Correspondence: Franciska T. de Vries, Soil Science Centre, Wageningen University and Research Centre, PO Box 47, NL-6700 AA Wageningen, The Netherlands. Tel.: +31 317 486 631; fax: +31 317 419 000; e-mail: firstname.lastname@example.org
Soil biological studies are often conducted on sieved soils without the presence of plants. However, soil fungi build delicate mycelial networks, often symbiotically associated with plant roots (mycorrhizal fungi). We hypothesized that as a result of sieving and incubating without plants, the total fungal biomass decreases. To test this, we conducted three incubation experiments. We expected total and arbuscular mycorrhizal (AM) fungal biomass to be higher in less fertilized soils than in fertilized soils, and thus to decrease more during incubation. Indeed, we found that fungal biomass decreased rapidly in the less fertilized soils. A shift towards thicker hyphae occurred, and the fraction of septate hyphae increased. However, analyses of phospholipid fatty acids (PLFAs) and neutral lipid fatty acids could not clarify which fungal groups were decreasing. We propose that in our soils, there was a fraction of fungal biomass that was sensitive to fertilization and disturbance (sieving, followed by incubation without plants) with a very high turnover (possibly composed of fine hyphae of AM and saprotrophic fungi), and a fraction that was much less vulnerable with a low turnover (composed of saprotrophic fungi and runner hyphae of AMF). Furthermore, PLFAs might not be as sensitive in detecting changes in fungal biomass as previously thought.
Fungi are important soil organisms. In agricultural soils, saprotrophic fungi are – together with bacteria – the main decomposing organisms and thereby form the base of the food web; mycorrhizal fungi play an important role in plant nutrition. Although considerable research has been conducted on the ecology of fungi and their impact on soil structure and ecosystem processes, they remain an elusive group of microorganisms because of their filamentous life form (Klein & Paschke, 2004). Field studies have shown that fungi are sensitive to physical disturbance and fertilization (Beare et al., 1997; Frey et al., 1999; Johnson et al., 2005; De Vries et al., 2006, 2007). Laboratory experiments are a useful tool to investigate the impacts of fungi on ecosystem processes. However, because of the filamentous life form and high sensitivity to physical disturbance of fungi, these experiments might not represent the situation in the field, especially when the soil is sieved and incubated without plants. More specifically, sieving might destroy hyphal networks and thereby kill the most sensitive fungi, and mycorrhizal fungi might not survive during incubation without a host plant.
Arbuscular mycorrhizal fungi (AMF) can make up a substantial amount of the total fungal biomass in agricultural soils (Gosling et al., 2006). It has been suggested that AMF are more susceptible to disturbance, fertilization and disruption by soil invertebrates than saprotrophic fungi (Kabir et al., 1997; Johnson et al., 2005; Bradley et al., 2006). Ectomycorrhizal fungi, which constitute a large part of the total fungal biomass in forest soils, have been shown to decrease in the absence of a host plant (Frostegård et al., 1996; Siira-Pietikäinen et al., 2001; Bååth et al., 2004). Similarly, AMF might decrease in the absence of plants. Therefore, soil biological incubation studies are – as a result of sieving and incubating without plants – likely to induce changes in the relative contribution of saprotrophic and AM fungal biomass. Especially if soils from different systems are compared, for instance from more and less fertilized grasslands that differ in the biomass of saprotrophic and AMF, incubation will have a differential effect on fungal biomass in these soils.
Changes in fungal biomass as a result of sieving or incubation might take place rapidly, even within the first weeks of incubation. AM fungal hyphae have been observed to die off within 5–6 days (Staddon et al., 2003), although this has been criticized by Zhu & Miller (2003), stating that the main components of AM fungal hyphae, chitin and glomalin, have much longer residence times in the soil. Olsson & Johnson (2005) observed that AM runner hyphae were persistent, whereas the average longevity of the finely branched absorbing hyphae was only 5.3 days, which corresponds with earlier findings of Friese & Allen (1991) and Bago et al. (1998). Less information is available on the longevity of saprotrophic fungal hyphae. Schmidt et al. (2007) reported microbial turnover times in soil to be between 9 and 18 days, but they did not distinguish between bacteria and fungi. In their paper, they stress that turnover of microbial communities is an important missing piece in our understanding of nutrient cycling in terrestrial ecosystems.
