, while P alone had no effect on 13C enrichment in shoots or roots. There was also a significant P × Shading interaction for 13C enrichment in shoots resulting from the fact that P increased 13C enrichment in shaded plants, but not in plants under normal light conditions.

View this table:

Influence of P fertilization (P) and shading (S) on the 13C enrichment in shoots and roots of Trifolium subterraneum colonised (AM) or not (NM) by the mycorrhizal fungus Glomus intraradices

Treatment13C enrichment (%)
C0.62 ± 0.150.35 ± 0.048
P0.38 ± 0.0880.38 ± 0.025
S0.43 ± 0.180.59 ± 0.14
PS1.1 ± 0.510.65 ± 0.086
C0.44 ± 0.160.43 ± 0.045
P0.45 ± 0.140.42 ± 0.036
S1.0 ± 0.190.53 ± 0.048
PS1.5 ± 0.460.58 ± 0.082
SP< 0.01P< 0.001
P × SP< 0.05NS
  • Controls (C) were neither shaded nor fertilized. The 13C enrichment was calculated as the %13C above the background level after a 7-day chase period. Means are given (± SE) and P values from three-way anova.

Shading resulted in a much higher rate of 13C enrichment by the fungal lipids inside the roots than under normal light conditions (Fig. 3c). The 13C enrichment in the soil mycelium after 13C-labelling was not significantly influenced by either P or shading (Fig. 3d).

We found no significant effect of P or shading on total retention of labelled C in the plant shoots although a decreasing effect was indicated (Fig. 4a). P fertilization under shaded conditions, on the other hand, increased the uptake in shoots (Fig. 4a). The transfer of C by plants to AM fungi was considerably reduced by P fertilization, particularly under normal light conditions (Fig. 4b). Shading, on the other hand, increased the C allocation to the fungus under fertilized conditions and it seemed as if P fertilization had no negative effect on C retention in the AM fungal mycelium under shaded conditions.


Effects of mycorrhizal colonization (AM) P fertilization (P) and shading (S) on C retention in (a) shoots of Trifolium subterraneum and (b) the AM fungus Glomus intraradices. Controls (C) were neither shaded nor fertilized. The total C uptake in plant shoots and in the AM fungal signature NLFA 16:1ω5 was calculated by multiplying the enrichment by the amount. Neither P nor shading had any effect on C retention in T. subterraneum, but P (P<0.01) and the P × shade interaction (P<0.05) had significant effects on C retention in G. intraradices.


The allocation of C to the mycorrhizal fungus was strongly reduced by P fertilization under good light conditions. When shading induced C limitation there was no reduction in the amount of C retained in the mycorrhizal mycelium. It has earlier been shown that T. subterraneum allocate much less resources to roots under shaded conditions (Tester et al., 1986) and we can show here that this does not reduce the C allocated to mycorrhizal structures, although the mycorrhizal colonization may decrease, as observed by Tester et al.. (1986) and this study. Mycorrhizal networks are extensive (Whitfield, 2007), and our results can help in understanding the source–sink dynamics in these networks. Several lines of evidence suggest that the plants in this study were C-limited: mycorrhizal colonization reduced plant growth, phosphorus fertilization did not increase plant growth, and shading reduced plant growth. This C limitation did not reduce colonization by the mycorrhizal fungus G. intraradices, a fungus known to be tolerant to high P levels (Douds & Schenck, 1990), which may indicate that it is able to act as a parasite under certain conditions. Johnson et al.. (1997) suggested that either fertilization (reduced benefits) or shading (increased costs) could generate mycorrhizal parasitism. The results of our study indicate that plants can control C allocation to their associated fungus when the soil is enriched with P, but not when light is limited. However, P fertilization did not reduce C allocation to the AM fungus in shaded plants as shown by a significant P × shade interaction. This could be an effect of higher total C uptake in those plants (Table 1), but it also emphasizes the ability of AM fungi to obtain C from shaded plants. When we transplanted T. subterraneum to the shaded conditions we also moved it to conditions far below its saturation point, which has been shown to be 650 μmol m−2 s−1 (Tester et al., 1986). One reason for the lack of a negative P effect in shaded plants could be that they did not have the energy requirement binding further C.

The reduced AM colonization in shaded and P fertilized plants was as could be expected from earlier studies (Tester et al., 1986; Bruce et al., 1994; Olsson et al., 1997). Here we show that these effects can be detected as rapidly as within 1–2 weeks after environmental change. Usually structures do not disappear, but root growth without new colonization will decrease the colonization rate in roots. The low colonization in the shaded plants did not prevent a high mycorrhizal C allocation. Therefore the reduced mycorrhizal C allocation in P fertilized plants could be due not to reduced colonization rate, but instead to specific regulation of C-flow over the mycorrhizal interface.

