at SciVerse ScienceDirect

Soil Biology & Biochemistry 60 (2013) 142e155

Contents lists available

Soil Biology & Biochemistry

journal homepage: www.elsevier .com/locate/soi lbio

The effect of free-living nematodes on nitrogen mineralisation in undisturbed anddisturbed soil cores

David Buchan*, Mesfin Tsegaye Gebremikael, Nele Ameloot, Steven Sleutel, Stefaan De NeveDepartment of Soil Management, Faculty of Bioscience Engineering, University of Ghent, Coupure Links 653, 9000 Ghent, Belgium

a r t i c l e i n f o

Article history:Received 12 April 2012Received in revised form24 January 2013Accepted 26 January 2013Available online 13 February 2013

Keywords:Microbial biomassNematodesGamma irradiationPhospholipid fatty acidsUndisturbed coresNitrogen mineralisation

* Corresponding author. Tel.: þ32 9 264 60 66; faxE-mail addresses: david.buchan@ugent.be, ecolod

mesfintsegaye.gebremikael@ugent.be (M.T. Gebrem(N. Ameloot), steven.sleutel@ugent.be (S. Sleutel), steNeve).

0038-0717/$ e see front matter � 2013 Elsevier Ltd.http://dx.doi.org/10.1016/j.soilbio.2013.01.022

a b s t r a c t

Soil fauna, particularly nematodes, are considered to strongly contribute to nitrogen mineralisationthrough grazing on microflora during decomposition. Demonstration of this effect has mostly relied oncalculation-based soil food web analyses or experiments involving simplified and artificially constructedfood webs. We carried out an incubation experiment in which defaunated soil cores were reinoculatedwith entire soil nematode populations extracted from bulk soil and during which nitrogen mineralisationwas measured. Both undisturbed and disturbed cores were prepared to investigate whether a repre-sentative pore structure influences the effect of entire free-living nematode populations on nitrogenmineralisation. Cores were subjected to a 5 kGy gamma irradiation dose sufficient to eliminate all soilfauna while leaving the microbial biomass largely intact. Half of the irradiated cores were reinoculatedwith nematodes extracted from a corresponding volume of bulk soil and incubated for 82 days. Themicrobial biomass was not strongly affected by gamma irradiation or nematode addition but declinedstrongly in all treatments during incubation. Reinoculation of nematodes was successful in establishingpopulations of a similar size and composition as in the control samples. Net nitrogen mineralisation fromindigenous soil organic matter was observed in all treatments throughout the incubation, but was alwaysmore pronounced in irradiated cores. Total mineral nitrogen concentrations did not differ significantlybetween simply irradiated and irradiated then reinoculated cores. However by the end of the incubationperiod nematode addition resulted in 87% and 23% more NO3

�eN g�1 dry soil in undisturbed anddisturbed cores respectively, while NH4

þeN g�1 dry soil decreased by 50% in both core types. Wefound no convincing evidence for a contribution of free-living nematodes on total nitrogen minerali-sation, but the activity of nitrifying organisms was clearly stimulated by nematode grazing.

� 2013 Elsevier Ltd. All rights reserved.

1. Introduction

Soil fauna have been estimated to collectively contribute toapproximately 30% of nitrogen mineralization (Verhoef andBrussaard, 1990) by excreting excess N in the form of ammoniumand by regulating the size and activity of the microbial community.Microbivorous nematodes contribute strongly to this process byoccupying key trophic positions in the soil food web. Soil free-livingnematodes include bacterivorous and fungivorous nematodes thatdirectly influence nitrogen mineralisation and microbial turnover(Hunt et al., 1987), as well as omnivorous and predatory nematodeswhich may indirectly influence nitrogen mineralisation by

: þ32 9 264 62 47.ave@gmail.com (D. Buchan),ikael), n.ameloot@ugent.befaan.deneve@ugent.be (S. De

All rights reserved.

regulating the population of microbivorous nematodes (Wardleet al., 1995).

