Phosphate transport by communities of arbuscular mycorrhizal fungi in intact soil cores

Authors

  • I. Jakobsen,

    Corresponding author
    1. Plant Biology and Biogeochemistry Department, Risø National Laboratory, DK-4000 Roskilde, Denmark;
      Author for correspondence: I. Jakobsen Tel: +45 46 77 41 54 Fax: +45 46 77 42 82 Email: iver.jakobsen@risoe.dk
    Search for more papers by this author
  • C. Gazey,

    1. Soil Science and Plant Nutrition, Faculty of Agriculture, The University of Western Australia, Nedlands, WA 6907, Australia;
    2. Present address: Plant Research and Development Services, Agriculture Western Australia, Baron-Hay Court, South Perth, WA 6151, Australia
    Search for more papers by this author
  • L. K. Abbott

    1. Soil Science and Plant Nutrition, Faculty of Agriculture, The University of Western Australia, Nedlands, WA 6907, Australia;
    Search for more papers by this author

Author for correspondence: I. Jakobsen Tel: +45 46 77 41 54 Fax: +45 46 77 42 82 Email: iver.jakobsen@risoe.dk

Abstract

  • • Variation in phosphate uptake capacity is reported here for natural communities of arbuscular mycorrhizal fungi associated with annual pasture plants.
  • • Tests were made of methodology for quantifying phosphate uptake by hyphae associated with clover in soil cores from pastures containing different morphotypes of the fungi. This provided a direct measure of the phosphate uptake capacity of hyphae from 32P-labelled soil in a root-free mesh bag inserted into the centre of intact soil cores.
  • • Bicarbonate-extractable phosphorus in the soils ranged from very deficient to close to adequate for plant growth. Uptake of 32P was related to an estimate of the length of hyphae formed in four of the five soils, but not to either the length or the proportion of roots colonized. In the fifth soil type, phosphate uptake by hyphae was negligible.
  • • Phosphate uptake by natural communities of arbuscular mycorrhizal fungi in intact soil cores can be assessed directly, and is shown to be highly variable. The experimental approach could be applied widely for field investigations of phosphate uptake by hyphal networks.

Introduction

Growth promoting effects of the ubiquitous arbuscular mycorrhizal (AM) fungi are easily demonstrated in pot experiments with semisterile soil, but their contributions in field soils are poorly understood and difficult to evaluate (Fitter, 1985; Jakobsen, 1994). Colonization of roots by AM fungi modifies the growth response of the plant to increasing supplies of phosphorus and the degree of modification depends on the initial soil P level and the quantity and form of inoculum (Abbott & Robson, 1981; Abbott et al., 1995). Symbiotic characteristics that may be responsible for the variation among AM fungi in their ability to influence plant growth include: the rate of mycorrhiza formation and the production of external hyphae (Sanders et al., 1977); the inherent P transport capacity of the fungus (Jakobsen et al., 1992b); and the functional compatibility between the host and fungus in terms of either fungal nutrient transport (Ravnskov & Jakobsen, 1995) or plant growth response (Graham & Eissenstat, 1994).

Communities of AM fungi vary in their effectiveness at increasing plant growth (Hamel et al., 1997), but the reasons for this have not been clarified. The potential for future management of field communities of AM fungi depends on understanding this intercommunity variation. Identification of major determinants of effectiveness of field communities may facilitate the development of a model to predict the mycorrhizal benefits at a field site and to evaluate the feasibility of inoculation (Abbott & Robson, 1991).

Bioassays with intact soil cores showed that the rate of root colonization is one important variable for predicting the contribution of naturally occurring AM fungi in field soils (Abbott et al., 1995). However, field communities of AM fungi are composed of a number of species that may coexist in the same root system (Blaschke, 1991; Merryweather & Fitter, 1998a) as well as in soil. AM fungi differ in their capacity to transport P (Jakobsen et al., 1992b). Therefore, it is likely that the P transport capacity of communities of AM fungi will depend not only on the rate and extent of root colonization, but also on the transport characteristics of the fungi that dominate within roots at times when the plant has greatest requirement for P. Realistic modelling of benefits of AM fungi in field soils may, therefore, depend on the incorporation of a variable describing the nutrient transport by the AM fungi present. This could be obtained from direct measurements of P uptake by hyphae from the soil into the plant (Jakobsen, 1994; Schweiger et al., 1999).

