Citation
Wal, A. van der. (2007, October 24). Soils in transition: dynamics and functioning of fungi.
Retrieved from https://hdl.handle.net/1887/12412
Version: Corrected Publisher’s Version
License: Licence agreement concerning inclusion of doctoral thesis in the Institutional Repository of the University of Leiden
Downloaded from: https://hdl.handle.net/1887/12412
Note: To cite this publication please use the final published version (if applicable).
Fungal biomass development in a chronosequence of land abandonment
Annemieke van der Wal, Johannes A. van Veen, Wiecher Smant, Henricus T.S. Boschker, Jaap Bloem, Paul Kardol, Wim H. van der Putten and Wietse de Boer
Published in: Soil Biology & Biochemistry 38 (2006) 51-60
Abstract
Based on biomass size, the contribution of fungi to nutrient cycling and soil properties is in general more important in natural ecosystems than in agro-ecosystems. Therefore, we expect an increase of fungal biomass after cessation of cultivation to values of a natural ecosystem.
However, so far, information on fungal dynamics in ex-arable land is limited. We quantified fungal biomass in a chronosequence of 26 ex-arable fields in the Netherlands ranging from 1- 34 years of abandonment. Agricultural lands and semi-natural heathlands were included as reference sites for initial and final stages of succession, respectively. Fungal biomass values were low at the start of land abandonment and increased during the first 2 years after abandonment. After this initial increase of fungal biomass no further increase was apparent, neither did we find any relations with time since abandonment and changes in soil acidity, organic matter content or organic matter quality (quantity of recalcitrant C and C to N ratio).
Therefore, we conclude that the initial increase of fungal biomass is caused by stopping agricultural management activities. A phase of stabilization occurs for at least three decades in which the size of the fungal biomass did not change significantly. We observed much higher values for fungal biomass, total and recalcitrant carbon in the heathland sites. We propose that a change in abiotic soil properties is a prerequisite for further increase of fungal biomass towards levels of representative heathlands.
Introduction
Fungi can play an important role in soil ecosystem processes, particularly in nutrient cycling, as they are able to decompose recalcitrant organic matter and transport nutrients from mineral layers to the litter layer (Frey et al., 2000; Klein and Paschke, 2004). Fungi also contribute to soil structures and thus to moisture retention and to resistance of soil against erosion (Tisdall and Oades, 1982; Beare et al., 1997). In addition, they influence vegetation dynamics by acting as pathogens (Van der Putten, 2003) and as mutualistic symbionts (van der Heijden et al., 1998; Allen et al., 2003).
Agricultural soil management practices and grassland intensification result in a decrease of the fungal biomass (Bardgett et al., 1996; Bardgett and McAlister, 1999; Frey et al., 1999; Stahl et al., 1999). This has been ascribed to physical disruption of the hyphal network by tillage (Wardle, 1995; Helgason et al., 1998), the negative influences of fertilization on the arbuscular mycorrhizal colonization of plant roots (Jasper et al., 1979;
Kahiluoto et al., 2001) and the lack of litter layers development with lignocellulose-rich material which forms the substrate for specialized saprotrophic basidiomycetes (Carlile et al., 2001). In contrast, bacterial abundance appears to be less affected by agricultural management (Bardgett et al., 1996; Frey et al., 1999).
As a result of the different effects of agricultural management on fungal and bacterial biomass, the fungal: bacterial ratio is usually substantially lower in agricultural soils than in more natural soils (Yeates et al., 1997; Bardgett and McAlister, 1999; Bailey et al., 2002).
Hence, it appears that fungi do not have an important contribution to soil ecosystem processes in intensively managed agricultural land, whereas this role is much greater in many natural ecosystems. Following land abandonment, fungi are expected to become an increasing component of the soil microbial biomass and their contribution to soil ecological processes relative to that of bacteria is supposed to increase over time. However, little information is available on the course and recovery of the fungal biomass during secondary succession of ex-arable land. The fungal biomass in semi-arid steppe soils increased with time after cessation of cultivation towards a maximum in the oldest (38 year abandoned) ex-arable site (Klein et al., 1995). Total fungal biomass of a native reference site was in the range of fungal biomass in fields between 12 and 38 years following abandonment, implying a fungal recovery time of at least 12 years. In an other study, re-establishment of ectomycorrhizal infectiveness required 25-30 years of abandonment (Boerner et al., 1996). Recovery of arbuscular mycorrhizal fungal propagules took even longer, as levels comparable with an uncultivated soil were only reached after 45 years (Roldan et al., 1997).
