Novel in-growth core system enables functional studies of grassland mycorrhizal mycelial networks


Author for correspondence: D. Johnson Tel: +44 114 222 0095 Fax: +44 114 222 0002 Email:


  •  A novel in-growth core system, enabling functional studies of natural communities of arbuscular mycorrhizal (AM) mycelia in soil is described and tested.
  •  The cores have windows covered with nylon mesh of 35 µm pore size that prevent in-growth of roots but permit penetration of AM hyphae. They were inserted into grassland turf and contained either sterilized sand and a ‘bait’ seedling of Trifolium repens or nonsterile natural soil without bait plants. The impacts of hyphal severance, achieved by periodic rotation of some of the cores, upon AM colonization of bait plants (experiment 1) and transfer of 33P from soil to plants outside the cores (experiment 2) were examined.
  •  Severance of AM hyphae reduced both AM colonization of bait plants and their shoot P concentrations. The shoot 33P concentrations of plants with mycelial access to 33PO4-labelled cores were 10-fold greater than those which had no mycelial access.
  •  It is concluded that this novel approach enables the functioning of mycorrhizal mycelial networks to be evaluated under conditions closely simulating those occurring in nature.


The roots of the majority of plant species found in temperate grasslands form arbuscular-mycorrhizal (AM) associations (Read et al., 1976; Sparling & Tinker, 1978). The mycelial systems produced by AM fungi form extensive networks with hyphal lengths reaching up to several m cm−3 in undisturbed temperate grassland soils (Miller et al., 1995). These provide the major nutrient–absorbing interface between plants and soil, and form important pathways for the rapid transport of nutrients to roots (Smith & Read, 1997). In addition to their roles in nutrient absorption and transport, the external mycelial systems of AM fungi are the primary sources of inocula which enable the spread of infection between root systems in permanent grasslands (Read & Birch, 1988).

Because external mycelial systems are less amenable to study than the colonized roots themselves, they have received little attention until recently. However, now there is emerging evidence that there are marked differences between species of AM fungi in the rates of growth and extent of their mycelial systems in soil and the effectiveness of these systems for nutrient (particularly phosphate) uptake and transport from different distances from plant roots (Jakobsen et al., 1992). Furthermore, most grassland communities contain a diverse assemblage of AM fungi (Johnson et al., 1992), and the extent of this diversity has been positively correlated with the overall effectiveness of phosphorus (P) uptake by plants with mycorrhizas (Van der Heijden et al., 1998).

These observations highlight the requirement to understand more fully the structure and functioning of the external mycelial systems of natural communities of AM fungi. Our present knowledge of these mycelia is mainly based on laboratory studies typically involving only single plant–fungal species associations and the fungi selected are predominantly those that are readily cultured from spores in sand or soil of low organic matter content (Johansen et al., 1993; Pearson & Jakobsen, 1993a,b). A major barrier to full understanding of the functioning of communities of AM fungal species in soil has been the requirement to eliminate mycorrhizal inoculum from control (nonmycorrhizal) treatments and to date this has only been achieved by sterilization using heat, irradiation or chemical treatments. Of the relatively few studies of AM mycelial functioning that have been undertaken under field conditions or in intact turfs, virtually all have either applied antifungal agents, such as benomyl, fosetyl-Al or phosphonate (Sukarno et al., 1998; Jakobsen et al., 2001) or broad-spectrum sterilants like methyl bromide (Jakobsen, 1986) to provide the required controls.

Treatments involving heat or irradiation sterilization of soil cause large releases of nutrients particularly from soils that are high in organic matter such as most permanent grasslands. Furthermore, these treatments can also generate phytotoxicity in the soil. As a consequence, there is an urgent need to develop more effective and less disruptive methods for controlling the presence and absence of mycorrhizal mycelial networks in soils in order to study their functioning particularly under field conditions.