The objective of this experiment was to study the variation in P uptake capacity of naturally occurring communities of AM fungi associated with annual pasture plants in six field sites in south-western Australia. We expected differences in P uptake by the fungi at different sites to be related to the length of hyphae formed in soil and to the extent of root colonized by AM fungi (Abbott & Robson, 1991). The objectives were addressed using intact soil cores, which contained root-free mesh bags with 32P-labelled soil. Uptake of 32P through AM hyphae into subterranean clover was assessed with and without the application of benomyl using the method of Schweiger & Jakobsen (1999). In addition, 32P uptake by hyphae associated with subterranean clover was assessed in bulked soil from three of the pasture soils, which were steamed and either inoculated or not inoculated with Glomus invermaium. This part of the study was included to assess 32P uptake in the experimental system using a fungus known to be effective when present at high inoculum ‘potential’.

Materials and methods

A glasshouse experiment was carried out with intact soil cores or bulked, steamed soil from pasture sites in south-western Australia. There were two parts to the study. Part (i) involved assessment of 32P uptake by hyphae associated with subterranean clover grown in intact field cores from 6 pasture sites either with or without application of benomyl, and Part (ii) involved assessment of 32P uptake by hyphae associated with subterranean clover grown in bulked, steamed soil from three pasture sites (Soils 1, 2 and 6) either inoculated or not inoculated with Glomus invermaium Hall.

A root-free mesh bag filled with sieved 32P-labelled soil was inserted into each pot (soil core or bulked soil) and the transport of 32P into shoots of subterranean clover growing outside the mesh bag was measured (Schweiger et al., 1999). P uptake, hyphal growth, root colonization by AM fungi and shoot fresh weights were measured for plants grown in soil cores ± the addition of benomyl (i). Plant growth and the fungal P transport were assessed in bulked and steamed soil from three of the sites (Soils 1, 2 and 6) with or without inoculation with G. invermaium (ii). All treatments in both parts of the study had three replicates.

Soils

Soil cores were sampled in early September (spring) from six subterranean clover (Trifolium subterraneum L.) pastures (Table 1). The pastures differed in rotation history and covered a range in some soil characteristics, especially bicarbonate extractable phosphorus (Table 1). Soil P levels were measured by two methods of bicarbonate extraction (Olsen et al., 1954; Colwell, 1963). Soils 1 and 3 had a high gravel content (33% and 18%, respectively) and Soils 3 and 6 had a high organic carbon content for this region (4.3% and 4.0%, respectively). Intact soil cores (11 cm diameter and 14 cm depth) were taken by means of a steel cylinder at each site soon after rain while soils were approximately at field capacity. The cores were transferred to pots of similar size that were lined with plastic bags. The total weight was recorded for each core and pots were placed in temperature-controlled tanks at a constant 20°C in a glasshouse. Additional cores (3 per site) were collected for harvesting prior to the main experiment being sampled, to estimate the extent of colonization of roots. This information was used to decide on the time for harvesting the pots of 32P treated soil. Bulk soil was collected (to the same depth as the soil cores) for Soils 1, 2 and 6 and steamed for 1 h on two consecutive days. No nutrients were added to any soils other than the 32P placed into the mesh bag.

Table 1.  Description (Northcote, 1979) and selected characteristics of experimental soils
SoilSoil type, location and crop rotationOlsen P (µg g−1 soil)Colwell P (µg g−1 soil)pH (10−2M CaCl2)Organic C (%)
  1. Phosphate levels were measured by two methods using extraction by NaHCO3 (Olsen et al., 1954; Colwell, 1963). All cores were taken in the pasture phase of all rotations.

1Yellow gravelly sand 2 45.13.7
 KS-Uc 4.21    
 Westdale (E 468738 N 6422273)    
 Permanent pasture    
2Reddish brown sandy loam 5134.71.9
 Dy 2.23    
 New Norcia (E424680 N6572951)    
 Pasture – pasture – wheat    
3Grey gravelly sand 8215.34.3
 KS-Dy 4.52    
 Wundowie (E439850 N6482835)    
 Pasture – wheat – lupin    
4Yellow gravelly sand 9175.02.3
 KS-Dy 4.52    
 Wundowie (E440826 N6482983)    
 Permanent pasture    
5Yellow sandy loam15324.82.8
 Dy 2.13    
 Katanning (E551427 N6272249)    
 Wheat – wheat – pasture    
6Brown alluvial loam28754.94.0
 Dy 3.12    
 Wundowie (E439329 N6482654)    
 Permanent pasture    