During land transition, soil organic matter, as well as the C: N ratio of the organic matter will increase (Knops and Tilman, 2000; Zeller et al., 2001) and the pH of the soil will decrease, which is assumed to favor fungal growth (Swift et al., 1979; Zeller et al., 2001).
However, these changes are slow and may take several centuries (Burke et al., 1995; Knops and Tilman, 2000). If soil pH, quantity and quality of organic matter are slowly recovering to levels of a natural ecosystem and form the major conditions for fungal growth, than a slow, gradual increase of fungal biomass is to be expected. On the other hand, if agricultural soil management is the major constraint for fungal development, a relatively quick recovery of fungal biomass might be realized. Conclusions on the relative importance of soil properties versus the absence of soil management on fungal biomass development in abandoned fields can not be drawn from the previous mentioned studies as they cover very few, wide time intervals and often relate to only specific fractions of the fungal community.
The aim of this study was to assess the development of fungal biomass in abandoned, formerly intensively managed agricultural soils. The specific questions addressed in this study were: 1) Is there a relatively quick recovery of fungal biomass in ex-arable land to values of representative natural reference sites? 2) To what extent is fungal biomass development related to abiotic soil properties?
For this study, 26 ex-arable fields were selected composing a chronosequence of land abandonment. We applied a multiphasic approach to avoid biases of individual methods for the assessment of fungal and bacterial biomasses. For instance, direct assessment of fungal biomass on the basis of hyphal lengths measurement assumes that all hyphae are filled with cytoplasm, whereas it is known that hyphae can be empty and non-active under field conditions (Klein and Paschke, 2004). Therefore, we used two other methods as well (e.g.
fungal PLFA and ergosterol) to estimate the more active part of the fungal biomass in soil (Federle, 1986; West et al., 1987).
Materials and methods
Soils and sampling procedure
We selected 26 ex-arable fields that were abandoned at different time intervals to establish a chronosequence of 34 years of land abandonment. All fields were managed by low-intensive grazing of natural and introduced herbivores (roe deer, fallow deer, red deer, horses, Scottish Highland cattle) to prevent forest development and had similar restoration targets. In addition, three agricultural fields were chosen to represent the starting point of the chronosequence whereas three semi-natural heathland sites, which have not been managed as agricultural land,
were taken as reference sites. The reference sites were chosen based on the targets for nature restoration of the ex-arable fields set by nature managers, i.e. heathlands. All fields are located on the same parental soil material of glacial sandy deposits in the central part of the Netherlands (Veluwe). In April 2003, soil samples were collected from the 32 fields. For each field, a composite soil sample was made of eighty soil cores of 3.5 cm diameter and 10 cm deep that were collected according to a stratified random pattern from a 50 x 50 m plot and bulked within each field. Samples were sieved (4 mm mesh) and stored at 4˚C for not more than 1 week until analyses. Soil samples for the PLFA measurements were freeze dried immediately after sampling. Soil characteristics and agricultural history of the fields are listed in Table 1. Using vegetation data as determined by Kardol et al. (2005), the plant association of each field was appointed using the program SynBioSys (Hennekens and Schaminee, 2001).