An important development in studies of the functioning of AM mycelial systems has been the use of mesh barriers to provide root-free compartments into which mycorrhizal mycelium can grow (Francis & Read, 1984; Schüepp et al., 1992; Jakobsen et al., 2001). In the laboratory, mycorrhiza-free compartments have been developed using membranes with pore sizes that are too small to be penetrated by AM hyphae (e.g. 0.45 µm; Tarafdar & Marschner, 1994) but these membranes are very fragile and they have much lower porosity than the nylon meshes used to divide root-free hyphal compartments. Their use in the field is compromised not only because of their high failure rate, but also because membranes of such fine pore size can influence nutrient and water transport between soil on either side.

In this paper we describe a procedure in which the colonization of soil by AM mycelium in grassland turfs is manipulated by regular rotation of mesh-bound soil cores in order to break in-growing hyphae. Undisturbed static cores permit colonization by AM mycelium and so serve as controls. The effectiveness of this methodology is assessed using both bioassays in which bait plants are used to evaluate the impacts of the barrier treatments on the colonization process, and by quantification of mycorrhiza-dependent P transfer to plants in grassland turfs traced by the addition of 33PO4 to cores containing natural soil.

Materials and Methods

Site description

Large intact rectangular monoliths (45 × 35 × 30 cm) supporting a closed grassland turf were removed from control plots of three blocks at the NERC Soil Biodiversity Thematic Programme field site, Sourhope, Scotland (NGR: NT 854196), and were installed in free-draining plastic boxes at the University of Sheffield Experimental Gardens. The species-rich grassland comprises 24 higher-plant species, 21 of which typically form arbuscular mycorrhizas (Harley & Harley, 1987). It is classified as U4d (Festuca-Agrostis-Galium with a Luzula multiflora-Rhytidiadelphus loreus subcommunity) under the British National Vegetation Classification (Rodwell, 1992). The soils are developed on locally derived drift from andesitic lavas of old red sandstone age and are acid brown earths (pH 4.5–5.0) with an organic C content of 25%.

Design of cores

Cores were constructed using acrylonitrile butadiene styrene (ABS) water pipe (18 mm ID, 22 mm OD in 150 mm sections). Four ‘windows’ were cut into each core, the openings produced being equivalent to approx. 50% of the below ground external surface area of the core. Nylon mesh (Plastok Associates Ltd, Birkenhead, Wirral, UK) with a pore size of 35 µm was attached to the cores using Tensol no. 12 adhesive (Evode Speciality Adhesives Ltd, Leicester, UK) and the completed cores were cured at 80°C for 24 h. The mesh covered the windows and the base of the core (Fig. 1). Provision of a permeable base enabled free drainage of soil water. A set of cores designed to exclude in-growth of AM hyphae was constructed using identical tubing without windows.

Figure 1.

Diagram showing details of the mesh core systems (mesh is shown cut-away for clarity). Note that 33P was added to cores without clover bait plants.

Using these cores, two experiments were undertaken. The first was designed to investigate the impact of hyphal severance upon the process of colonization of bait plants grown inside the cores (experiment 1). The second evaluated the effects of severance upon the transfer of P from soil in plant-free cores to surrounding plants in an intact grassland turf (experiment 2).

Experiment 1: analysis of the effects of hyphal severance upon AM colonization of bait plants

Experimental design

The cores were filled with approx. 40 g f. wt of sieved (2 mm) and autoclaved (1 h at 125°C) Bunter sand (pH 5.2; derived from Triassic Bunter Sandstone) obtained from Blaxton Common, near Doncaster, England. This material was used since it has very low organic matter content and microbial biomass making it amenable to steam sterilization and a pH closely matching that of the soil into which the cores were inserted. A total of 42 cores (28 with mesh barriers, 14 solid) were installed in a random array of holes cut into three replicate turfs using a 20 mm diameter cork borer. Into half of the cores were placed, as bait plants, single 2 wk old seedlings of Trifolium repens L. which were previously propagated on autoclaved Bunter sand in a controlled environment growth room (20°C, 18 h day 15°C, 6 h night). A 15 cm clear acetate collar was wrapped around the neck of each core for the duration of the experiment in order to reduce the loss of bait plants through herbivory. Typically 5 replicate seedlings survived for the duration of the experiment. The plant-free cores were used to measure the length of extractable AM mycelium.

