Nutrient transfer from soil nematodes to plants: a direct pathway provided by the mycorrhizal mycelial network


Correspondence: Professor D. J. Read. Fax: + 44 114 222 0002; e-mail:


A pathway for the transfer of nutrients from dead nematodes to mycorrhizal plants is described for the first time. Plants of Betula pendula were grown in transparent microcosms in the mycorrhizal (M) or non-mycorrhizal (NM) condition, either with or without nematode necromass of known nitrogen (N) and phosphorus (P) contents as the major potential source of these elements. Plants colonized by the mycorrhizal fungus Paxillus involutus produced greater yields and had larger N and P contents in the presence of nematodes than did their NM counterparts. The symbiotic systems were shown to exploit the N and P originally contained in necromass more effectively, and to transfer the nutrients to the plants in quantities approximately double those seen in NM systems. Even so, NM plants obtained sufficient N and P from dead nematodes to enable some enhancement of growth. Our observations confirm that mycorrhizal fungi provide the potential for the recycling of nutrients contained in this quantitatively important component of the soil mesofauna and demonstrate that the symbiotic pathway is considerably more effective than that provided by saprotrophs alone. The consequences of this nutrient transfer pathway for nutrient recycling in temperate forest ecosystems are considered.


In recent times, there has been increasing recognition of the ability of ectomycorrhizal (ECM) fungi to mobilize plant nutrients contained in complex organic substrates (Leake & Read 1997). Although the potential for such activities was first revealed in studies using monoxenic cultures of fungi grown in the presence of pure compounds such as proteins (Abuzinadah & Read 1986a, b) and organic phosphates (Dighton 1983; Hilger & Krause 1989; Antibus, Sinsabaugh & Linkins 1992), emphasis has shifted, of late, to more realistic scenarios in which the fungi are grown in association with their hosts and are provided with substrates in the form in which they are likely to be encountered in nature. These have included detrital material collected from the fermentation horizon (FH) of forest soils (Bending & Read 1995) and pollen, a relatively well-defined potential substrate that is deposited annually in large quantities onto the forest floor (Perez-Moreno and Read 2001).

Among the other potential sources of nutrients for ECM plants are soil animals. Interestingly, whereas the mesofauna of forest soils has been well characterized both in terms of the quantitative representation of the various taxa (Petersen 1982) and of their contribution to the dynamics of decomposer food webs (Verhoef & Brussaard 1990), relatively little attention has been paid to the processes whereby the nutrients contained in the animals themselves are recycled through these webs. In the context of mycorrhizal systems, emphasis has been placed upon the possibility that fungivorous components of the mesofauna may reduce the effectiveness of the mycelial network in soils (Gange 2000), while the possibility that the necromass of soil animals might constitute a significant source of nutrients for mycorrhizal fungi has been largely overlooked. However, Klironomos and Hart (2001) have recently shown that one ectomycorrhizal fungus, Laccaria bicolor (Maire) Orton, might act as a predator on collembolans, so extracting nitrogen from their bodies.

Among the mesofaunal groups known to be present in very large numbers in temperate forest soils of the kind that support ECM trees are the nematodes. Petersen (1982) estimated that the biomass of these organisms in temperate forest ranged between 15 and 3040 mg dry weight (DW) m−2, these values being equivalent to 370 × 103 and 29·8 × 106 individuals m−2. As each individual nematode is a nutrient-enriched package that, on death of the organism, becomes a potential substrate for ECM fungi, it is of great interest to determine the extent to which mycorrhizal infection enables these resources to be captured by the plant.

Here, using birch seedlings grown in microcosms with the symbiotic fungus Paxillus involutus (Fr.) Fr., we test the hypothesis that a mycorrhizal symbiont has the ability to mobilize nutrients from dead nematodes. In addition, we evaluate the response of the colonized plant, in terms both of growth and nutrient contents, associated with the mobilization process.