Determination of fungal and bacterial biomass
Microscopic slides were prepared as described by Bloem (1995) for the determination of hyphal length and bacterial numbers. Fungal hyphae in the soil suspension were stained with DFS solution (34.8 mg europium (III) thenoyltrifluoroacetonate and 0.5 mg calcofluorwhite fluorescent brightener in 10 ml of 95% ethanol). Bacteria were stained with DTAF (5-(4,6- dichlorotriazin-2-yl)aminofluorescein). Fungal hyphal length (active + inactive) was measured using a Leitz epifluorescence microscope at 10 x 25 magnification. Melanized (brown) hyphae were viewed with conventional light microscopy. Total hyphal length was estimated by the line intercept method (Olson, 1950) as applied by Bloem (1995). The equation V= (π/4) W2 (L-W/3) where V= volume (μm3), L= length (μm) and W= width (μm) (Krambeck et al., 1981), a mean hyphal diameter of 2.5 μm and a specific carbon content of 130 fg C μm-3 (van Veen and Paul, 1979; Bakken and Olsen, 1983) were used to estimate fungal biomass. Bacterial biomass was calculated using a specific carbon content of 320 fg C μm-3 and bacterial cell numbers and volume were determined by confocal laser scanning microscopy combined with an image analysis system (Bloem et al., 1995).
Phospholipid (PLFA) and neutral (NLFA) lipid fatty acids were extracted and analyzed by chromatography (Boschker, 2004; Hedlund, 2002). Briefly, fresh soil samples were sieved (2mm mesh) and freeze dried before analysis. The soil was extracted using one- phase chloroform-methanol-buffer extractant. Lipids were dissolved in chloroform and subsequently separated into neutral lipids, glycolipids and phospholipids using a silicic-acid column. Neutral and phospholipids fractions were collected and transesterified with mild alkaline to recover the lipids as methyl esters in hexane. Next, the lipid fractions were separated on a gas chromatograph with a flame ionization detector and a a-polar HP 5MS
column (55 m length, 0.25 mm diameter, 0.25 µm film). Peak areas were calculated relative to the internal standard methyl nonadecanoate fatty acid (19:0) which was added before methylation. The following fatty acids were summed up to estimate bacterial biomass: i15:0, a15:0, 15:0, i16:0, 16:1ω9, 16:1ω7t, i17:0, a17:0, 17:0, cy17:0, 18:1ω7 and cy19:0 (Frostegård and Bååth, 1996). PLFA 10Me16:0, 10Me17:0 and 10Me18:0 were used to calculate the biomass of actinomycetes (Frostegard et al., 1993). The PLFA 18:2ω6 was used as marker for fungal biomass (Federle, 1986). Arbuscular mycorrhizal fungal biomass was estimated from NLFA 16:1ω5 (Hedlund, 2002).
Ergosterol, a sterol only found in fungal cell membranes, was used as a fungal-specific biomarker and extracted and quantified as described by HPLC analysis as in Bååth (2001).
Accumulation of an easily extractable pool of the protein glomalin, a protein that is produced by arbuscular mycorrhizal (AM) fungi (Wright et al., 1996), was used as an indicator for the presence of (dead) AM-fungi in the plots (Driver et al., 2005). Easily extractable glomalin was extracted as described by Wright and Upadhyaya (1996), where air-dried sieved (2 mm mesh) replicate 0.5 g soil samples were extracted with 20 mM sodium citrate, pH 7.0 at 121˚C for 30 min. After extraction, the samples were centrifuged at 10.000 x g for 5 min. and protein content in the supernatant was measured spectrophotometrically at 595 nmby the Bradford dye-binding assay with bovine serum albumin as standard.
Soil microbial activity
Total soil microbial activity was determined by measuring the amount of CO2 production from 40 gram fresh soil incubated for 48 hours at 20˚C. To avoid elevated CO2 release caused by soil disturbance during the weighing procedure, serum bottles were sealed with parafilm after filling and incubated for 24 hours at 4˚C. Prior to the incubation period, bottles were flushed with air to remove excess CO2 and capped with a gas-tight septum. CO2
concentrations (1 ml headspace gas) were analyzed using a gas chromatograph (Carlo Erba GC 6000) equipped with a hot wire detector (HWD 430). Helium was used as a carrier gas.
Physical and chemical analyses
Soil water content was determined as weight loss after drying overnight at 105˚C. Soil mineral N was extracted by (10 g dry weight) shaking with 50 ml 1 M KCl for 2 hours.