Half of the mesh cores were rotated approx. 45° around their vertical axes every week in order to break any hyphae penetrating the core. The experiment was maintained during the summer of 2000 (11th May–21st July) during which time the boxes remained out of doors where they were provided with supplementary water as required.

Assessment of AM colonization and nutrient status of bioassay plants

The bioassay plants were removed after 10 wk and their shoots were dried (80°C), weighed and coarsely ground. A c. 25 mg subsample was digested in 1 ml of a salicylic/sulphuric acid-lithium sulphate mix for 5 h at 350°C (Bremner & Mulvaney, 1982), diluted to 10 ml and analysed for total nitrogen (Scheiner, 1976) and P (John, 1970). The roots were washed and cut into approx. 1 cm lengths from which a subsample was removed and frozen. The remaining tissue was cleared (10% KOH at 80°C for 2 h), acidified (10% HCl), stained for 2 h (trypan blue in lactoglycerol), de-stained (50% glycerol) and scored for mycorrhizal colonization at × 300 magnification (McGonigle et al., 1990). Root lengths were measured by the grid intercept method (Tennant, 1975).

Impact of severance on AM hyphal lengths

Rotated and static sets of cores lacking bait plants were removed after 11 wk incubation in the turfs and hyphal lengths were measured in a 5 g subsample using the method of Brundrett et al. (1994).

Experiment 2: analysis of the effects of hyphal severance upon transfer of 33P from cores into grassland turfs

Experimental design

Blocks of turf were divided and each piece inserted into one of eight pots of 1 l capacity which were maintained in a controlled environment growth room for the duration (41 wk) of the experiment (20°C, 18 h day 15°C, 6 h night). Into each pot were inserted 2 mesh-bound cores filled with approx. 40 g f. wt sieved (2 mm) Sourhope soil, which were maintained plant free. The cores were left in position for 12 wk to enable in-growth of AM mycelium. After this period, the cores in four of the pots were rotated to break mycelial connections and a 250 µl aliquot of 33P-orthophosphoric acid with a specific activity of 5.776 GBq µg−1 was injected into the centre of each core (providing each with approximately 682 mBq) using a disposable 1 ml syringe. The solution was injected in such a way so as to release the isotope when the needle was being withdrawn from the soil, thus ensuring that it was distributed along the vertical axis of the core.

Half of the cores were undisturbed for the duration of the experiment. The remaining cores were rotated twice weekly until the second harvest (17 d after addition of the label). Thereafter rotation was stopped in order to allow hyphae to re-colonize the cores, until the fifth harvest (111 d after addition of the label), after which rotation was restarted twice weekly until the end of the experiment (201 d after addition of the label).

Analysis of 33P content in recovered plant shoots and soils

Live shoot biomass was harvested to 2 cm height at 3, 16, 40, 73, 110, 158 and 201 d after addition of the label. The plant material was oven-dried (80°C), weighed and coarsely ground before a c. 25 mg subsample was taken and analysed for total P as described previously. To 10 ml of scintillation fluid (Packard Scintillation Plus) was added 1 ml of the wet-ashed samples and the radioactivity determined in a Packard Tricarb 1600 TR liquid scintillation counter, and corrected for quench, decay and background.

The soil cores were removed at the final harvest and the water-extractable 33P determined by shaking a 10 g f. wt subsample with 20 ml distilled water for 45 min from which a 2 ml subsample was added to 10 ml scintillation fluid and measured as described. Total 33P was determined following digestion of 500 mg oven-dry soil as described for shoot P analysis.

Statistical analyses

Treatment effects were analysed by ANOVA and Tukey multiple comparison tests using Minitab 10.5. All data were log or arc-sine transformed to satisfy test assumptions where appropriate.