This contribution of nematodes to nitrogen mineralisation haspreviously been investigated by means of two distinct approaches:empirically determined in additive microcosm experiments(Anderson and Ineson, 1982; Woods et al., 1982; Ingham et al.,1985; Griffiths, 1986; Bruckner et al., 1995; Ferris et al., 1998;Bardgett and Chan, 1999) or theoretically derived from soil foodwebmodels (e.g. Hunt et al., 1987; Brussaard et al., 1990;Moore andde Ruiter, 1991). Although food web models are based on fieldmeasurements for the estimation of the biomass of the differentfunctional groups, they are sensitive to a number of uncertain inputparameters such as the C:N ratios of bacteria and their substrates(De Ruiter et al., 1993) and the assimilation and production effi-ciencies of various faunal groups (Verhoef and Brussaard, 1990).Lastly food web models do not explicitly take account of indirectcontributions of soil fauna to N mineralisation through the mod-ification of factors limiting microbial growth (Beare et al., 1992).

D. Buchan et al. / Soil Biology & Biochemistry 60 (2013) 142e155 143

The use of microcosms to study the influence of soil fauna onecosystem processes such as nitrogen mineralisation has seennumerous improvements and developments in complexity over thedecades (Huhta, 2007). However, besides a clear bias of mostmicrocosm studies towards small volumes and relatively short in-cubation times (Kampichler et al., 2001), most microcosm experi-ments have lacked realism compared to natural soil in at least oneof the following three ways: (1) a modified soil physical structuredue to soil pre-treatment effects such as sterilisation; (2) a reino-culated microbial community of highly reduced or uncharacteriseddiversity; (3) an artificially assembled soil food web with oftena single species of fauna representing a functional group.

Despite proven links between processes carried out by soil biotaand the physical heterogeneity of the soil habitat (Crawley et al.,2005), the establishment of a representative soil pore structurehas seldom received attention in microcosm studies. However ina series of experiments dealing with the dynamics of nematodepopulations in response to changes in matric potentials, undis-turbed cores were used but nematode populations were notexplicitly manipulated in a controlled manner (Görres et al., 1999;Neher et al., 1999). In a similar type of experiment, Yeates et al.(2002) first defaunated undisturbed cores and then reinoculatedthem with the bacterivorous nematodes Cephalobus, Pristionchusand Rhabditis to study the effects of matric potential on populationdynamics and feeding activity.

To prepare microcosms, soil is usually first mixed, sieved andrepacked; occasionally soil is also dried and re-wetted prior to in-cubation. In some cases a structurally simple and homogenousmedium such as acid-washed sand has been used instead of soil(Ferris et al., 1998; Chen and Ferris, 1999). Common sterilisationprocedures such as autoclaving have been shown to modify the soilpore structure (Clarholm, 1985). Kampichler et al. (1999) showedthat defaunation by freezing allowed intact soil monoliths to beused as experimental units without much physical disruption;however the effect of freezing on microflora was not studied.Gamma irradiation is one technique that can be used to eithersterilise or selectively remove soil fauna from soil without affectingits physical structure (McNamara et al., 2003; Buchan et al., 2012).The influences of such pre-treatments on processes of interest inmicrocosm experiments have, to the best of our knowledge, notbeen systematically studied yet.

A controlled biotic community is mostly established by com-plete sterilisation followed by reinoculation with a single orhandful of micro-organisms easily cultured in laboratory environ-ments at the worst (Coleman et al., 1978; Anderson et al., 1981;Griffiths, 1986; Chen and Ferris, 1999), or a microbial slurry filteredfrom a soil suspension at the best (Clarholm, 1985; Jones et al.,1998; Sulkava and Huhta, 1998; Bardgett and Chan, 1999; Xiaoet al., 2010). Except for one study where microbial communitiesreinoculated into sterile soil were compared to the controls by PLFAanalysis (Griffiths et al., 2008), we found no attempts to comparethe composition or size of reinoculated microbial communitieswith the original ones in undisturbed soil in field conditions.