Concentrations of NH4+-N and NO3--N in the KCl extract were determined colorimetrically using a Traacs 800 auto-analyzer. The pH was measured in 1: 2.5 (dry weight) soil: water suspensions. Total organic C was determined by the Walkley-Black potassium dichromate- concentrated sulfuric acid oxidation procedure (Nelson and Sommers, 1982). Content of total N and total P were measured by digestion of samples with a mixture of H2SO4-Se and
salicyclic acid (Novozamsky et al., 1984). Calcium chloride-extractable P was determined in 0.01 M CaCl2. The quality of organic matter was assessed by fractionating soil organic matter into three pools indicated by labile I (extraction with 5 N H2SO4; mainly free carbohydrates), labile II (extraction with 26 N H2SO4 ; mainly cellulose) and a recalcitrant pool (residual material after digestion) (Rovira and Vallejo, 2002). Soil texture was determined by sieving through <53, 75, 106, 150, 212, 300, 425, 850 and 1400 μm mesh sieves.
Statistical analyses
The relation between fungal biomass and time since abandonment were calculated by the coefficient of the linear equation at P = 0.05 using STATISTICA (release 6.1, Statsoft, Inc.).
The agricultural fields were considered as starting points of the chronosequence. As it was not possible to assign a period of abandonment to the reference sites, these fields were excluded from the regression analysis.
The mean of each soil- and microbiological property was calculated in order to compare soil characteristics of the agricultural fields and the 1st year ex-arable fields with all other ex-arable sites, and to compare these ‘old’ ex-arable sites with the semi-natural reference sites. A 95% confidence interval was computed for the mean of each variable from the ex-arable fields and the three semi-natural reference sites using STATISTICA (release 6.1, Statsoft, Inc.). The two computed confidence intervals were examined for overlap to observe similarities between agricultural fields and the 1st year ex-arable fields, ‘old’ ex- arable fields and natural sites.
To analyze the response of the fungal and bacterial biomass to soil properties, a redundancy analysis (RDA) was performed by means of Canoco version 4.5 (ter Braak and Smilauer, 2002). Redundancy analysis is a principal component analysis (PCA) with a restriction on the site scores by multiple regression on the environmental variables (ter Braak, 1995). A large reduction in explained variance from the axes in a RDA compared to a PCA is thus indicating that the environmental variables measured do not give a reliable estimate of the relation between microbiological properties and soil properties. Scaling is focused on inter-sample distances. Species scores, in this case microbiological parameters, were divided by standard deviation. Loam fraction was used as covariable in addition to the environmental variables to adjust for soil structure differences between fields. Centering was used for microbiological parameters. Monte Carlo permutation test (499 permutations under reduced model) was performed to test the significance of the eigenvalue related to the first RDA canonical axis. Since two methods were used for estimating bacterial biomass and three methods for fungal biomass, bacterial data and fungal data were attributed a weight of respectively three and two.
Results
Fungal biomass development in relation with time since abandonment
Although a significant and positive relation between microscopically determined fungal biomass and the period of abandonment seemed to occur when all arable and ex-arable sites were included in the regression analysis (R2 = 0.26, n = 29), the significance of the relation was mostly due to the low fungal biomass values in the agricultural fields and the 1st year ex- arable fields (Fig. 1A). Fungal biomass did not significantly increase anymore with the period of abandonment when the agricultural fields and the 1st year ex-arable fields were excluded (R2 = 0.04, n = 23) from the regression analysis. The same relation between the period of abandonment and ergosterol content was found as for microscopically determined fungal biomass (Fig. 1B). Ergosterol content was much lower in the agricultural – and 1st year ex- arable fields and the increase was not significant anymore when these sites were excluded (a decrease from R2 = 0.19, n = 29 to R2 = 0.01, n = 23). Fungal-specific PLFA was not significantly related with length of abandonment (Fig. 1C, R2 = 0.00, n = 29). The mean of fungal biomass measured by the three methods in agricultural fields and 1st year ex-arable fields were significantly lower than all other ex-arable sites (Table 2a, no overlap of confidence intervals).