Materials and methods

Mycorrhizal synthesis

Mycorrhizas were synthesized on Betula pendula Roth. seedlings using the ectomycorrhizal fungus Paxillus involutus following a procedure adapted from Brun et al. (1995). Seeds of B. pendula were surface-sterilized by exposure to H2O2 for 25 min, then rinsed with sterile distilled water and placed to germinate on water agar in Petri dishes. Paxillus involutus (strain USPI97 isolated from Cropton Forest, north Yorkshire, UK) was grown on cellophane overlying agar supplemented with modified Norkrans (1953) nutrient solution (see Perez-Moreno & Read 2000). Once the mycelium had covered approximately one-third of the area of the Petri dish, the birch seedlings were aseptically transferred, their roots being placed in direct contact with the mycorrhizal fungus. All lateral roots were colonized within 15 d, at which time seedlings were transferred singly to microcosms.

Construction of microcosms

Microcosms were constructed from 15 × 15 × 1 cm sheets of transparent plastic as described by Perez-Moreno & Read (2000). Over the surface of the lower plate, a film of water agar was poured, onto which a thin layer of inert clay (leca) granules was distributed. Finally, a layer of Sphagnum peat (2·5 g DW per microcosm) was spread over the granules to provide a nutrient-poor seminatural matrix for support of the seedlings and growth of the vegetative mycorrhizal mycelium. Single mycorrhizal (M) or non-mycorrhizal (NM) birch seedlings were transferred from the water agar plates to each of a series of such observation chambers to which were clipped an upper plate forming a sandwich. Shoots of the plants projected into the atmosphere. Roots were free to forage in the clay-peat substrate. The microcosms were wrapped in aluminium foil and incubated in a controlled environment growth room under an 18 h photoperiod (150 µmol m2 s−1) with a day : night temperature of 15 : 10 °C.

Extraction of nematodes and addition to the microcosms

The nematode selected for use, Heterorhabditis megidis, was obtained as a pure culture from a commercial source (Scarletts Plant Care, Colchester, UK). The cultures were stored at +5 °C until required, then large aggregates of nematodes, together with the clay carrier in which they were embedded, were added to distilled water and centrifuged at approximately 1000 ¥ g for 10 min in order to separate the nematodes from the culture medium. The centrifugation produced a discrete layer of nematodes at the base of the water column, overlying the clay carrier. From this layer, pure suspensions of the nematodes could then be carefully removed using a 250 µL pipette. The nematodes were placed onto sterile Petri dishes, which were then exposed to direct sunlight to produce rapid dehydration. The dry nematode aggregates were checked under the microscope to ensure that only pure samples were used in the experiments. They were then stored at + 5 °C until required. Preliminary experiments in which these nematodes were rehydrated following desiccation confirmed that mortality was 100%.

When the vegetative ectomycorrhizal mycelium of P. involutus had grown sufficiently to cover most of the peat-granule surface, the trays containing dead nematodes were added to microcosms containing M or NM plants using a modification of the approach used for pollen by Perez-Moreno & Read (2001). There was an additional set of M and NM chambers to which no nematodes were added. Dead air-dried nematodes were added in small (3 cm × 3 cm) plastic trays to half the M and NM microcosms. Each tray contained, as a support matrix, 2 g dry wt of clay (leca) granules in 2 mL of water agar (0·8%), onto the surface of which the nematodes were spread. The nematodes were added in two aliquots, the first consisting of 15 mg added at time zero and the second of 35 mg, which was added 45 d later. Two nematode-containing trays (i.e. 50 mg DW nematodes in total) were added per chamber together with a single tray containing support matrix only, the latter being used to provide a comparison of mycelial development in trays with and without nematodes.

The quantity of nematode material added to each microcosm when expressed on an area basis, was equivalent to 2220 mg DW m−2, and hence lay within the range 15 and 3040 mg DW m−2 reported by Petersen (1982) as being typically found in temperate forest soils. There were three replicate microcosms with nematodes (designated: + nematodes) and three without (designated: – nematodes) in both M or NM treatments.