There has been a long tradition in microcosm experiments totest ecological theories using an artificially constructed soil foodweb with a select amount of species, especially in the case ofnematodes (Woods et al., 1982; Ingham et al., 1985; Griffiths, 1986).Yet related species of nematodes with similar feeding habits havebeen shown to differ in their effect on nitrogen mineralisation(Ferris et al., 1998). In a few controlled experiments, nematodespecies from trophic groups other than bacterivores and fungivoreshave been used to test the effect of omnivory in a simple soil foodweb (Mikola and Setälä, 1999). We found no studies includingpredatory and/or omnivorous dorylaimid nematodes, presumablybecause these are not easily cultivated in laboratory media. Free-

living root-feeding nematodes such as Tylenchidae, which inmany soils make up a significant part of the nematode community,are also not commonly included in experimental studies of nitro-gen mineralisation. Because of these limitations, potentiallyimportant ecological processes such as interspecific competition orpredation have been overlooked. In addition there is a risk ofmaking generalised conclusions concerning the relationship be-tween functional diversity and an ecosystem process, when in factspecies-specific effects are being observed (Mikola, 1998). Howeversome experiments have been conducted using whole nematodecommunities extracted from soil (Sulkava and Huhta, 1998) orderived from soil and intentionally reared so as to be dominated bybacterivores (Xiao et al., 2010).

To address the issues highlighted abovewe chose to carry out anincubation study using both undisturbed and disturbed (i.e. sievedand homogenised) soil cores, in which we attempted to leave themicroflora intact and only selectively remove soil fauna using non-disruptive low-dose gamma irradiation. Most importantly, wechose to reinoculate nematode populations extracted directly fromsoil into our defaunated microcosms. We made no use of amend-ments and instead focused on mineralisation of indigenous soilorganic matter over a longer time period.

The aims of this study were to assess whether entire nematodepopulations in microcosms caused increased nitrogen mineralisa-tion compared to defaunated microcosms, and whether this influ-ence differed between undisturbed and disturbedmicrocosms. Twosecondary aims relating to the methodology we employed were toverify whether the microbial community differed between controland irradiated microcosms, and whether reinoculated nematodepopulations differed from those in control microcosms.

2. Materials and methods

2.1. Field sampling and core preparation

Undisturbed cores and bulk soil were collected in October 2008from an organic agriculture trial field (ILVO, Merelbeke, Belgium)on which clover had been sown earlier in the spring. The soil hada sandy loam texture, and at sampling the pHKCl averaged 5.3,organic C content was 1.03% and C:N ratio was 11. Due to poorestablishment of the clover, it was possible to collect bare, plant-free soil for this experiment. Six undisturbed samples were col-lected to determine bulk density (1230 � 27 Mg m�3) and fieldmoisture content (0.165 � 0.002 g g�1) in the 0e10 cm layer withina ca. 10 � 10 m area at least 5 m from the edges of the field. Fromthis same area 54 undisturbed cores were collected from the upper7.5 cm by inserting bevelled PVC tubes (height 7.5 cm, diameter7.5 cm). These were capped on both ends and stored with minimalphysical disturbance. A large amount of bulk soil was also collectedfrom the same area and depth layer. Bulk soil was first mixed inbuckets and coarsely sieved (10 mm mesh) to remove stones androot fragments and then further gently homogenised. Bulk soil wasused to refill another 54 disturbed cores with 280.0 g fresh soil,which was gently compacted to the same bulk density as measuredin the field. The remaining bulk soil was stored in black plastic bagsin a basement at around 17 �C until nematode extraction (no morethan two days later).

2.2. Gamma irradiation and nematode reinoculation

One third (n ¼ 18) of both type of cores were set aside ascontrol cores and kept overnight in a basement at ca. 17 �C alongwith the remaining bulk soil. All the other cores were subjectedto gamma irradiation at a dose of 5 kGy (Sterigenics industrialfacility, Fleurus, Belgium), following a previously determined