Figure 1: The relation between the period of abandonment and A) microscopically determined fungal biomass B) ergosterol C) fungal PLFA and D) total phosphate. Since relationships in A, B and C were not linear (see text)
No relation was found between length of abandonment and both estimates of bacterial biomass, i.e. microscopically determined bacterial biomass (R2 = 0.00, n = 29) and bacterial PLFAs (R2 = 0.03, n = 29). The fungal: bacterial ratio based on microscopic counting was not significantly related with time since abandonment when the agricultural fields and 1st year ex- arable fields were excluded from the analysis (a reduction from R2 = 0.16, n = 29 to R2 = 0.11, n = 23). In addition, other microbiological parameters, i.e. soil respiration, fungal PLFA, mycorrhizal NLFA and actinomycetal PLFAs were not related with length of abandonment (all R2 < 0.12, n = 29).
Figure 2: RDA triplot depicting the relation of microbiological parameters with abiotic soil properties in a chronosequence of ex-arable land. The first 2 axes explain 62.1% of the microbiological variance and 93.7% of the variance of the relationship between microbiological and abiotic soil properties. Dots are representing field locations with corresponding field ages. Microbiological parameters are presented in italic, soil properties in standard letters. Microbiological parameters are represented by slim lines and codes are: soilresp = soil respiration, fung plfa = fungal PLFA, bact plfa = bacterial PLFAs, actin plfa = actinomycetal PLFAs, ergost = ergosterol, FB micr = F:B ratio based on microscopically determined fungal and bacterial biomass, FB plfa = F:B ratio based on fungal and bacterial PLFAs, nlfa = mycorrhizal NLFA, bact bi = microscopically determined bacterial biomass, fung bi = microscopically determined fungal biomass and glomalin = glomalin. Soil properties are represented by lightly weighed lined and codes are: lab I = labile I C fraction, recalc C = recalcitrant C fraction, org C = total organic C and the rest of the codes are self-explaining.
Relation between fungal biomass and abiotic soil properties
Comparison between the first two axes of PCA (explained variance 93%) and RDA (explained variance 61.2 %, and high species-environment correlation of 0.842) indicated that the most important environmental variables were measured in the study. The RDA triplot shows that fungal and bacterial parameters were highly correlated as arrows are pointing in the same direction (Fig. 2). The fungal and bacterial parameters were positively related with total N and labile I carbon and negatively with the C: N ratio of soil organic matter except for glomalin, which was positively related with C: N ratio. There was no relation between pH and microbiological parameters. Field sites did not cluster with respect to the time of abandonment, e.g. the oldest field (34 years) is plotted nearby an agricultural field.
Simple linear regressions were calculated to indicate single relations between length of abandonment and- soil and microbiological parameters. No abiotic soil properties had a significant relation with length of abandonment, except for P total which was negatively related with the length of abandonment (Fig. 1D, R2 = 0.49). A significant negative relation was found between labile P (calcium chloride-extractable P) and the fungal biomass indicators ergosterol and fungal hyphae length (R2 = 0.27 and 0.29 respectively). In addition, microscopically determined fungal biomass was negatively related with P total (R2 = 0.16, n = 29), but the relation disappeared after excluding agricultural fields and the 1st year ex-arable sites with the lowest fungal biomass values (R2 = 0.03, n = 23).
Comparison of natural sites with abandoned fields
A comparison between the confidence intervals of the mean of microbiological properties of ex-arable fields (agricultural and 1st year ex-arable fields excluded) and the reference sites showed significantly higher values for the fungal biomass indicators ergosterol and fungal PLFA in the latter (Table 2a). Microscopically determined fungal biomass was not significantly different due to large variation in the reference sites. The fungal: bacterial ratio based on PLFAs as well as actinomycetal PLFAs were much higher in reference sites. In addition, soil respiration, an indicator of microbial activity, was higher in reference sites.