Traditionally, total fungal biomass in soil has been quantified microscopically. AM fungal hyphae can be distinguished visually from hyphae of other fungi: in contrast to saprotrophic fungi, AM fungal hyphae generally have nonseptate hyphae and irregular cell walls and show angular, unilateral branching (Bonfante-Fasolo, 1986). However, distinguishing AMF from saprotrophic fungi on the basis of these morphological structures is a highly time-consuming method.
Most recent studies measure saprotrophic fungal biomass using ergosterol, which is a compound of the membrane of ectomycorrhizal and saprotrophic fungi, or the phospholipid fatty acid (PLFA) 18 :2ω6,9, which is only present in the cell membranes of ectomycorrhizal and saprotrophic fungi. PLFA 16:1ω5 has been used to quantify AM fungal biomass, but this fatty acid is also present in bacteria. Therefore, the neutral lipid fatty acid (NLFA) 16:1ω5, which is mainly present in AM fungal spores and therefore does not represent the active fraction, is a more specific indicator for the presence, but not always the biomass, of AMF (Olsson, 1999).
The aim of the present study was to follow the dynamics of fungal hyphae during short-term incubation without plants. We hypothesized that as a result of sieving and incubating without plants, the total fungal biomass decreases, and we expected this decrease to be caused primarily by AMF dying off. To test this hypothesis, we conducted three incubation experiments. In all three experiments, we compared a less or an unfertilized soil with a (more) fertilized soil. We expected the total and AM fungal biomass to be higher in the less or the unfertilized soils. Thus, we expected fungal biomass in the less or the unfertilized soils to decrease more than in the (more) fertilized soils.
Materials and methods
Two plots from an experimental field trial at Heino in the eastern part of the Netherlands (52°25′N and 6°15′E) were sampled. Both plots were sown with a grass–clover mixture (Lolium perenne L. and Trifolium repens L.) in 2001 and have since been fertilized with farm yard manure: one plot at a rate of 40 kg N ha−1 year−1, and the other plot at a rate of 80 kg N ha−1 year−1. The two plots were chosen on the basis of their widely differing fungal/bacterial biomass ratio (determined by microscopy): 0.19 for the plot fertilized with 80 kg N ha−1 year−1, vs. 0.69 for the plot fertilized with 40 kg N ha−1 year−1. The field trial was situated on a sandy soil; for a detailed description of the field trial, see De Vries et al. (2006).
A bulk soil sample, consisting of 100 cores (0–10 cm depth, 3.5 cm diameter), was taken from each of the two plots in November 2004. The two samples were sieved (5 mm), homogenized and stored at 4 °C for 1 week. The equivalent of 50 g of dry soil was weighed into a 250-mL serum bottle and adjusted to a moisture content of 25% (w/w). Flasks were closed with a cotton wool plug to prevent drying of the soil, but allow gaseous exchange. Flasks were weighed and rewetted to the initial moisture content weekly.
Of each soil, a control treatment without the addition of straw was incubated, as well as a treatment that received 2.5 mg milled wheat straw (C/N=137) per gram of dry soil (soil C content was 18.5 mg C g−1 and 20.5 mg C g−1 for the 80 kg N ha−1 year−1 plot and the 40 kg N ha−1 year−1 plot, respectively). The addition of the wheat straw was carried out for each experimental unit separately. The contents of the flasks were thoroughly mixed and incubated at 20 °C. The flasks were sampled destructively and fungal biomass was determined after 0, 1 and 8 weeks of incubation. Each treatment was replicated four times. We assessed the dynamics of fungal biomass using microscopy.
We conducted this experiment to test whether the observed decrease in fungal biomass in Experiment 1 would also occur in another soil type. Soil samples were taken in March 2005 from four plots of the Ossekampen Fertilizer Experiment in Wageningen. This trial was performed in 1958 on a heavy riverine clay soil. Two of the four plots have been unfertilized ever since (plots 13O and 16O: KCl, pH 4.0, 20.0% organic matter in the top 5-cm soil layer), and the other two plots have received 160 kg N, 74 kg P2O5 and 375 kg K2O ha−1 year−1 (plots 11NPK and 19NPK: KCl, pH 3.8, 23.7% organic matter in the top 5-cm soil layer) (for details, see Elberse et al., 1983).