Experiment 1: analysis of the effects of hyphal severance upon AM colonization and nutrient status of bait plants and length of extractable hyphae

Mycorrhizal colonization of host roots was significantly lower (P < 0.001; F = 16.7; n= 5) in plants harvested from the rotated mesh cores compared to those harvested from the static mesh cores (Table 1). In the static cores, the mean root length colonized by AM was 55% compared with < 10% for the rotated cores. The control plants harvested from the solid cores remained non-mycorrhizal.

Table 1.  Percentage of root length colonized by AM fungi, vesicles and arbuscules, shoot P and N concentrations and N : P ratios of Trifolium repens bait plants grown in either solid or static and rotated mesh-bound cores inserted into intact grassland turfs (brackets indicate SE). Hyphae were extracted from identical cores that contained no bait plants
Barrier treatment% root length colonized by AM% root length occupied by arbuscules% root length occupied by vesiclesLength of extractable hyphae (cm g d. wt−1)Shoot P concentration (mg g−1 d. wt)Shoot N concentration (mg g−1 d. wt)N : P ratioShoot biomass (mg d. wt)
  1. Values sharing a letter are not significantly different (P > 0.05).

35 µm mesh static55.7a(10.2)12.1a(3.7)27.5a(10.2)9.1a(1.5)7.8a(2.74)24.2a(2.45) 6.3b(2.55) 9.0ab(2.9)
35 µm mesh rotated 9.5b(5.9) 0.58b(0.58) 5.4b(4.8)2.4b(0.6)3.1ab(0.44)24.4a(1.07) 8.7ab(1.28)18.6a(4.0)
Solid 0.0b(0.0) 0.0b(0.0) 0.0b(0.0)0.7b(0.08)1.3b(0.30)31.4a(3.91)28.1a(7.30) 4.6b(1.4)

The percentage root length occupied by vesicles declined significantly (P < 0.001; F = 19.3; n= 5) from 28% in plants in the static cores to 5% in the rotated mesh cores. Formation of arbuscules mirrored the pattern seen for vesicles but their occurrence ranged from 0 to 12% of root length colonized (Table 1). Significant reductions (P < 0.05; F = 9.6; n= 5) in arbuscule colonization were seen in the plants from the rotated and solid cores relative to the situation observed in the static cores.

Mean shoot P concentrations of bait plants ranged from 7.8 mg g−1 d. wt in the static mesh cores to 1.3 mg g−1 d. wt in the solid cores (Table 1). The mean shoot P concentrations in the plants from the rotated cores were 60% lower than those from the static cores although this was not significantly different. The variability associated with the shoot P concentrations of the plants from the static cores was greater (SE = 2.74 mg g−1 d. wt) than those from both the rotated (SE = 0.44) and solid (SE = 0.30) cores (Table 1). The shoot P concentrations in the nonmycorrhizal control plants in the solid cores were significantly lower than in the mycorrhizal plants from the static cores (P = 0.032; F = 4.9; n= 5). Shoot N concentrations ranged from 24 to 32 mg g−1 d. wt but were not significantly affected by either hyphal exclusion treatment. The shoot N : P ratio increased significantly (P = 0.011; F = 7.1; n= 5) from 6.3 in the static mesh cores to 28.1 in the solid cores, reflecting the decline in P concentrations. Mean shoot biomass ranged from 4.6 to 18.6 mg d. wt, the highest values being observed in the rotated cores which were significantly greater (P = 0.027; F = 5.1; n= 5) than those from solid cores.

The lengths of hyphae extracted from the cores were low in all three treatments, ranging from 0.7 cm cm−3 in the solid to 9.1 cm cm−3 in the static cores. Values generally followed a pattern similar to those seen with mycorrhizal colonization. Relative to the static mesh cores, hyphal lengths were significantly lower (P < 0.001; F = 27.3; n= 6) in the rotated mesh and solid cores (Table 1).

