• A large-scale tree-girdling experiment enabled estimates in the field of the contribution of extramatrical mycelium of ectomycorrhizal (ECM) fungi to soil microbial biomass and by ECM roots and fungi to production of dissolved organic carbon (DOC).
• Tree-girdling was made early (EG) or late (LG) during the summer to terminate the flow of photosynthate to roots and ECM fungi. Determination of microbial C (Cmicr) and microbial N in root-free organic soil was performed by using the fumigation–extraction technique; extractable DOC was determined on unfumigated soil.
• Soil Cmicr was 41% lower on LG than on control plots 1 month after LG, whereas at the same time (that is, 3 months after EG), the Cmicr was 23% lower on EG than on control plots. Extractable DOC was 45% lower on girdled plots than control plots.
• Our results, which are of particular interest as they were obtained directly in the field, clearly demonstrate the important contribution by extramatrical ECM mycelium to soil microbial biomass and by ECM roots to the production of DOC, a carbon source for other microbes.
Biogeochemists, ecologists and soil microbiologists can be divided into those that neglect and those that recognize the role of mycorrhizal fungi in ecosystems. A reason for this division is the problem of assessing the contribution of mycorrhizal fungi to the soil microbial community, especially in the field. More precise estimates are available on the carbon (C) cost of ectomycorrhizal (ECM) symbiosis in laboratory model systems (Rygiewicz & Anderson, 1994) but only rough calculations are available for forest ecosystems (Söderström, 1992; Smith & Read, 1997). To date, it has not been possible to quantify the contribution made by ECM fungi to total microbial biomass in the soil, despite the need to distinguish these symbiotic fungi from other mycorrhizal fungi and from saprotrophic fungi and other decomposers. However, the phospholipid fatty acid (PLFA) 18 : 2ω6,9 and ergosterol fungal biomarkers were monitored inside and outside cores (2.0 dm2) that were root isolated (Wallander et al., 2001). The root isolation killed ECM fungi inside trenches and Wallander et al. calculated the ECM biomass by subtracting the amount of fungi outside, from the values obtained from inside trenched cores. The ECM biomass, including that of fungal mantels, was calculated to be approx. 800 kg ha−1. The amount of ECM mycelium produced in mesh bags filled with quartz sand, an inert substrate mainly colonized by mycorrhizal fungi, was approx. 160 kg ha−1. It is unclear if this estimate applies to the situation in the organic mor-layer, which is the horizon of greatest biological activity in the boreal forests.
Ectomycorrhizal fungi use C from their plant hosts, and it seems unlikely that the fungi ensheathing these plant roots leak much C to the surrounding soil. However, there is evidently a flora of bacteria on the surfaces of ECM hyphae (Garbaye, 1994; Timonen et al., 1998), and several studies report that ECM fungi and roots produce large amounts of certain organic acids (Griffiths et al., 1994; Wallander et al., 1997). The contribution made by ECM roots and ECM fungi to the production of dissolved organic C (DOC) in the mor-layer of boreal forests, where concentrations of DOC are high (van Hees et al., 2000), is not known. This gap in our knowledge is particularly serious given that DOC affects rates of weathering of soil minerals (Lundström et al., 2000) and represents important sources of C for microbes.
In a recent large-scale tree-girdling experiment (Högberg et al., 2001) a 50% loss of soil respiration was interpreted as a loss of activity by ECM roots and their extramatrical mycelium. Here, we make use of this unique field experiment to quantify the contribution by the extramatrical mycelium of ECM fungi to total soil microbial biomass in the organic mor-layer. We also used this experiment to estimate the production of extractable DOC by ECM roots and fungi.
Materials and Methods
Field site, experimental design and soil sampling
The forest was a naturally regenerated 45- to 55-year-old Scots pine (Pinus sylvestris L.) located on a weakly podzolized sandy silt sediment at Åheden, northern Sweden (64°14′ N, 19°46′ E, 175 m above sea level). The climate is cold with a mean annual temperature of 1.0°C, and a mean annual precipitation of 600 mm. There is usually snow cover for 6 months between late October and early May. There was a sparse understorey of Calluna vulgaris L. and Vaccinium vitis-idaea L. The bottom layer consisted of mainly Cladonia spp. lichens and Pleurozium schreberi moss. The organic mor-layer (F + H horizons) was 2 cm thick and had the following characteristics (n = 9, mean ± SD): bulk density 0.16 ± 0.05 g cm−3, C : N ratio 40 ± 5, organic matter content (weight loss on ignition) 76 ± 5%, pHH20 4.0 ± 0.1, water content at the time of sampling 193 ± 41% (g g−1 dry wt).