Measurement of 32P uptake by naturally occurring AM fungal hyphae in intact soil cores from pastures (1)

Experimental design: 6 soils × ±benomyl (3 replicates). Six replicate cores of soil from six sites were collected and three of the cores from each site were treated with benomyl. Shoots of plants growing in the cores at the time of collection in the field were cut just below the soil surface to prevent regrowth of plants already present. A soil core of 3 cm diameter was removed from the centre and to the full depth of each core, and a nylon mesh bag (37 µm mesh size) filled with 100 g soil was inserted into the space. This soil had originated from the corresponding field site, immediately adjacent to where the intact cores were collected, and was sieved (<2 mm), air dried and labelled with 32P soil.

The radioisotope was supplied as a carrier-free solution of H332PO4 to the surface of weighed soil lots, allowed to dry and then mixed carefully into the soil to be placed in the mesh bag. The radioactivity contained in the soil for one mesh bag corresponded to 527 kBq at the time of planting. Half of the pots with intact soil cores received 50 mg benomyl (50% a.i.) per mesh bag together with the 32P in order to impair growth and function of the AM fungi. The benomyl was supplied in aqueous suspension and allowed to dry before mixing into the soil. Pots with benomyl mixed into the soil in the mesh bag also received 10 mls of an aqueous suspension of 100 mg benomyl on the soil surface outside the mesh bag in order to reduce mycorrhiza formation. This surface application was repeated 22 d after sowing to the pots which had previously received benomyl resulting in a total of 250 mg benomyl per pot. The surface-applied benomyl would move into the soil cores at watering, but not at an even distribution due to the low mobility of benomyl in soil (Pedersen & Sylvia, 1997).

Measurement of 32P uptake by hyphae of Glomus invermaium inoculated into 3 steamed pasture soils (2)

Experimental design: 3 soils (bulked and steamed) × ± inoculation (3 replicates). Additional treatments were prepared with steamed Soils 1, 2 and 6 in order to measure both the 32P uptake from the mesh bags in nonmycorrhizal plants and the 32P uptake in plants colonized with an inoculant AM fungus. Only three of the soils, covering the range of P levels of the soils collected, were selected for this component of the study because of limited experimental capacity for 32P assessment. Mesh bags were filled with sieved and steamed soil that had been supplied with 32P and the corresponding steamed, bulk soil was packed around the bags in pots lined with plastic bags. Crude soil inoculum (50 g) of G. invermaium Hall (WUM 10(1), BEG 44) from a dried pot culture with subterranean clover in Lancelin sand (pH 5.3; Olsen P < 8 µg g−1 soil) was mixed uniformly into the steamed bulk soil for half of the pots. The total weight of dry soil per pot was 1600, 2100 and 1600 g for Soils 1, 2 and 6, respectively.

General methods

The 18 pots containing soil cores alone and the 54 pots (i + ii) containing mesh bags were placed in the temperature-controlled tanks and watered to maintain the soil at field capacity. Pots of steamed soil were maintained at 70% field capacity throughout the experiment in the same tanks. After 1 wk, six germinated seeds of T. subterraneum L. cv. Seaton Park were sown in each pot. Each seed received 0.5 ml of a dense suspension of Rhizobium leguminosarum biovar trifolii (WU1) in 1% sucrose. After emergence, the seeds were thinned to three per pot and the soil surface was covered with Alkathene® (Quenos Pty. Ltd., Altona, Victoria, Australia) beads to reduce evaporation.

Harvest and analyses

The untreated cores without mesh bags were harvested after 21 d and roots were washed out of the soil for assessment of early root colonization by AM fungi. Plants in pots of all other treatments were harvested after 33 d.

At harvesting, shoots were excised and mesh bags removed from each pot. Intact root systems were washed and old root pieces not originating from the experimental plants were removed. A weighed subsample was cleared and stained (Gazey et al., 1992) for measurements of root length using the line intercept method (Newman, 1966); mycorrhizal colonization was recorded simultaneously. The relative abundance of morphotypes of AM fungi in roots of cores treated with 32P in Part (i) were assessed (Abbott, 1982). Fungi in 50 root intercepts per treatment were characterized. Plant material, except for untreated pots in Part (i), was dried at 70°C, weighed and analysed for total phosphorus content by the molybdovanado-phosphate method (Boltz & Lueck, 1958) after acid digestion (Johnson & Ulrich, 1959).