D. Buchan et al. / Soil Biology & Biochemistry 60 (2013) 142e155144

dosage proven to eliminate nematodes but not significantlyaffect microbial biomass (Buchan et al., 2012) and storedtogether with control cores. Cores were always kept coveredduring storage to avoid moisture loss (verified by mass balance).Entire nematode populations were extracted from 150.0 g ali-quots of bulk soil using the zonal centrifuge method (Hendrickx,1995). Preliminary experiments with the same soil showed thatthe zonal centrifuge extraction method gave the same thenematode abundance and composition as a Cobb’s sieving anddecanting extraction followed by migration through cotton woolfilters for 48 h; more than 90% of the nematodes extracted withthe zonal centrifuge were recovered after migration throughcotton wool filters (unpublished data). Nematodes were col-lected in 100 ml water and then washed over a 5 mm sieve toremove MgSO4 and kaolinite from the extraction solution, aswell as dissolved or particulate organic matter and bacteria thatmay have been extracted along with the nematodes. Half of theirradiated cores (n ¼ 18 for each core type) were reinoculatedwith nematodes extracted from a corresponding amount of bulksoil, assuming an extraction efficiency of approximately 70% (i.e.nematodes extracted from three 100.0 g bulk soil samples wereused to inoculate each core). Although it could not be verifiedwhether extraction efficiency differed between specific taxa, thecomposition of the nematode extracts used for reinoculation didnot differ from those of the populations extracted from controlsoil.

2.3. Experimental setup and sampling

The cores preparation steps generated six treatments: CU (un-disturbed control), GUþ (irradiated undisturbed and reinoculated),GU� (irradiated undisturbed), CD (disturbed control), GDþ (irra-diated disturbed and reinoculated), and GD� (irradiated dis-turbed). Irradiated and reinoculated cores are henceforth referredto as ‘reinoculated’, simply irradiated cores as ‘defaunated’. Coreswere brought to and maintained at ca. 50% water-filled pore spaceand incubated at 18 �C for 82 days. The amount of water to be addedwas derived from the bulk density and initial moisture content ofthe soil; weekly water additions were calculated using simple massbalance of each individual core. Three replicates from each treat-ment were destructively sampled after 5, 12, 26, 38, 53 and 82 daysof incubation.

At each sampling event 3 randomly selected cores per treatmentwere removed, gently homogenised and each sampled for thedetermination of chemical and biochemical parameters. Gravi-metric moisture content was determined and all parameters areexpressed per gram of dry soil.

2.4. Nematode analyses

Nematodes were extracted as above from 100.0 g soil from eachcore, after thorough mixing. The zonal centrifuge extractionmethod as described by Hendrickx (1995) was used, which usescentrifugation to separate free-living nematodes from mineral andmost organic soil constituents based on their specific density.Extracted nematodes were washed as above and concentrated in5e10 ml distilled water for storage at 4 �C until further processing.Within days of extraction nematodes were counted under a binoc-ular microscope; for each sample the percentage of moving, activenematodes was estimated after counting. A previous experiment(Buchan et al., 2012) indicated that recently irradiated nematodeswere extracted and counted along with viable living nematodes;hence we paid attention to the visual state of nematodes duringcounting and taxonomic determination. After counting, nematodeswere fixed with 4% hot formaldehyde (70 �C) for further

conservation. During counting we estimated the percentages ofmobile, active nematodes and seemingly dead, immobilenematodes.

Nematode identification was carried out only on control andreinoculated samples from the beginning (5 days) and from the end(82 days) of the incubation period. In addition nematodes fromthree replicates of bulk soil sampled on the field were identified toestablish initial conditions. Mass slides were prepared and nema-todes identified to family level or genus where possible accordingto Bongers (1988) and allocated to trophic groups according toYeates et al. (1993). The percentage of degraded or otherwise un-identifiable nematodes was determined to differentiate betweenirradiated and viable nematodes and calculate the abundances ofall identified nematodes. Given the yet unresolved issue of theexact feeding habits of nematodes in the family Tylenchidae(Christensen et al., 2007), we preliminarily considered only Fil-enchus to be fungivorous and other genera in the family to be root-feeders. Dauer larvae were counted separately but not differ-entiated by family. To further assess whether the overall functionaldiversity of nematode communities was affected by experimentaltreatments, we calculated several ecological indices as summarisedby Bongers and Ferris (1999) and Ferris et al. (2001): the MaturityIndex (MI), the same index excluding enrichment opportunists (MI-2e5), the plant-parasitic index (PPI), the enrichment index (EI), thestructure index (SI), the basal index (BI) and the channel index (CI).