All microcosms were incubated in a controlled environment growth cabinet (day–night temperature 15–10 °C, day length 18 h, irradiance 150 µmol m−2 s−1) for a further period of 195 d after which time harvest was carried out. During the incubation period the pattern of development of the vegetative mycelium of P. involutus over the trays containing dead nematodes, was followed. Regular visual inspection of all microcosms confirmed that roots neither entered nor contacted the trays themselves.

Nutrient analysis of plants and nematodes

At harvest, both M and NM plants grown in + or – nematode microcosms were partitioned into roots and shoots and oven-dried at 80 °C before their dry weights and tissue N and P contents were determined in replicate subsamples. The tissues were digested until clear in a concentrated H2SO4 (98%)–salicylic acid mix (33 g of salicylic acid dissolved in 1 litre of H2SO4) to which a catalyst was added (lithium sulphate : copper sulphate in a 10 : 1 ratio). The resulting samples were allowed to stand until they were cold and then were diluted with water. The extracts were stored at −20 °C before colorimetric analysis using a Tecator 5012 Analyser (Foss UK Ltd, Didcot, Oxfordshire, UK). N and P concentrations in the nematodes were determined at time zero and at harvest by the same methods as those described for plant materials.


Growth responses of the plants

Over the course of the experiment, birch plants grown in both the M and NM condition showed a progressive increase in vigour in the presence of dead nematodes. In contrast, plants grown in the equivalent treatment without nematodes were of stunted appearance and leaf development was weak (Fig. 1). At harvest it was found that in the presence of nematodes mycorrhizal colonization facilitated a significant increase of total yield of plants relative to those seen in NM treatments (476·1 ± 18·0 versus 307·0 ± 32·0; means ± SE, n = 3; Student’s t-test P < 0·05) (Fig. 2). Both shoots and roots of M plants had significantly greater biomass than did their NM counterparts (Fig. 2). Whole plant, shoot and root yields in the + nematode treatment were significantly greater than those in – nematode treatment in both M and NM microcosms (Fig. 2).

Figure 1.

(a–d) Appearance of representative B. pendula plants immediately before harvest. (a) Non-mycorrhizal plant grown in a microcosm without the addition of dead nematodes showing very weak development. (b) Mycorrhizal plant grown with P. involutus in a microcosm lacking nematodes showing development of mycorrhizal tips (single arrows), weak development of mycorrhizal mycelium (double arrows) and small plant growth response. (c) Non-mycorrhizal plant in microcosm supplied with nematodes showing extent of plant response achieved in the nonsymbiotic condition. (d) Mycorrhizal plant grown with P. involutus in a nematode-containing microcosm. Considerably greater plant vigour is evident in this, relative to the other treatments. Nematodes were added to the two right-hand trays (single arrows) in this, and in the non-mycorrhizal microcosm (c). In the mycorrhizal treatment intensive colonization of nematode-containing trays by Paxillus mycelium is evident, as is increased mycelial density in the peat ‘downstream’ of the trays (double arrows). Heavy mycorrhizal colonization of the roots of the B. pendula plants in this microcosm is seen.

Figure 2.

Dry weight yields of B. pendula plants grown with the ectomycorrhizal fungus P. involutus (above) or in the non-mycorrhizal condition (below) for 195 d in microcosms with (closed bars) and without (open bars) nematodes. Values are expressed as means ± SE. Asterisks indicate significant differences within tissue category according to Student’s t-test: *P < 0·05; **P < 0·005.