Radioactivity in the digests was measured by Cerenkov counting (Kessler, 1986) on a Packard (Packard Instrument Co., Meriden, CT, USA) 1500 scintillation counter; results were corrected for counting efficiency and isotope decay.

The soil in the mesh bags was air dried and stored dry for 4 months until levels of radioactivity were very low. Hyphal lengths were then measured in soil from the mesh bags after storage using a membrane filter technique (Jakobsen et al., 1992a).

Statistics

Treatment effects were detected by two-way ANOVA and means were compared by means of LSD0.05. Relationships between 32P uptake and hyphal length or root colonization were calculated by the method of least square fit using data points from individual pots.

Untreated cores of Soil 6 in Part (i) were not included in the analysis due to the presence of high densities of root-infecting nematodes (Meloidogyne arenaria) and retarded plant growth. This soil requires further investigation.

Results

32P uptake by naturally occurring AM fungal hyphae in intact soil cores from pastures (1)

Mycorrhizal colonization

Roots of seedlings of subterranean clover became rapidly colonized by the fungi present in each soil (Tables 2 and 3). Different combinations of fungi were present in roots of plants grown in each soil at the time that 32P was assessed in plants (Table 3). Roots of plants in Soils 1 and 3 were dominated by a Glomus species resembling G. invermaium. Two different Glomus morphotypes were abundant in Soils 2 and 5 although the relative amount of each was different. In Soil 4, roots were jointly colonized by fungi resembling Acaulosporalaevis and G. invermaium.

Table 2.  Colonization of roots by AM fungi in intact soil cores from pastures resown with Trifolium subterraneum grown for 21 and 33 d
 Root length colonized (%)
Soil*21 d33 d
  1. (n = 3; SE in parentheses)*Data for Soil 6 not included due to nematode infestation.

167 (11)50 (8)
288 (2)61 (6)
356 (4)62 (6)
456 (3)42 (4)
571 (3)52 (11)
Table 3.  Proportion of colonized roots of Trifolium subterraneum sown into intact cores of soil from five pasture sites occupied by morphotypes of naturally occurring AM fungi
 Colonization by morphotype of arbuscular mycorrhizal fungia (%)
SoilGlomusb >4 µmGlomusc <3 µmAcaulosporaFine endophyteGigaspora/ScutellosporadUnknowne
  1. (n = 3; SE in parentheses).aMorphotypes based on morphology in Abbott (1982).bHyphae within roots >4 µm in diam.cHyphae within roots <3 µm in diameter.dDistinction between fungi in these genera not made.eUnidentified morphotypes.

1 098 (2) 02 (2)00
248 (11)38 (11) 1 (1)9 (4)3 (3)0
3 2 (2)87 (4) 0006 (1)
4 2 (1)39 (6)55 (6)006 (1)
516 (12)83 (12) 1 (1)001 (1)

Fractional root colonization at the time of harvest (33 d) varied only little and was accordingly not related to the soil P level (r2 = 0.02). The colonized root length was not different between soils, but was reduced by the benomyl treatment (P < 0.05, Table 4).

Table 4.  Shoot dry weight, root length and hyphal length density of Trifolium subterraneum-AM fungal symbioses grown in intact soil cores without or with benomyl
   Root length 
Soil No.Soil treatmentPlant dry weight (mg per plant)Total (cm per plant)ColonizedHyphal length density (m g−1)
  1. Hyphae length was measured in a mesh bag that excluded roots (n = 3).

1Untreated 66271183 9.9
 Benomyl 9031919010.5
2Untreated11639332124.7
 Benomyl 87314184 9.0
3Untreated16147126613.9
 Benomyl113383145 7.0
4Untreated202666372 9.9
 Benomyl143435218 7.6
5Untreated27944431412.5
 Benomyl25947421114.7
LSD0.05  61202156 3.5
ANOVA, probabilities     
Soil   0.000  0.013  0.276 0.000
Benomyl   0.100  0.152  0.007 0.000
Interaction   0.557  0.288  0.585 0.000

The highest hyphal length within the mesh bags was observed in Soil 2 (Table 4). Lengths of hyphae in Soil 3 were higher than those in Soils 1 and 4. The effect of benomyl on hyphal length interacted with soil (P < 0.05) and a significant reduction was observed in Soils 2 and 3.