2.5. Chemical analyses

Mineral N (as NO3� and NH4

þ) was determined from 30.0 g freshsoil extracted with 60 ml 1 M KCl and measured colourimetricallyby a continuous flow auto-analyser (Chem-lab 4, Skalar Analytical,Breda, The Netherlands). Microbial biomass carbon (Cmic) wasdetermined using the fumigationeextraction technique (Vanceet al., 1987). Fumigated soil and non-fumigated controls (30.0 gfresh soil) were extracted with 60 ml 0.5 M K2SO4 and storedat �18 �C until analysis. Organic carbon contents of the extractswere determined with a TOC analyser (TOC-VCPN, Shimadzu Cor-poration, Kyoto, Japan). For conversion from organic C in extracts toCmic in soil a kEC value of 0.45 was assumed (Joergensen, 1996). Thefumigation procedure at days 12 and 53 failed due to an incompletevacuum therefore these data were omitted from further analyses.

2.6. PLFA extractions

Phospholipid fatty acids (PLFAs) were extracted usinga modified Bligh and Dyer (1959) technique. The procedure fol-lowed is described in more detail by Moeskops et al. (2010).Briefly, lipids were extracted from 4.00 g freeze-dried soil witha one-phase mixture of phosphate buffer (pH 7.0), chloroformand methanol in volume ratios of 0.8:1:2 respectively. Lipidswere then fractionated into neutral, glyco- and phospho-lipidsusing SiOH SPE cartridges (Chromabond, MachereyeNagelGmbH, Düren, Germany). Neutral and glycolipids were dis-carded and the phospholipid fraction was subjected to mildalkaline methanolysis to give fatty acid methyl esters (FAMEs).These were dried down under N2 and re-dissolved in 300 mlhexane containing 8 mg L�1 19:0 (nonadecanoic acid methylester; SigmaeAldrich Inc., St. Louis, USA) as an internal standard.Samples were analysed by capillary gas chromatographyemassspectrometry (Thermo Focus GC coupled to Thermo DSQ MS;Thermo Fischer Scientific Inc., Waltham, USA), using splitlessinjection, helium as a carrier gas and a Restek Rt-2560 capillarycolumn (100 m length � 0.25 mm internal diameter, 0.2 mm filmthickness; Restek, Bellefonte, USA). FAMEs were identified bychromatographic retention time using the total ion count method

Fig. 1. Microbial biomass carbon in undisturbed (a) and disturbed (b) soil.

D. Buchan et al. / Soil Biology & Biochemistry 60 (2013) 142e155 145

and by comparison with a standard quantitative FAME mixture(Restek Food Industry FAME mix 35077) and a qualitative FAMEmixture (Supelco Bacterial Acid Methyl Ester mix 28039U, towhich individual standards of a16:0, a17:0, 16:1u5, 10Me16:0 and10Me18:0 were added). FAMEs were quantified using serial di-lutions of the quantitative FAME mix; for FAMEs for which onlyqualitative standards were available, the calibration curve of thenearest structural analogue was used, so that the concentrationof all FAMEs could be expressed in nmol g�1 dry soil. Only FAMEswhich could be accurately identified based on the retention timeof known standards were measured and included in furtheranalysis, after which no major peaks remained unidentified inany of the sample chromatograms. Of the 49 standard FAMEswhich could be detected with certainty, 25 consistently occurredin significant amounts, always representing more than 90% of thetotal measured FAME concentration. Only these FAMEs wereretained for further statistical analysis and the sum of these ishenceforth referred to as the ‘total’ amount of PLFAs (expressedin nmol g�1). The nomenclature of fatty acids was adapted fromZelles (1999); potential taxon-specific biomarker FAMEs weretaken from Moeskops et al. (2010) as follows: the sum of i15:0,a15:0, i16:0, a16:0, i17:0 and a17:0 for Gram-positive bacteria;cy17:0 for Gram-negative bacteria (cy19:0 co-eluted with18:2u6,9c and was hence ignored); the sum of 10Me16:0 and10Me18:0 for the actinomycetes, 18:1u9c for saprotrophic fungi(as an alternative to 18:2u6,9c e cf. Joergensen and Wichern,2008) and 16:1u5c for arbuscular mycorrhizal fungi (AMF). Theratio of the cyclopropyl fatty acid cy17:0 to its monounsaturatedprecursor 16:1u7c was used as an indicator of nutrient stress ofGram-negative bacteria, with higher values indicating reducedactivity (Peterson et al., 1997). The bacterial:fungal ratio wascalculated as the sum of the Gram-positive and Gram-negativebiomarkers in addition to fatty acids 15:0 and 17:0 divided bythe fungal fatty acid 18:1u9c.