Nutrient contents of the plants

Evaluation of nutrient contents of the plants showed a similar overall pattern to that revealed by analysis of their growth. The total N and P contents of M plants grown in microcosms with nematodes were, respectively, 2·1 and 1·7 times greater than those of their equivalent NM treatments, the differences between these treatments being highly significant (Student’s t-test, P < 0·01)(Table 1). The shoots and roots of both M (Fig. 3a,c) and NM (Fig. 3b,d) plants grown with nematodes contained significantly greater N and P contents than those of plants in microcosms without nematodes, the differences between + and – nematode treatments being significant in all cases except that of root P contents of NM plants (Fig. 3d).

Table 1.  Total N and P contents of B. pendula grown in microcosms for 195 d, with or without dead nematodes as a potential nutrient source, either in the mycorrhizal condition with the fungus P. involutus, or as non-mycorrhizal plants
TreatmentWhole-plant contents (µg)Gain of nutrients by plants (µg) a
  1. Values are means ±SE, n = 3. aGain of nutrients by plants = (plant nutrient contents in microcosms with nematodes) − (plant nutrient contents in microcosms without nematodes). In this case, the ±SE were calculated by a postiori ranking of microcosms with mycorrhizal and non-mycorrhizal plants. Asterisks indicate significant differences between nutrients in microcosms with and without nematodes for whole-plant contents and between nutrients in microcosms with mycorrhizal and with non-mycorrhizal plants for gain of nutrients using t-tests at *P < 0·05 and **P < 0·01.

Microcosms with non-mycorrhizal plants  1122·0 ± 31** 48·2 ± 11·0*
With nematodes1971·9 ± 148* 99·3 ± 5·8  
Without nematodes 849·9 ± 179 51·1 ± 15·0  
Microcosms with mycorrhizal plants  2810·5 ± 98106·8 ± 8·0
With nematodes4077·3 ± 131**166·1 ± 8·5**  
Without nematodes1266·8 ± 188 59·3 ± 16·0  
Figure 3.

Nitrogen (N) and phosphorus (P) contents of B. pendula plants grown with the ectomycorrhizal fungus P. involutus (a, c) or in the non-mycorrhizal condition (b, d) for 195 d in microcosms with (closed bars) and without (open bars) dead nematodes as a potential nutrient source. Values are expressed as means ± SE. Asterisks indicate significant differences within the mycorrhizal or non-mycorrhizal category according to Student’s t-test *P < 0·05.

The M plants had higher N and P contents than those in the NM category. The N contents of shoots and roots of M plants grown in microcosms with nematodes were 2·3 and 1·9 times greater than those in the equivalent NM treatments, there being significant differences (Student’s t-test, P < 0·05) between both M and NM shoots (1642·5 ± 203·0 versus 709·1 ± 47·0; means ± SE, n = 3) and roots (2434·8. ± 266·0 versus 1263·0 ± 142·0; means ± SE, n = 3) (Fig. 3a,b). Similarly in the case of P (Fig. 3c,d), the shoot and root contents were, respectively, 1·4 and 1·8 times greater in the M + nematode than those in the NM + nematode treatment. The whole-plant quantities of N and P were greater in + nematode than in – nematode treatments in both M and NM plants with the exception of P in NM plants (Table 1).

Nutrient depletion in nematodes

The growth responses and gains of nutrients in M and NM plants grown in the presence of nematodes were paralleled by significant reductions of both N and P concentrations in nematodes (Fig. 4). However, in the microcosms containing M plants larger amounts of nutrients were transferred from the nematodes than in those containing NM plants, there being significant differences between treatments (P < 0·05). The reductions in N and P concentrations of the nematode necromass of microcosms containing M plants were approximately double those seen in microcosms supporting NM plants (Fig. 4). Colonization of dead nematodes by P. involutus was associated with a reduction of their total N and P contents after 195 d of 68·2 and 65·7%, respectively. The differences between these and the NM systems, where the depletion of the same nutrients was only 37·1 and 24·5%, respectively, were significant at P < 0·05 (Table 2).

Figure 4.