From each plot, 50 cores (10 cm depth, 3.5 cm diameter) were bulked into one sample. After sieving and homogenizing, the moisture content was adjusted to 50% water holding capacity. Triplicate samples from each plot (100 g soil) were incubated in polyethylene bags at 12 °C (average soil temperature). Bags were sampled destructively and samples were analysed for fungal biomass after 0, 1, 2 and 4 weeks. As in Experiment 1, we assessed the dynamics of fungal biomass using microscopy.
This experiment was conducted to assess which fungal groups were causing the changes in fungal biomass observed in Experiment 2. The soil used for this experiment also came from the Ossekampen Fertilizer Experiment. In September 2005, we sampled the same plots as in Experiment 2. This time, duplicate samples from each of the four plots were incubated under the same conditions as in Experiment 2, and sampled destructively after 0, 1, 2 and 4 weeks. To assess the dynamics of the total fungal biomass and AMF, we analysed fungal biomarkers and, as in the previous two experiments, we used microscopy. We measured hyphal diameters in the unfertilized samples before and after incubation.
Measurements of fungal biomass, fungal biomarkers and hyphal diameters
Microscopic slides were prepared and hyphal length was measured by epifluorescence microscopy as described by Bloem & Vos (2004). In addition, using transmitted light, hyphae were categorized into blue-stained and melanized hyphae. Fungal biomass (C) was estimated from hyphal lengths and biovolume using the calculations described by Bloem & Vos (2004). Hyphal diameters were measured at × 1000 magnification. Of each slide, diameters of 100 fragments of blue-stained hyphae were measured and classified as septate or nonseptate.
Fungal biomass data of Experiment 1 were analysed by a three-way anova with the factors fertilizer level (40 and 80 kg N ha−1 year−1), straw amendment (straw and no straw) and incubation time (0, 1 and 8 weeks). Data of fungal and bacterial biomass and fungal biomarkers of Experiments 2 and 3 were analysed by a two-way anova with the factors fertilizer level (unfertilized and fertilized) and incubation time (0, 1, 2, and 4 weeks). Hyphal diameters were divided into classes and the differences were analysed by t-tests. All statistical tests were performed with the statistical package spss (SPSS Inc., Chicago).
At the start of the incubation, fungal biomass in the 40 kg N ha−1 year−1 field was three times higher than in the 80 kg N ha−1 year−1 field. Fungal biomass in the two soils converged rapidly within 1 week of incubation (Fig. 1). Fungal biomass was significantly affected by fertilization level and straw amendment (P<0.001 and P=0.05, respectively) – being the highest in the straw-amended treatments – but not by week (P=0.52). However, fungal biomass in the 40 kg N ha−1 fertilized soil responded differently to incubation than fungal biomass in the 80 kg N ha−1 fertilized soil (P=0.017 for the field × week interaction effect).
Fungal biomass in the heavy riverine clay soil from the Ossekampen Fertilizer Experiment was 10 times higher than fungal biomass in the sandy loam soil of Experiment 1. At the start of the incubation, fungal biomass in the unfertilized treatments (13O and 16O) was two times higher than in the fertilized treatments (11NPK and 19NPK) (Fig. 2a). Fertilization level and time had a significant effect on fungal biomass (P<0.001 for both fertilization level and time). An interaction effect of fertilization level and time was present (P=0.02), indicating that fungal biomass in the unfertilized treatments responded differently to incubation than in the fertilized treatments. Already after 1 week of incubation, fungal biomass in the unfertilized treatments decreased, whereas fungal biomass in the fertilized soil hardly decreased or remained constant. Between weeks 1 and 4, fungal biomass increased in all treatments. The effect of time on fungal biomass was significant in the unfertilized treatments (P=0.029), but only marginally so in the fertilized treatments (P=0.077).
In contrast to the total fungal biomass, melanized fungal biomass was higher in the fertilized treatment than in the unfertilized treatment (Fig. 2b, P=0.001). In both treatments, melanized fungal biomass was affected by time (P<0.001). The effect of time differed between the treatments (P=0.005 for the fertilization level × time interaction): melanized fungal biomass increased throughout the experiment in the fertilized soils, but culminated after 2 weeks in the unfertilized soils.