2.7. Statistical analyses

The following 3-way factorial design was used in the analysis ofthe mineral N, nematode abundance and microbial biomass C:a factor core with 2 levels (undisturbed and disturbed), a factorgamma with 3 levels (control, defaunated and reinoculated) anda factor time with 6 levels (0, 5, 12, 26, 37, 53 and 82 days of in-cubation). Because the response to irradiation and reinoculationdiffered markedly between undisturbed and disturbed cores andmostly resulted in interactions effects involving the factor core, thisfactor was fixed and split-plot 2-way ANOVA of the remainingfactors was undertaken instead. For PLFA profiles and the nematodepopulation composition a similar split-plot design was used exceptthat the factor time had either 3 (disturbed cores: 0, 5 and 82 days)or 2 (undisturbed cores: 5 and 82 days) levels. For nematodeabundance the factor gamma obviously had only 2 levels (controland reinoculated). For any significant factor withmore than 2 levelsTukey’s HSD post hoc test was used to compare groups. Wherenecessary, data were log-transformed to meet assumptions ofnormality and homoscedasticity prior to statistical analyses. Toinvestigate possible changes in microbial community structure, theconcentrations of all consistently occurring FAMEs (expressed inthe nmol g�1) from days 5 and 82 were subjected to principalcomponent analysis (PCA) based on the correlation matrix. Corre-lations were investigated using Spearman’s correlation coefficient,and where significant, the data set was additionally split accordingto the factors time, gamma and/or core to verify if correlations heldtrue under all conditions. All statistical tests were carried out ata significance level of 0.05, using the statistical software pro-gramme SPSS 17.0.

3. Results

3.1. Microbial biomass carbon

Microbial biomass carbon (Cmic) displayed similar dynamics inall treatments in both undisturbed and disturbed soils: a peak after5 days of incubation then a sharp decline until reaching a morestable baseline from 38 days onwards (Fig. 1, p< 0.001 for the factortime for both core types). In disturbed cores, Cmic of reinoculatedand defaunated treatments didn’t differ from each other but weresignificantly lower than that of the control (p ¼ 0.009). In undis-turbed cores Cmic was lower in the defaunated than in the controltreatment, while the reinoculated was intermediate between thetwo and significantly different (p¼ 0.032 for factor gamma). In bothundisturbed and disturbed cores the gamma treatment had a sig-nificant effect on Cmic (p ¼ 0.013 for both, model with interaction)

D. Buchan et al. / Soil Biology & Biochemistry 60 (2013) 142e155146

being on average higher in the control than in irradiated treat-ments. In the undisturbed control Cmic peaked much more stronglythan in irradiated cores and in its disturbed counterpart, whosepeak was the same as the defaunated treatment.

3.2. PLFA profiles

The PLFA composition for all treatments was measured after 5and 82 days of incubation (Table 1). Almost all significant differ-ences in biomarkers were found among disturbed cores only. TotalPLFA concentration was significantly affected by incubation time indisturbed soils, but not in undisturbed soil cores. The biomarker forfungi was significantly affected by the factor time, being lower in alltreatments after 82 days of incubation than after 5 days. As a resultthe bacterial:fungal ratio increased significantly with time in alltreatments. The marker for actinomycetes was lower after 82 daysthan after 5 days in disturbed cores, as was that for Gram-positivebacteria. In undisturbed cores only the bacterial:fungal ratio wassignificantly affected by the factor time even though there were nosignificant differences in the concentrations of bacterial and fungalbiomarkers. The ratio of fatty acids cy17:0 to 16:1u7c, an indicatorof stress, significantly increased with incubation time in disturbedcores. In undisturbed cores there was a weak but significantinteraction between the factors time and gamma, due to a decreasewith time in the defaunated treatment only (Table 1); in contrast inall other treatments the ratio cy17:0 to 16:1u7c increased from day5 to day 82.