Concentrations of nitrogen (N) and phosphorus (P) in nematode tissues at the time of their addition (open bars) and after 195 d in microcosms containing non-mycorrhizal plants (hatched bars), and plants colonized by the mycorrhizal fungus P. involutus (closed bars). Values are expressed as means ± SE. A change of letter over a bar indicates significant difference in concentration of the element according to the Tukey multiple comparison test (P < 0·05).

Table 2.  Nitrogen and phosphorus depletion of nematode tissues after 195 d incubation in microcosms supporting B. pendula grown with the mycorrhizal fungus P. involutus or in the non-mycorrhizal condition
 Total nutrient contents (µg) of added nematodesReduction of nutrients in nematodes in relation to initial contents
Nutrient and treatmentInitialAfter 150dWeight (µg)%
  1. Values are means ±SE, n = 3. Asterisks indicate significant differences between nutrients in microcosms with mycorrhizal and with non-mycorrhizal plants using t-tests at *P < 0·05.

Microcosms with non-mycorrhizal plants3073·17 ± 1771933·2 ± 125*1140·0 ± 125*37·1 ± 4·1*
Microcosms with mycorrhizal plants 977·5 ± 1502095·7 ± 15068·2 ± 4·9
Microcosms with non-mycorrhizal plants 223·5 ± 11·4 168·8 ± 10·9* 54·7 ± 10·9*24·5 ± 4·9*
Microcosms with mycorrhizal plants 76·6 ± 11·8 146·9 ± 11·865·7 ± 5·3

Relation between nutrient gains in plants and their losses from nematodes

Comparisons of nutrient gain values in M and NM systems indicate that the symbiosis facilitated significantly more nutrient transfer to the plants (Table 1). The total N and P transfer from nematodes in M plants was more than double that seen in their NM counterparts (Table 1). In the case of P the gain by M plants amounted to 106·8 µg (Table 1) whereas the loss from the nematodes was 146·9 µg (Table 2). Thus, only a portion, approximately 73%, of the P removed from the dead nematodes was transferred from the fungus to the plant. The quantity (106·8 µg) of P acquired by the M plants in the + nematode systems represented 48% of the 223·5 µg originally present in the animal tissues. In contrast, in the NM systems, the P gain amounted only to 48·2 µg (Table 1), this being half of that transferred in M systems, and representing only 22% of the total originally present in nematodes.

A different overall pattern was seen in the case of N, where the gain by M plants, amounting to 2810·5 µg (Table 1), was greater than the loss from the nematodes (2095·7 µg) (Table 2), there being significant differences between these values (Student’s t-test, P < 0·05). The possible basis of the apparent discrepancy between loss of N from the putative source and gain by the plant is discussed below (Discussion). By contrast, in NM systems the N gain by plants of 1122·0 µg (Table 1) and the losses from nematodes, 1140·0 µg (Table 2), were similar.


To our knowledge this is the first demonstration that ectomycorrhizal colonization can provide access to the nutrient pools contained in the bodies of soil nematodes. It is an important observation because nematode populations are typically very large in the normally nutrient-impoverished forest soils that support ectomycorrhizal plant communities (Petersen 1982). The results indicate that a significant proportion of the most important plant nutrients that are initially sequestered in the bodies of soil nematodes can be effectively recycled by mycorrhizal fungi to their autotrophic partners rather than being immobilized or transferred to other levels within the food web.