We have shown that parasitic behavior can be independent of P fertilization, and that perhaps shaded plants are more likely to be subject to the parasitic action of AM fungi. Fitter (2006) suggested that plants allocate C to places in the root where there is an in-flow of P, a mechanism that would prevent the plants from being parasitized by cheating fungi. The reduced mycorrhizal C allocation in P-fertilized plants under normal light conditions is in line with this hypothesis. Still, the mycorrhizal plants had a higher P content than nonmycorrhizal ones. Indeed, it is well-known that AM fungi may contribute significantly to plant P uptake without increasing the growth (Smith et al., 2003). Therefore, it could be that both AM inoculation and P fertilization may in the long run increase fitness even when no growth response was found during the time-span of this experiment and in particular during the short time-span following P fertilization. P-mediated benefits in the long run could be due to increased seed quality or pathogen resistance (Newsham et al., 1995). Such benefits from a better P status could be one reason for the relatively high C allocation to the fungus in spite of a lack of an obvious P limitation in the plants in this study.

The reduction in colonization often observed following P fertilization (see e.g. Bruce et al., 1994) has been explained by a reduction in germ tube formation by AM fungal spores, reduced exudation of branching factors by roots, and by faster root growth (Smith & Read, 2008). None of these factors were, however, possible in our study since the AM colonization was already established when P fertilization was applied, and because we saw no growth stimulation following P addition. Instead, we found that P fertilization reduced the flow of C to the fungus indicating that a mechanism in the carbohydrate transfer system was involved.

In some systems, indications of plant-to-plant transfer of C via the mycorrhizal network have been observed, perhaps toward shaded plants (Francis & Read, 1984; Grime et al., 1987; Simard et al., 1997). Our results, showing that extraradical AM fungal mycelium connected to a shaded plant has a high C sink capacity, support the suggestion that the transfer of C is unlikely in many species (Robinson & Fitter, 1999; Zabinski et al., 2002; Voets et al., 2008), but still certain species may have special adaptations to achieve plant-assimilated C (Cameron et al., 2008). The significant allocation of C to extraradical mycelium in the shaded plants may be a strategy that has evolved to find new hosts when the C allocation capacity of the current host is reduced. Such a strategy may be important when vernal flora is shaded later in the season by plants of other functional types. This kind of C flow may also explain the finding that seedlings connected to mature plants with a common mycorrhizal network benefit when mature plants are defoliated (Pietikäinen & Kytöviita, 2007).

The fact that C was reallocated to extraradical mycelium in shaded hosts may lead to more aggressive colonization of other plants. The more rapid turnover in soil mycelium could also lead to a reduction in C flow to other plants, but the possibility of fungus–plant C transfer is still under debate. We propose that the high C sink capacity of certain fungi may draw C from shaded plants, thus regulating the community structure in shaded environments.

Although vesicles are lipid storage organs, their formation is not correlated with lipid accumulation over time. Earlier results indicate that vesicles are formed first, and then later filled with lipids (Van Aarle & Olsson, 2003). The vesicles also showed a rather late response to the high P fertilization in the current study, while the lipid indicators reacted much faster. Overall, our results showed that using 13C-labelling of indicator lipids is an efficient way of tracing C flows under symbiotic conditions.

We have proposed a mechanism for soil C redistribution within mycorrhizal networks in response to changes in the availability of resources that alter the costs and benefits of the symbiosis. This mechanism should be considered in the modelling of nutrient exchange in mycorrhizal symbiosis (Fitter, 2006; Landis & Fraser, 2007). These results are also of importance for our conception of mycorrhizal dynamics during periods of shade. A change in climate may affect the allocation of C in the fungal mycelia that colonize plants. C may leave the mycorrhizal structures and return rapidly to the atmosphere (Johnson et al., 2002; Staddon et al., 2003), or may be used to form stable structures such as spores (Zhu & Miller, 2003; Olsson & Johnson, 2005). We used a single mycorrhizal fungus, which is believed to be P tolerant. The effects on other fungi may be even more drastic. It is clear that P fertilization may reduce the C pathway via mycorrhizal fungi, and that this may lead to the loss of C from ecosystems due to higher respiration rates and less formation of stable mycorrhizal C pools (Zhu & Miller, 2003; Wilson et al., 2009).


Financial support from FORMAS and The Swedish Research Council is gratefully acknowledged. We thank Nancy Johnson and Ylva Lekberg for valuable comments on the manuscript.


  • Editor: Philippe Lemanceau


View Abstract