Plant growth

Plant dry weights in the soil cores differed up to four-fold between soils in accordance with large differences in soil P levels (Tables 1 and 4). Total root lengths in Soils 1–4 increased with increasing soil P level, but less so than the dry weights; the root length was lower in Soil 5 than in Soil 4. A similar pattern was observed for total root length and benomyl had a negative effect on plant dry weights in Soils 3 and 4 (Table 4).

Phosphate uptake

Plants grown without benomyl had higher shoot P concentration in Soil 2 than in Soils 3, 4 and 5 in spite of their higher levels of available P (Table 5). The relatively high P status and P uptake in plants from Soil 2 corresponded with a high content of 32P in those plants. In untreated soil cores the uptake of 32P from mesh bags in Soil 2 was four–six-fold that in plants in Soils 3, 4 and 5 (Table 5). Plants grown in Soil 1 were extremely P deficient and contained a very low level of 32P (Table 5). The uptake of 32P in plants grown in untreated soil cores increased with increasing hyphal length in the mesh bag (Fig. 1). In contrast, linear regression analysis showed that the uptake of 32P in untreated pots was not related to the fractional root colonization (r2 = 0.08) or to the colonized root length (r2 = 0.002).

Table 5.  Phosphorus concentration, total P content and content of 32P in Trifolium subterraneum grown in intact field cores with or without benomyl. 32P was taken up from a root-free mesh bag inserted into each core (n = 3)
  P concentration (%)  
Soil No.Soil treatmentShootRootP content (µg per plant)32P content (Bq per plant)
1Untreated0.080.11 63  19
 Benomyl0.040.09 71   6
2Untreated0.330.313767366
 Benomyl0.230.23201 142
3Untreated0.180.202991790
 Benomyl0.130.16158  12
4Untreated0.240.275141332
 Benomyl0.200.23309 291
5Untreated0.230.337551289
 Benomyl0.200.26592 388
LSD0.05 0.040.051771157
ANOVA, probabilities     
Soil 0.0000.000  0.000   0.000
Benomyl 0.0000.001  0.007   0.000
Interaction 0.1360.444  0.680   0.000
Figure 1.

Relationship between 32P taken up from soil in root-free mesh bags by Trifolium subterraneum and hyphal length density in the soil within the mesh bag. Plants were grown in untreated soil cores for 33 d. Symbols correspond to soil Nos: × (1), filled circle (2), filled square (3), filled triangle (4), and filled diamond (5). The curve represents the equation y = 18.8x2 – 335.4x + 2304.5; r2 = 0.87.

P uptake was markedly reduced by benomyl (P < 0.05), except in Soil 1 where there was little P uptake. This effect was even more pronounced for the content of 32P in the plants (P < 0.05, Table 5). Plants from the benomyl-treated cores for Soils 2, 3, 4 and 5 contained 2, 1, 21 and 30%, respectively, of the 32P content in the corresponding plants from untreated soil cores. The proportion of the variance in 32P uptake in benomyl-treated pots, which could be explained by the variance in hyphal length was relatively small (y = 7.25x2 − 132.2x + 698.7; r2 = 0.33) as compared to untreated pots (Fig. 1).

32P uptake by hyphae of Glomus invermaium inoculated into three steamed pasture soils (2)

Fractional root colonization of G. invermaium-inoculated plants grown in bulk, steamed Soils 1, 2 and 6 was high (>78%, Table 6). The hyphal length density measured in steamed soil was increased by inoculation only in Soil 2 (Table 6).

Table 6.  Plant dry weight, root length and hyphal root length density of Trifolium subterraneum grown in steamed soils inoculated with Glomus invermaium or not inoculated. Hyphal length was measured in a root-free mesh bag inserted into each core (n = 3)
   Root length 
Soil No.Inoculant fungusPlant dry weight (mg per plant)Total (cm per plant)ColonizedHyphal length density (m g−1)
1G. invermaium11931527614.1
 None 68290  012.5
2G. invermaium35380566618.1
 None195579  010.1
6G. invermaium36258145411.3
 None174612 1414.7
LSD0.05  28174155 3.3
ANOVA, probabilities     
Soil   0.000  0.000  0.008 0.572
G. invermaium   0.000  0.137  0.000 0.037
Interaction   0.000  0.096  0.008 0.001

Plants grew better in the bulked Soils 1 and 2 after steaming and inoculation with G. invermaium than in untreated soil cores in Part (i) sampled after 33 d. Growth responses to the introduced fungus in steamed soil were greater than differences between untreated and benomyl-treated soil cores in Part (i) (Tables 2 and 6).