The ability of some saprotrophic soil fungi to use dead nematodes as nutrient sources has been previously recognized. Dyer, Boddy & Preston-Meek (1992) supplied two basidiomycetous wood decomposer fungi, Phanaerochaete velutina (DC.Pers.) Parmasto and Stereum hirsutum (Willd.Fr.) S.F. Gray with necromass of the nematode Panagrellus redivivus (L.) as a sole nutrient source. In the presence of this substrate they observed some increase in mean radial extension growth as well as of protease and acid phosphatase activity of P. velutina. Under the same experimental conditions increases in activity of this enzymes were also observed in S. hirsutum. Since the ectomycorrhizal fungus P. involutus employed in the present study, showed clear evidence of growth stimulation in and around the nematode-containing trays and is known to have the potential for release of these extracellular enzymes, it is possible to hypothesize that the improved growth and nutrient acquisition seen in the colonized plants was attributable to direct attack by the fungus upon the bodies of the nematodes. However, as considerable time elapsed between placement of the necromass-containing trays into the microcosms and their colonization by P. involutus, it is more likely that the primary nutrient mobilizing event were initiated by saprotrophs, the propagules of which, be they fungal or bacterial, would be associated with the bodies of the nematodes at the time of their death. In this context it is of interest that Tsuneda & Thorn (1995), while noting that the fungal saprotroph Pleurotus ostreatus (Jacq.Fr.) Kummer was able to colonize the corpses of dead nematodes, observed that decomposition was initiated by bacteria and that penetration by the fungus of the bacterial biomass rather than direct attack appeared to facilitate its access to nutrients.

If, as our observations suggest, the success of the symbiosis in this case is based upon secondary exploitation of nutrients released by earlier colonists the situation is clearly different from that seen in the specialized group of nematophagous fungi that have the ability both to kill their prey and directly mobilize the nutrients contained in them by means of a variety of well-characterized enzymatic processes (Toth, Toth & Nordbring-Hertz 1980; Schenck et al. 1980; Tunlid et al. 1994; Araujo et al. 1997).

The nutrient-foraging strategy deployed by ectomycorrhizal fungi, which is dependent upon their access to carbohydrate supplies from the autotroph (Söderström & Read 1987), is to produce an extensive mycelial system that enables effective exploitation of the soil volume (Read 1992). This ensures that while they may not be in direct contact with individual food depots such as the senescing bodies of nematodes, they are close enough to respond to any nearby enrichment of nutrient availability by localized proliferation of hyphal branching (Bending & Read 1995; Perez-Moreno & Read, 2000, 2001).

The present experiment provides a test of the relative effectiveness, for the plant, of mycorrhizal and saprotrophic fungal nutrient mobilization pathways, because in neither case did roots make contact with the trays containing the animal necromass.

As was shown when pollen was provided as the sole potential source of nutrients in mycorrhizal and non-mycorhizal microcosms (Perez-Moreno & Read 2001), saprotrophic activities alone enabled sufficient mobilization of N and P to facilitate some growth stimulation and nutrient enrichment in the plant. However this process is relatively ineffective presumably in part because saprotrophs will transfer nutrients into the soil in all directions from a localized source rather than towards individual roots. In contrast, the mycelium of the symbiont P. involutus, can allocate its resources directly to the interconnected mycorrhizal roots through the mycelial network that is sustained by carbon supplied by the host plant. The rapidity with which the symbiont mycelium proliferates after contacting the substrate indicate that it will be able to respond effectively to short-term pulses of nutrient release.

The extent of N and P mobilization in mycorrhizal systems, 68 and 65%, respectively, from dead nematodes was somewhat less than that reported from pollen (Perez-Moreno & Read 2001) where the equivalent values were, respectively, 76 and 97%. However, the quantities of N and P mobilized from these animals were greater than those reported from detrital material collected from the forest soil FH (Entry, Rose & Cromack 1991; Bending & Read 1995; Perez-Moreno & Read, 2000). Amounts of nutrient release from FH substrates ranged from 0 to 32% in the case of N and from 3 to 40% in that of P, being in all cases much smaller than those from nematodes.