Experiment 2: effects of hyphal severance upon transfer of 33P from cores into grassland turfs

In the first harvest (3 d after addition of the label), the shoot 33P concentration was 10-fold higher (0.3 compared to 0.03 ng 33P g d. wt−1; P < 0.001; F = 34.4; n= 4) in plants removed from pots containing static cores compared to those containing rotated cores (Fig. 2). A similar difference (P = 0.002; F = 27.6; n= 4) was also measured in the second harvest (16 d after addition of the label) where shoot 33P concentrations were 6.3 and 0.7 ng g d. wt−1 in plants removed from pots containing static and rotated cores, respectively. The rate of 33P uptake declined after the third harvest in the pots containing static cores with shoot concentrations ranging from 7.8 to 9.5 ng 33P g d. wt−1.

Figure 2.

Concentrations (ng g d. wt−1) of 33P (± SE) in shoots removed sequentially from pots containing either static (closed circles) or rotated (open circles) cores inserted into grassland turfs injected with 33P orthophosphoric acid. Core rotation was stopped immediately after the second harvest (16 d after addition of the label) and re-started after the fifth (110 d after addition of the label). Asterisks indicate significant differences (*, P < 0.05; **, P < 0.01; ***, P < 0.001). NS, not significantly different.

When core rotation was stopped after the second harvest the shoot 33P concentrations increased rapidly. Forty days after addition of the label, the shoot 33P concentrations remained significantly (P = 0.021; F = 9.7; n= 4) lower (1.7 compared to 8.1 ng 33P g d. wt−1) in these plants than in those which had constant mycelial access to the cores, while 73 and 110 d after addition of the label the concentrations were not significantly different. Resumption of core rotation after the fifth harvest led to a further 11% reduction in shoot 33P concentration to 5.5 ng g d. wt−1 at the sixth harvest and a 37% reduction (P < 0.01) to 3.9 ng g d. wt−1 at the final harvest. In the final harvest, the concentrations were 56% lower (P < 0.01; F = 13.6; n= 4) in plants removed from pots containing static cores compared to those containing rotated cores.

The accumulation of 33P in plant shoots during the initial five harvests (0–110 d after addition of the label) followed a linear relationship (y = 0.46x + 0.085; r2 = 99.6; P < 0.001; F = 842; n= 4) in pots containing static cores, and an exponential relationship (y = 0.07e0.041x; r2 = 97.5%) in pots containing cores that were rotated in the first and second harvests but static thereafter (Fig. 3). Throughout the duration of the experiment, 16% and 8% of the source isotope was accumulated in plant shoots in pots containing the static and rotated cores, respectively. These differences were reflected by the static cores having residual total 33P concentrations that were 36% lower (P = 0.06) than those in the rotated cores (Fig. 4). The water-extractable 33P concentrations were approximately 2.5 ng 33P g d. wt−1 in both of the core treatments.

Figure 3.

Regression analysis of percentage of supplied 33P that was accumulated in plant shoots during 110 d following addition of the label to cores inserted into grassland turfs which were either permanently static (closed circles; y = 0.46x + 0.085; r2 = 99.6) or initially rotated (open circles; y = 0.07e0.041x; r2 = 97.5%. Cores were rotated twice weekly for 17 d after addition of the isotope and thereafter were undisturbed).

Figure 4.

Residual water-extractable and total concentrations of 33P (ng g d. wt−1) in static (open squares) and rotated (hatched squares) cores (± SE). Bars are not significantly different (P > 0.05).


We have described and tested a novel method enabling the control of mycorrhizal mycelial colonization to be achieved under field conditions, without the use of broad-spectrum chemical treatments or steam sterilization. The fall in the amounts of extractable external mycelium, mycorrhizal colonization (total, vesicular and arbuscular) and shoot P concentrations associated with severance of the hyphal connections into the cores provides a clear demonstration of the importance of the mycelial network for the formation and function of the symbiosis in soil. Although such links have been observed previously (Sanders et al., 1977; Merryweather & Fitter, 1995a,b; Jakobsen et al., 2001) the methodologies described here provide for the first time a simple mechanism enabling manipulation of mycorrhizal colonization without causing major alteration to soil properties. The shoot P concentration of the bait plants removed from the static cores were more variable than those from both the rotated and solid cores. A similarly large variation in the percentage root length occupied by arbuscules was also observed in these plants which may reflect differences in the P transfer characteristics of the fungi colonizing their roots.