The experiment comprised nine quadratic plots of 900 m2 each (with c. 120 trees each) and was divided into three blocks (Fig. 1). Girdling was performed in early June 2000 (early girdling, EG) and in mid-August 2000 (late girdling, LG) on three plots at a time leaving three plots as control plots. Girdling had no effects on soil temperature and moisture (Högberg et al., 2001). Seventy-two days after girdling, the number of sporocarps and their biomass were reduced by 98.4% and 99.4%, respectively, on the central 100 m2 of EG plots compared with control plots.
On 12 September 2000, soil from the F and H horizons was sampled by use of a 0.1-m diameter corer. Sampling was performed along the border of the central 100 m2 of each plot. Five composite samples made up from 10 cores each were taken from each plot.
Fumigation and determinations of Cmicr and DOC
Soil samples were stored at 4°C overnight. Roots were thereafter sorted out by hand. After a day at 16°C, microbial C, Cmicr, and microbial N, Nmicr, were determined by the fumigation–extraction (FE) method (modified from Brookes, 1985a,b; Vance et al., 1987). Approximately 12 g (w : w) root-free soil was put into each of 45 50 ml glass beakers, which were placed in a desiccator (18 dm3 volume). Forty-five millilitres of ethanol-free CHCl3 (Lichrosolv, Merck no. 2444, Merck KGaA, Darmstadt, Germany) was used as fumigant (22°C, 20 h). At the same time as the fumigation process was started, the nonfumigated soil was shaken (150 rev min−1) for 30 min with 50 ml 0.5 m K2SO4 (mean soil : solution ratio = 1 : 13, w : v) and filtered (Munktell 00H filters (equivalent to Whatman no. 42), Munktell Filter AB, Grycksbo, Sweden). The fumigated soil was extracted as described above after removal of the CHCl3 from the soil by repeated evacuations. The extracts were kept frozen at −30°C before analysis.
Extracts were analysed for total organic C on a TOC-5000 (Shimadzu Corporation, Kyoto, Japan): the organic C component was combusted to CO2 at 680°C and detected on an infrared gas analyser. Extractable DOC was determined as total organic C in extracts from nonfumigated soil. The sum of organic N and NH4-N in the K2SO4 extracts was determined as NH4-N at 590 nm by flow injection analysis (FIAstar, FOSS TECATOR, Höganäs, Sweden) after preincubation and micro-Kjeldahl digestion (Wyland et al., 1994). The Cmicr and Nmicr were obtained after correcting for the efficiency of extraction of microbial biomass C and N, respectively. Cmicr was calculated as Cmicr = Cf/kEC, where Cf is (organic C extracted from fumigated soil) − (organic C extracted from unfumigated soil) and kEC is 0.4. The Nmicr was calculated as Nmicr = Nf/kEN, where Nf is (organic N extracted from fumigated soil) − (organic N extracted from unfumigated soil) and kEN is 0.4. These values of kEC and kEN were from a similar soil in a Finnish Pinus sylvestris forest of C. vulgaris type and were calibrated by microscopic counting (Martikainen & Palojärvi, 1990). Soil dry weight was determined after drying at 105°C for 24 h. The organic matter content was determined by loss on ignition (600°C, 4 h) and pH was measured in water (soil : solution ratio = 1 : 5, v : v).
Statistical analyses were performed using sigmastat 2.0 (SPSS Science, Chicago, IL, USA). Effects of treatments and blocks on Cmicr, Nmicr, microbial C : N and DOC were tested by two-way anova using the mean values for each plot. If a significant effect (P < 0.05) was found, Tukey’s post hoc test was performed to test for significant differences among treatments and blocks.
Results and Discussion
Microbial C and N contributed 1.6% to soil organic C and 6.6% to soil organic N, respectively, on control plots. These figures are in agreement with the mean values of 1.8% for C and 8.5% for N given for a similar Finnish forest soil (Martikainen & Palojärvi, 1990).