Microscopically measured fungal biomass (Fig. 3a, Table 1) and all fungal biomarkers were higher in the unfertilized treatments than in the fertilized treatments (Fig. 4a–c, Table 1). However, in the unfertilized soils, microscopically measured fungal biomass decreased during the first 2 weeks of incubation and then increased, whereas in the fertilized soils, it remained more or less constant. In all treatments, fungal biomarkers remained constant during the whole incubation period. Incubation time affected fungal biomass more strongly in the unfertilized treatments than in the fertilized treatments (Fig. 3a, P<0.001 for the interaction effect of fertilization level × time, Table 1). This difference in the effect of time between unfertilized and fertilized treatments was not present in any of the fungal biomarkers (Table 1). Biomass of melanized fungi was – unlike the total fungal biomass – higher in the fertilized treatments than in the unfertilized treatments (Fig. 3b, Table 1). In contrast to the previous experiment, the biomass of melanized fungi was not affected by incubation time (Table 1).
Table 1. Two-way anova summary statistics of the effects of fertilization level and incubation time on fungal biomass and biomarkers, in Experiment 3
Melanized fungal C
Fungal C, microscopically determined fungal biomass; melanized fungal C, microscopically determined biomass of melanized fungi; PLFA 18:2ω6,9, biomarker for saprotrophic fungi; NLFA 16:1ω5, biomarker for AM fungal spores; PLFA 16:1ω5, biomarker for AM fungal hyphae.
Fertilization × time
The distribution of hyphal diameters in the unfertilized treatments had shifted as a result of the incubation (Fig. 5). After 4 weeks of incubation, the frequency of hyphae with diameters smaller than 0.5 μm had decreased (P<0.001), while the frequency of thicker hyphae with diameters between 1 and 1.5 μm and between 1.5 and 2 μm had increased (P=0.045 and P=0.025, respectively). Percentage of septate hyphae was higher as a result of incubation: 27.5±1.73 after incubation vs. 18.7±5.68 before incubation (P=0.013).
We aimed to monitor the dynamics of fungal hyphae in short-term incubation experiments without plants. We hypothesized that the total fungal biomass would decrease as a result of sieving and incubating, and that the total and AM fungal biomass would be higher in the less or the unfertilized treatments than in the fertilized treatments. This was the case for microscopically measured fungal biomass in all three experiments. Also, biomarkers for saprotrophic and AMF were the highest in the unfertilized treatments of Experiment 3. As we hypothesized, fungal biomass decreased during the incubation, but only in the less or the unfertilized treatments. This decrease set in rapidly: either after 1 week (Experiments 1 and 2) or after 2 weeks of incubation (Experiment 3).
We observed a large difference in fungal biomass between the sandy soil and the riverine clay soil. Fungal biomass in the clay soil was 10 times higher than in the sandy soil. As organic matter and clay content have been reported to be positively correlated to (saprotrophic) fungal biomass (e.g. Van der Wal et al., 2006), these factors may explain the difference in fungal biomass between the two soils.
We expected AM fungal hyphae to decrease more than saprotrophic fungal hyphae. To test this, we measured AM fungal biomarkers and saprotrophic fungal biomarkers. However, all fungal biomarkers remained constant during the incubation. Nonetheless, percentage of septate hyphae increased during the experiment, indicating that septate hyphae were less affected by the incubation than nonseptate hyphae. Because AMF are nonseptate (Bonfante-Fasolo, 1986), this could indicate that AMF were dying off during the first weeks of incubation. This rapid decline of (presumably AM) fungal biomass is remarkable, since Bååth et al. (2004) performed incubation for 10 months to estimate ectomycorrhizal fungal biomass by the decline in the total fungal biomass (measured by ergosterol and PLFA 18:2ω6,9, which are both present in ectomycorrhizal fungi and in saprotrophic fungi).
Melanized fungal hyphae did not decrease during the experiment. Melanins in fungal hyphae are associated with protection from environmental stress and microbial degradation (Bell & Wheeler, 1986; Butler & Day, 1998). Either melanized fungi were not affected by the incubation, or their hyphae were too recalcitrant to decompose during the incubation. Although AM fungal spores can be melanized (Purin & Rillig, 2008), AM fungal hyphae are not. From the distribution of hyphal diameters before and after incubation, we can conclude that after 4 weeks, almost all blue-stained hyphae with a diameter smaller than 0.5 μm had disappeared (Fig. 5). These thin hyphae might represent branched absorbing structures (BAS) of the extraradical mycelium of AMF (Bago et al., 1998). Third-order BAS branches can be very thin and are known to have a lifespan between 2.5 and 10 days (Bago et al., 1998). However, the hyphae that disappeared in our experiment are much thinner than the 2 μm that Bago and colleagues reported. The observed shift towards thicker hyphae could just denote the higher decomposability of thinner hyphae and might not indicate the dying off of a specific group of thin fungi.