Phosphate uptake was higher in Soils 1 and 2 after steaming and inoculation with G. invermaium than for the untreated soil cores in Part (i) (Tables 5 and 7). The P uptake of G. invermaium-inoculated plants grown in Soil 6 was similar to that recorded for inoculated plants grown in Soil 2 although there was almost five-fold more extractable P in Soil 6 than in Soil 2.

Table 7.  Phosphorus concentration, total P content and content of 32P in Trifolium subterraneum grown in steamed soils inoculated with Glomus invermaium or not inoculated. 32P was taken up from a root-free mesh bag inserted into each core (n = 3)
  P concentration (%)  
Soil No.Inoculant fungusShootRootP content (µg per plant)32P content (Bq per plant)
1G. invermaium0.130.16 170 6722
 None0.020.06  33   57
2G. invermaium0.240.36102414170
 None0.120.12 173  456
6G. invermaium0.260.371097 5379
 None0.100.10 173   72
LSD0.05 0.010.04  80 2732
ANOVA, probabilities     
Soil 0.0000.000   0.000    0.001
G. invermaium 0.0000.000   0.000    0.000
Interaction 0.0000.000   0.000    0.001

P concentrations in shoots and roots of nonmycorrhizal plants in steamed Soils 1 and 2 were lower than the values recorded for the corresponding benomyl-treated soil (in Part (i)) (Tables 5 and 7). The highest uptake of 32P by G. invermaium-inoculated plants was observed in Soil 2 (Tables 5 and 7). This was twice the corresponding measurement for the untreated soil cores from the same site measured in Part (i). High 32P levels were also present in inoculated plants grown in Soils 1 and 6. The high P uptake by G. invermaium-inoculated plants grown in Soil 1 contrasted with the marginal level of uptake of 32P by plants grown in untreated cores of soil from this site in Part (i) (Table 5). In steamed soil, nonmycorrhizal plants contained only 1–3% of the 32P levels present in inoculated plants.

Discussion

This experiment demonstrated that the mesh bag technique previously applied to pots with disturbed soil, for example by Schweiger et al. (1999), has the potential to measure the P uptake by hyphae of naturally occurring AM fungi in intact soil cores taken from the field. Although the methodology is tedious, it could be used to assess the rate of uptake of 32P by hyphae from the root free soil. The study also highlighted the complex reasons for differences between soils in the functioning of naturally occurring AM fungi under field conditions.

Plant roots in intact cores of five pasture soils were colonized by the naturally occurring AM fungi to an extent which is normally assumed to benefit P uptake and plant growth but hyphal uptake of 32P markedly differed among soils. This variation was generally related to the length of hyphae present in the root-free mesh bags, but not to the percentage or length of roots colonized by AM fungi. The close relationship between hyphal P uptake and hyphal length density for the fungal communities did not support a previous conclusion that P uptake may depend more on the distribution of hyphae around roots than on length of hyphae (Jakobsen et al., 1992b).

The soil for which the greatest length of hyphae occurred had a dominant morphotype of AM fungus, a coarse-hyphaed Glomus sp., within roots that was different to that present in plants grown in the other soils. The effective 32P uptake in this soil might alone have been due to a more extensive growth of this morphotype, but a higher length-specific P uptake could also have played a role. The negligible uptake of 32P into plants grown in Soil 1 could have been due to the absence of hyphal growth into the mesh bags as judged from the hyphal length measurements. However, in some other soils, the benomyl or steaming treatment resulted in reduced uptake of 32P without any reduction in the measured hyphal length. Accordingly, the fungitoxic effect of benomyl is primarily linked to its influence on function of microtubuli and hence to transport processes (Davidse, 1986). The lack of an increase in hyphal length after inoculation was more difficult to explain for steamed Soils 1 and 6, but this corresponds with relatively low levels of hyphal growth of G. invermaium when inoculated into soil in a previous study (Abbott & Robson, 1985).

The very low level of P in Soil 1 might also have minimized the capacity of the naturally occurring fungi to take up P. However, when this soil was steamed and inoculated with an AM fungus, P uptake and plant growth were extensive. Further study is required to determine whether steaming released P from inorganic or organic pools in this soil, allowing the inoculant fungus to take up P more effectively than the naturally occurring fungi or whether steaming eradicated some adverse biotic components.