Likewise, the soil properties C: N ratio, total and recalcitrant carbon were significantly higher in reference sites than in ex-arable fields (Table 2b). In contrast, P total and mineral nitrogen (NO3--N ) was significantly lower in reference sites than in the ex-arable fields. Confidence intervals of other parameters were overlapping, indicating that there were no significant differences between other abiotic soil and microbiological properties from ex-arable fields and those of the natural reference (heathlands).
Discussion
Fungal biomass, regardless of the method used, was significantly lower in agricultural fields and 1st year ex-arable fields than in the other ex-arable fields. When these youngest fields were excluded from regression analysis, fungal biomass did not significantly increase further with time since abandonment for at least three decades. This could indicate that agricultural soil management is suppressing fungal biomass, and that fungal biomass can be quickly recovered. However, the fungal biomass and the fungal: bacterial biomass ratio in the reference sites are much higher than in the abandoned fields, indicating that fungal biomass is still far from the level of heathland soils. This pattern of fungal biomass development with a short-term initial increase followed by an apparent intermediate stabilization phase has not been revealed by earlier studies on ex-arable fields where only a few, wide time intervals of abandonment were included (Klein et al., 1995; Zeller et al., 2001).
This apparent stabilization phase of fungal biomass implies no further increase, which can be explained in two ways (Fig. 3). The first explanation is that the apparent stabilization could point to a constant maximum carrying capacity for fungal biomass for at least 3 decades. This carrying capacity may be determined by important factors for fungal growth like organic matter quantity and quality, pH and levels of mineral nitrogen (Swift et al., 1979;
DeForest et al., 2004). In addition, the increase in soil fungal biomass during primary succession has been attributed to an increase in the amount of organic matter and C: N ratios as well as to a decrease in soil pH (Ohtonen et al., 1999; Pennanen et al., 2001; Bardgett and Walker 2004; Wardle et al., 2004). In ex-agricultural fields it is predicted that total soil carbon requires 230 years to recover to pre-agricultural levels (Knops and Tilman, 2000). We observed a very slight positive trend of microscopically determined fungal biomass in fields abandoned for more than 1 year (R2 = 0.04), total carbon (R2 = 0.07) and amount of recalcitrant carbon (R2 = 0.10) and a negative trend in soil pH (R2 = 0.11) and in NO3--N (R2
= 0.09) with time since abandonment. Hence, during the period of abandonment studied, the carrying capacity for fungi may have been changing so slowly that a significant increase could not be determined given the additional variation that exists between field sites (Fig. 3, model 1). An increase in fungal biomass would thus occur slowly in relation with changes in abiotic soil properties, an increase of the C: N ratio and quantity of soil organic matter.
The second explanation could be that the apparent stabilization phase is due to threshold values of important environmental properties that have to be exceeded before key fungal groups can enter the ecosystem (Fig 3, model 2). This could for instance be the case for lignocellulose degrading basidiomycetes and ericoid mycorrhizal fungi that make up a significant part of the fungal biomass in woodland and heathland ecosystems (Boddy and
Watkinson, 1995; Read et al., 2004). Lignocellulose degrading- and ericoid mycorrhizal fungi are probably not yet dominating in arable soils, since there is no large amount of recalcitrant organic matter present and no predominance of ericaceous plants in ex-arable soils compared to the references. Before lignocellulose degrading- and ericoid mycorrhizal fungi can proliferate, they may require a minimal amount of woody material or specific host plants, respectively.
It should be realized, however, that the determination of fungal biomass consists of only one time-point in a dynamic process of growth and death of fungi. It is possible that in this stabilization phase the size of the fungal biomass is subject to top-down regulation by predation, e.g. by fungal-feeding nematodes, which do show an increase with time since abandonment (Kardol et al., 2005). Hence, gross increase of fungal biomass in the stabilization period can not be excluded.