The nutrient analysis showed that in the case of M systems, the N gain in the plants was greater than was the loss from nematodes. A possible explanation for the observed discrepancy could be the occurrence of N2-fixing activity in the conspicuous ECM mycelial patches associated with the dead nematodes. Both associative N2-fixation and the presence of bacterial species, e.g. Bacillus spp., known to have N2-fixing potential have been previously found in the mycorrhizosphere of forest ECM trees (Li & Hung 1987; Li, Massicote & Moore 1992; Nurmiaho-Lassila et al. 1997; Timonen et al. 1998). It has been pointed out by Perez-Moreno & Read (2000) and Sen (2000) that the vigorous growth of external vegetative ECM mycelium would lead, by enhancement of CO2 release, both to low oxygen tension and high C-availability, providing therefore an ideal niche for the development of associative N2-fixing bacteria. The occurrence of N2-fixation requires confirmation, preferably by direct quantification using 15N labelling methods. Perez-Moreno & Read (2000) found a similar discrepancy between N loss from FH materials and gains by mycorrhizal plants. In that study, the excess N observed in the plants was even greater than that observed in the present study.

An additional possibility is that exploitation of the nutrient contents of the nematodes has a ‘priming’ effect upon the fungus enabling it more effectively to scavenge for N from the surrounding growth environment. It was observed that formation of dense mycelial patches over trays enriched with dead nematodes, was followed by enhancement of mycelial growth in areas distal to the trays and along lower lateral edges of the microcosms (Fig. 1b). Stimulation of hyphal growth in territories ‘downstream’ of nutrient enriched patches has been observed in saprotrophic fungi (Rayner 1996). An effect of this growth enhancement, which may be facilitated by release from P limitation, could be to increase the ability of the fungus to scavenge for, or mineralize N from, the peat.

It is necessary to consider the extent to which and the mechanisms whereby the processes observed in this study might operate in the field. Clearly the distribution of nematodes is unlikely to be as clustered in nature as in the microcosm systems. However there is evidence (De Goede, Verschoor & Georgieva 1993) that, under field conditions, a significant proportion of the numerically large population of these organisms is concentrated in that part of the soil profile, the FH horizon, in which ECM mycelium proliferates most intensively. Indeed this concentration may in part arise specifically because, for the fungivorous nematodes at least, this horizon is the richest source of nutrient provision. It is not clear from the literature whether this group of nematodes does feed extensively upon mycorrhizal mycelia but the size of its biomass indicates that major impacts upon some components of the fungal system are likely.

Nematodes, be they detritivores, bacterivores or fungivores have nutrient-enriched bodies in which the protein contents range from 50 to 80% (Lee & Atkinson 1976). To the extent that these organisms, except in the egg stage, do not produce structural nitrogenous polymers like chitin, they probably represent a more labile potential source of nutrients than do some other mesofaunal groups, for example collembolans. This would lead to relatively rapid turnover of the nutrients originally sequestered in the population. Since Laccaria bicolor has been shown (Klironomos and Hart 2001) to attack the bodies of living as well as dead collembolans, and to facilitate transfer of nitrogen from the animals to its host plant, it appears that even these more complex nutrient sources may be accessible to some mycorrhizal associations.

Whether or not mycorrhizal fungi such as Paxillus, with their acknowledged abilities to release the extracellular enzyme protease (Abuzinadah & Read 1986a; Finlay, Frostegard & Sonnerfeldt 1992) and phosphatase (Hilger, Thomas & Krause 1986; Hilger & Krause 1989; Kieliszewska-Rokicka 1992) are primary decomposers, it is clear that the ability of their mycelia to occupy the FH and to rapidly proliferate within it in response to localized nutritional cues will facilitate effective involvement in the decomposer food web, to the direct advantage of the plant. Within this web, the key to the foraging success of the ECM fungi, whether it involves primary degradation of nematode corpses or secondary attack upon the first waves of saprotrophs, is likely to be their freedom from carbon limitation.


We thank Ms Irene Johnson, Mr Glynn Woods and Mr David Hollingworth for excellent technical assistance. We also thank the British Chevening Scholarships Programme, CONACyT (Consejo Nacional de Ciencia y Tecnología) and Colegio de Postgraduados (Mexico) for provision of financial support to J.P.-M.