If this approach is extended to studies other than simple bioassays of AM in-growth, a number of factors will need to be taken into consideration including the growth period and optimum duration of the core rotation to ensure bait plants remain uncolonized. Furthermore, their use in the field must be carefully evaluated on a site-specific basis to ensure that the contact of the mesh walls of the cores with the bulk soil is not affected by soil physical conditions, such as shrinkage and expansion cycles. However, it is clear that this approach has much potential for development as a relatively unintrusive method of controlling mycorrhizal colonization which is very amenable for use in highly replicated field experiments.

The core systems enabled us to investigate the role of external AM fungi in the uptake of P by plants growing in whole turf monoliths from the Sourhope system. The amount of P transferred into shoots from static cores was 10-fold that transferred from rotated cores. This provides striking evidence of the importance of the external mycelial network for P transfer in the upland grassland community from which the turf was obtained. The difference in shoot P concentrations between plants with access to either the static or rotated cores was consistent at both of the initial harvests (3 and 16 d after addition of the 33P). The subsequent cessation and continuation of core rotation provided additional information on the time dynamics of P transfer from uncolonized patches of soil. We measured an exponential increase in the proportion of source isotope transferred into shoot tissue during the 110 d period in which hyphae were permitted to recolonize the (initially) rotated cores. Johansen et al. (1993) demonstrated that maximum hyphal 32P uptake and subsequent transport to host shoots occurred 0–7 d after addition of the isotope 2 cm distant from the roots of 42 d-old T. subterraneum plants mycorrhizal with Glomus intraradices. The more rapid uptake reported by Johansen et al. (1993) is likely to be influenced, at least in part, by their use of a simplified model system comprising clay/sand growth medium and single species of both host and symbiont. The data reported in the present study using the novel core systems are more likely to reflect the situation occurring in nature in which mixed communities of both host plants and mycorrhizal fungi are present.

As with other studies using mesh barriers in the range 5–50 µm to provide root-free hyphal compartments (Johansen et al., 1993; Pearson & Jakobsen, 1993a,b), the meshes covering the cores could have been penetrated by root hairs, which typically have diameters between 5 and 20 µm (Wulfsohn & Nyengaard, 1999). Although we do not know for certain the importance of root hairs in P transfer from the static cores, their turnover is typically very fast. Cytoplasmic disintegration in root hairs has been shown to occur within 2–3 d (Fusseder, 1987) while vital staining of nuclei indicates root hairs live for 1–3 wk, although this method may overestimate root hair life span (Eissenstat & Yanai, 1997). The exponential increase in movement of source isotope from the cores during the 110 d period after cessation of rotation strongly indicates that hyphal re-growth, and not root hair growth, is likely to have been the main mechanism for transfer of 33P to plant shoots. In addition, whereas the root hairs of grass species have been found to achieve lengths in the range from 400 to 1000 µm (Caradus, 1980), mycelia of AM fungi have been known to extend for several cm from plant roots (Powell, 1978; Jakobsen, 1999). Since the isotope was injected into the centre of the cores 9 mm from the mesh barriers, root hairs are unlikely to have had extensive contact with the 33P. Movement of the isotope in the soil solution towards the edge of the core is also likely to have been limited due to the strong binding capacity of P in soil such as that studied here in which there was a high organic matter content (Marschner, 1995).

We have demonstrated the importance of natural communities of AM mycelium in the transport of P to their host plants in conditions simulating closely those occurring in nature. Further work using modified core systems is underway to determine the spatial scales at which hyphal P transfer occurs.


We gratefully acknowledge funding by the Natural Environment Research Council under the Soil Biodiversity Thematic Programme (Grant no: GST/02/2117). Dr Sarah Buckland (site manager) kindly assisted with the removal of turfs from the field and Irene Johnson provided skilled technical assistance.