The low values of fungal biomass at the start of abandonment may be attributed to a negative effect of artificial fertilizers and pesticides or tillage on fungal growth and maintenance of the hyphal network (Wardle, 1995; Helgason et al., 1998; Chen et al., 2001;
Bittman et al., 2005). We found high values of mineral nitrogen, total and labile phosphate in
Figure 3: Conceptual model for fungal biomass development in secondary succession. Model 1 = very slow gradual increase towards the level of a reference site, model 2 = stabilization phase of at least 3 decades followed by a gradual increase towards the level of a reference site. Solid lines are representing the most likely development according to statistical analyses; dashed lines are representing possible alternative and future development.
agricultural land and 1st year ex-arable land (Fig. 2D and Table 1). Mineral nitrogen can negatively affect growth of fungal decomposers (DeForest et al., 2004). In addition, high phosphate levels have been reported to reduce the colonization of roots by AM-fungi (Azcón et al., 2003). Since hyphae of AM-fungi can make up a significant amount of the fungal biomass in soil (Kabir, 1996; Olsson et al., 1999), high values of phosphate could explain the low values of fungal biomass at the start of abandonment. However, there was no relation between the arbuscular mycorrhizal NLFA marker and total or labile phosphate. In addition, there was a significant negative relation between labile (CaCl2-extractable) phosphate and fungal biomass based on ergosterol. This may point to a negative impact of labile phosphate on saprophytic fungi as AM-fungi do not contain ergosterol (Grandmougin-Ferjani et al., 1999; Olsson et al., 2003). The initial increase of fungal biomass may, therefore, be attributed to a recovery of saprophytic fungi. The potential negative effect of phosphate on fungal biomass is, however, not the only driving factor for fungal biomass development, since fungal biomass did not further increase with a decreasing amount of total phosphate after the first 2 years of abandonment.
The three methods used for fungal biomass measurements were all significantly positively related, indicating that we reliably estimated total and active fungal biomass. With direct counting of fungal hyphal lengths it is assumed that all hyphal lengths are filled with cytoplasm, although it is known that a large proportion of hyphae are empty and non-active (Klein and Paschke, 2004). The positive relation between the more active part of the fungal biomass (measured by fungal PLFA and ergosterol) and total fungal biomass (measured by direct observation) is suggesting that the proportion of living hyphae in the total fungal biomass is similar in the different fields.
The bacterial component of the microbial community appeared to be unaffected by land abandonment, in accordance with Klein et al. (1995). Therefore, an initial increase of the fungal: bacterial ratio during land abandonment is caused by the initial increase of fungal biomass and a constant amount of bacterial biomass. Land transition may therefore lead to a more fungal-based detritus food web compared to agricultural land (Yeates et al., 1997;
Bardgett et al., 2001).
All microbiological parameters were related to total N and the absolute amount of labile I carbon, and negatively correlated with the C: N ratio of the soil. This indicates that fungal and bacterial biomass in this chronosequence are positively influenced by a high quality of organic matter, i.e. a low C: N ratio and a high proportion of easily decomposable substrate (carbohydrates). Indicators for fungal and bacterial biomass were all intercorrelating, suggesting that the preference of bacteria and fungi for organic substrates has not (yet) diverged. Since a low C: N ratio is expected to stimulate sugar and opportunistic
cellulolytic fungi (Swift et al., 1979), these fungal groups may dominate the decomposition processes in this stage of secondary succession. Although microbial composition is not an indicator for microbial functioning, these results could indicate that cellulolytic and ligninolytic wood-degrading basidiomycetes may not yet prevail; competition by r-selected species or a relatively small proportion of recalcitrant organic matter could be limiting their growth. In future work this hypothesis could be tested by using enzyme activities for lignin and cellulose degradation. Based on the results we hypothesize that the functional composition of the fungal community has not yet changed in 34 years of land abandonment, but this is likely to happen in time when abiotic soil properties and resource quality alter.
Acknowledgments
We thank André ten Hoedt, André Westendorp, Aalt Boonen, Tjitske Lubach, Evert Tijmes, Machiel Bosch and Arno Bremer for permission to sample the ex-arable and natural fields and for providing us with information about the history of the sites. We are grateful to An Vos and Meint Veninga from Alterra for their support with microscopic analysis. The study was funded by TRIAS-SKB. This is publication number 3575 of the NIOO-KNAW Netherlands Institute of Ecology.