Microbial C was 41% lower on LG plots than on control plots (P < 0.05), while on EG plots it was 23% lower than on control plots (difference was nonsignificant) (Fig. 2a). For Nmicr, there were no differences among treatments (Fig. 2b). However, the C : N ratios of the soil microbial biomass were significantly lower (P < 0.05) on both EG and LG than on control plots (Fig. 2d). In this case, there was also a significant block effect.
In the girdling experiment, in which up to 56% of soil respiration was lost during the first year after girdling (Högberg et al., 2001), measured total soil respiratory activity on control plots included that of ECM roots and their fungal sheaths in addition to that of the root-free soil studied here. In root-free soil, there should be respiratory activity by the extramatrical ECM mycelium, extramatrical ericoid mycorrhizal mycelium, saprophytic fungi, bacteria and other soil organisms. In this study, the extramatrical ECM mycelium is the major functional component, along with mycorrhizosphere organisms, that could be negatively affected by the girdling. Conversely, the activity and growth of the other organisms could be enhanced because they could use the dying ECM mycelium as a substrate and/or benefit from a release from competition for space and nutrients. Thus, based on the average loss of Cmicr in the treatments EG and LG, at least 32% of the soil microbial biomass was contributed by extramatrical ECM mycelium. This contribution was calculated to be equivalent of 145 kg ha−1, corresponding to 58 kg C ha−1 at a fungal biomass carbon content of 40% by dry weight.
Potential changes in microbial respiratory activity, growth and community composition could have been going on for 3 months in EG plots compared with only 1 month in LG plots. This difference in time since girdling may help to explain the more drastic decline in Cmicr in LG plots. Therefore, the 41% loss of Cmicr in the LG treatment may be the more relevant estimate of ECM biomass. The difference in biomass between EG and LG plots could also relate to seasonality. For respiration, the highest calculated contribution by ECM roots and mycelium was found in late summer (Högberg et al., 2001), which is in line with observations on C allocation patterns in temperate conifers (Hansen et al., 1997). This means that the LG treatment was conducted when the ECM fungal biomass would be expected to be greatest. Several studies show a seasonal pattern in fungal biomass in forest soils. High values for fluorescein diacetate (FDA)-active fungi were found in early spring and autumn (Söderström, 1979; Bååth & Söderström, 1982) and growth of extramatrical mycelium of ECM fungi was highest in the autumn (Wallander et al., 2001). At the time of this study, the respiratory activity was 37–39% lower in EG and LG plots than in control plots, which suggests a rough correlation between biomass and respiration.
The lower microbial C : N ratios of 7.1 and 6.2 in the EG and LG soils, respectively, compared with 8.9 for the control soil, may reflect a lower abundance of ECM fungi (lower fungi : bacteria ratio), since the C : N ratio in bacteria is mostly lower than in fungi (Paul & Clark, 1996). Alternatively, there are no changes in the microbial community with respect to the abundance of fungi and bacteria. Thus, the lower C : N ratios are simply the results of the same amounts of N being associated with smaller amounts of C.
Levels of extractable DOC were 49% and 41% lower on EG and LG plots, respectively, than on control plots (Fig. 2c). At the same time, there were no differences in extractable dissolved organic nitrogen (DON) among treatments (data not shown). This suggests the loss of organic DOC with a high C : N ratio of 51. It would be of considerable interest to know the molecular composition of the DOC in the different treatments, since low molecular weight organic acids are thought to play a major role in the mineral weathering by ECM fungi (van Breemen et al., 2000) and as DOC comprises potentially important C sources for other microbes.
We suggest that our estimate of a 32% contribution by extramatrical ECM mycelium to the total microbial biomass is a conservative one, primarily because the dead ECM mycelium could be used as a substrate by other organisms. The particular strength of our estimates of the contribution of ECM fungi to microbial biomass and by ECM roots and fungi to the production of DOC is that they relate to an organic soil in a field setting, in which the only major manipulation of the studied system is the removal of the C source of the functional group of interest. Högberg et al. (2001) demonstrated the vital importance of the flux of current photosynthates to ECM roots for soil respiratory activity. Our results show that this flux similarly directly supports a considerable portion of the soil microbial biomass and is of utmost importance for the production of DOC.
We thank Birgitta Olsson and Bengt Andersson for conducting the FIA and TOC analysis, respectively. This study was supported by grants from the Swedish Council for Forestry and Agricultural Research (SJFR), the EU (project FORCAST) and the Swedish Natural Sciences Research Council (NFR).