The discrepancy between the microscopic measurements and the concentrations of fatty acids is noteworthy. Although both methods showed comparable differences between treatments at the start of the incubation, microscopy indicated decreasing fungal biomass while PLFAs remained constant throughout the incubation. This is surprising, because PLFAs are assumed to decompose rapidly after cell death – although actually there is only one reference for this (White et al., 1979). However, in a study on heating effects on peat, PLFAs degraded more slowly at low temperatures, which was attributed to slower enzymatic reactions (Ranneklev & Bååth, 2003). Therefore, the low incubation temperature of Experiment 3 (12 °C, which is the average annual soil temperature) might explain why PLFAs did not disappear in the current experiment.
Another explanation for the discrepancy between microscopy and PLFAs could be that the decrease in fungal biomass was mainly caused by the disappearance of thin hyphae. Because PLFAs are a compound of the cell membrane, the amount of PLFAs would be linearly related to hyphal surface, and thus increase with a factor four if hyphal diameter increases by a factor two. The high contribution of thick hyphae towards PLFAs could mask the decrease of thin hyphae. Especially melanized hyphae, which did not disappear during the incubation, and that can be up to 10 times thicker than blue-stained hyphae, might have had a relatively high contribution to the amount of PLFAs. In addition, much of the carbon present in thin hyphae can be translocated into storage structures and thick hyphae (Bago et al., 2002; Olsson & Johnson, 2005). However, because most PLFAs are found in nonmobile cell membranes, translocation might not explain why PLFAs did not disappear at the same rate as fungal hyphae.
Microscopic counts have been criticized, because they do not accurately represent the living fraction of fungal biomass (Frankland, 1975). Because hitherto fungal hyphae have been assumed to degrade slowly in soil, microscopic counts would not be suitable for detecting negative treatment effects. In the current experiment, we demonstrated that microscopically determined hyphal lengths can reflect changes in fungal biomass even more rapidly than analysis of fungal PLFAs.
The rapid degradation of fungal hyphae we found has never been reported before. The rapid decrease of fungal hyphae is in line with a recent – although criticized (Zhu & Miller, 2003; Olsson & Johnson, 2005) – study by Staddon et al. (2003), but is in contrast with the general assumption that fungal hyphae degrade slowly in soil. This assumption is based on the relatively recalcitrant components present in the fungal cell wall (Martin & Haider, 1979; Kassim et al., 1981), and on protection from degradation through interaction of fungal hyphae and soil aggregates (Simpson et al., 2004). As we discussed above, nonmelanized fungal hyphae might degrade more rapidly than melanized fungal hyphae. In addition, after sieving the soil, as we did in the current incubation experiments, the aggregate structures that protect fungal hyphae from degradation will be disrupted.
In summary, microscopic counts showed that during short-term incubation, part of the fungal biomass died off. As we hypothesized, fungal biomass was higher in the less or unfertilized treatments, and these treatments also showed the largest decrease in fungal biomass. Although there were some indications, we could not confirm that AMF were decreasing more than saprotrophic fungi. We propose that in our soils, there was a fraction of fungal biomass that was sensitive to disturbance and fertilization with a very high turnover, and a fraction that was much less vulnerable with a low turnover. The first fraction possibly consisted of fine hyphae of both AMF and saprotrophic fungi; the second fraction possibly consisted of saprotrophic fungi and runner hyphae of AMF. Our results show that incubation or pretreatment of soil samples can rapidly nullify long-lasting differences in fungal biomass between unfertilized and fertilized soils. Furthermore, PLFAs might not be as sensitive in detecting changes in fungal biomass as previously thought.
Thanks are due to Rob Dijcker, who carried out Experiment 1, to Nick van Eekeren (Louis Bolk Institute, Driebergen) for allowing us to sample Aver Heino, and to Rob Geerts (Plant Research International, Wageningen) for allowing us to sample the Ossekampen Fertilizer Experiment. Thanks are also due to An Vos and Meint Veninga for help in the laboratory. We thank Ellis Hoffland and Lijbert Brussaard for helpful comments on the manuscript.