The lack of transport of P in the intact cores of Soil 1 was possibly related to the types of the fungi present and their capacity to form hyphae in this soil, but not to their infectivity. It is, therefore, interesting to note that the morphotype of AM fungus within the roots of plants grown in Soil 1 was similar to that of G.invermaium, the inoculant fungus. AM fungi with similar morphology have been shown to differ in their effectiveness (Abbott & Robson, 1978) but these effects could be compounded by differences in inoculum quantity and quality (Abbott & Robson, 1981). Furthermore, the time of sampling could have influenced between-site differences in 32P uptake by AM fungi because the composition of the fungi within roots changes seasonally (Scheltema et al., 1987; Rosendahl et al., 1989; Merryweather & Fitter, 1998b). A more detailed description of the rate and extent of colonization of roots and growth of hyphae may be necessary to understand the inconsistent observations for Soil 1.

Benomyl was successfully used to suppress P uptake of hyphae of AM fungi in root-free compartments buried in the field (Schweiger & Jakobsen, 1999). The benomyl treatment in the present work had marked effects on P uptake from the mesh bags, where it was thoroughly mixed throughout the soil, but it was insufficient to prevent colonization by AM fungi in soil cores due to its application to the soil surface. This problem of dispersing benomyl through undisturbed field soil has been identified previously (Pedersen & Sylvia, 1997). Based on previous research with this experimental system (Schweiger & Jakobsen, 1998), it was concluded that the hyphae in the mesh bags treated with benomyl were not functional. The low levels of 32P taken up in the benomyl treatments originated from uptake by root hairs penetrating the mesh bags and from diffusion into the soil cores. The observed variation in 32P uptake between the different benomyl-treated cores was probably due to differences in P adsorption capacity and rates of P diffusion out of the bags.

The experimental system used in this study has several advantages over other methods for assessment of the P transport effectiveness of communities of AM fungi in field soils. First, soil disturbance is low when intact cores are sampled from the field. This is of particular importance in perennial systems where intact mycelia and colonized root pieces are the major sources of new root colonization (Jasper et al., 1991; McGonigle & Miller, 1996). Second, the hyphal P transport is measured directly as 32P transport from a mesh bag excluding roots. Potential confounding effects of 32P diffusion into the surrounding soil or of root hairs entering the bag may be reduced by inclusion of a buffer zone of unlabelled soil around the labelled soil in the mesh bag (Schweiger et al., 1999). Root uptake of 32P in this experimental system can be reduced in control treatments by addition of benomyl to soil or by soil steaming. In future studies comparing different soils, the specific activity of 32P in extracts of the labelled soil should also be measured in order to calculate the total uptake of P by the hyphae. This will reduce problems in data interpretation caused by differences in isotope dilution between soils. The third advantage of this experimental system for assessing the P uptake capacity of field communities of AM fungi is that the labelling compartment (the root-free mesh bag) can be kept relatively small compared to the main compartment with plants. Inter-root distances in the field are typically as small as 0.2–1.0cm (Barber, 1984) and the AM fungal hyphae only need to transport P over these relatively short distances. Although hyphae may extend much further into the soil (Jakobsen et al., 1992a), their P uptake over the typical interroot distances will represent the most realistic field situation. The scale used in this study is, therefore, more appropriate than an experimental system with an adjacent large root-free compartment used in other investigations of P uptake by AM fungi (Li et al., 1991; Jakobsen et al., 1992a).

This research demonstrated that root-free mesh bags with 32P-labelled soil provide a feasible tool for assessing the difference in P transport effectiveness by naturally occurring communities of AM fungi in field soils. The application of this approach illustrated the complexity of functioning of AM fungi in field situations, in terms of P uptake by hyphae and its transfer to the plant. Soil characteristics, including P, can influence both the extents of both colonization of roots and hyphae formed in soil by fungi that may differ in their effectiveness at taking up P. However, simple relationships between growth of hyphae in soil, colonization of roots and consequences for P uptake do not occur under field conditions. Furthermore, changes in the community of AM fungi occur with time and time-course studies using this experimental system could help define the functioning of communities of AM fungi in field soils.

Acknowledgements

This research was made possible with support of a Grains Research and Development Visiting Fellowship to I. Jakobsen.

Ancillary