The principal component analysis of all FAMEs segregatedsamples mainly according to sampling time and more weakly ac-cording to core type (Fig. 2), although clustering was by no meansclear-cut and individual data points were spread out and over-lapped between groups. The first principal component accountedfor 41.1% of the total variance and the second for 17.1%, or combined58.4% of the total variance. The projection of the loadings of indi-vidual FAMEs as vectors on the ordination plot showed that

Table 1Concentration of total PLFA and selected biomarkers in mol g�1 dry soil, except for the B:F(by core type) two-way ANOVA analysis for the factor gamma and time. Where the inteinteraction effect in the model. Significant effects (p < 0.05) are indicated in bold.

Factors Total PLFA Gram-positive Gram-negative A

DisturbedControl0 31.24 (1.25) 6.77 (0.24) 0.97 (0.04) 25 32.60 (3.83) 7.78 (0.96) 1.09 (0.13) 282 27.61 (2.96) 6.66 (0.67) 0.79 (0.10) 2Reinoculated5 34.15 (0.56) 7.49 (0.12) 0.96 (0.01) 282 26.64 (2.72) 5.59 (0.46) 0.94 (0.20) 1Defaunated5 33.05 (0.68) 7.82 (0.19) 0.89 (0.04) 282 28.58 (1.49) 7.09 (0.52) 0.88 (0.16) 2P (gamma) 0.968 0.348 0.764 0P (time) 0.021 0.011 0.460 0P (gamma � time) ns ns ns nUndisturbedControl5 32.97 (2.33) 7.47 (0.41) 1.00 (0.09) 282 34.10 (1.62) 8.16 (0.36) 1.05 (0.07) 2Reinoculated5 33.29 (2.35) 7.61 (0.77) 1.17 (0.15) 282 31.81 (0.51) 7.21 (0.23) 1.04 (0.01) 2Defaunated5 29.92 (1.29) 7.02 (0.47) 0.89 (0.01) 282 33.40 (3.25) 7.09 (0.52) 1.12 (0.38) 2P (gamma) 0.665 0.314 0.822 0P (time) 0.542 0.762 0.715 0P (gamma � time) ns ns ns n

a number of bacterial and fungal FAMEs contributed to segregatesamples by sampling date in approximately equal measures (Fig. 3).It stands out that no single FAME dominates the share of any otherin terms of loadings, meaning the observed clustering in function ofsampling time is a reflection of the overall decrease in the con-centration of a range of microbial FAMEs, confirming the trend seenin the individual biomarker data.

3.3. Nematode population dynamics

Nematode abundance was strongly affected by incubation timeand gamma treatment (Fig. 4), but due to a strong interaction be-tween these factors in undisturbed cores (Table 2), results arepresented separately for each gamma treatment. Nematodes wererecovered from non-inoculated cores throughout the duration ofthe experiment, although this amount steadily declined throughoutthe experiment to ca. 5 individuals g�1 dry soil for both core types(Fig. 4). Subsequent microscopic observation confirmed thesenematodes were dead. Unexpectedly high amounts of nematodeswere found in disturbed defaunated samples after 11 days,although these were also clearly immobile and dead.

In undisturbed cores nematode abundance in the controlsreached 30e35 individuals g�1 dry soil after 26 days of incubationand stabilised at around 20 individual g�1 dry soil by day 53(Fig. 4a). In disturbed soil population levels for the controls did notdeviate much from this same mean value and there was no distinctpeak (Fig. 4b).

Inoculation of nematodes was successful in establishing pop-ulations of the same magnitude as in the control samples: after aninitial increase between 5 and 12 days after inoculation, abundancein both types of core decreased to around 75% of control levels.

The composition of the nematode community was determinedafter 5 and 82 days of incubation in the control and reinoculatedsamples, and also in disturbed cores at sampling (0 days). Theproportion of unidentifiable nematodes reflected the recent death

(bacterial:fungal) ratio and stress (cy17:0/16:1u7c) ratio and results of the split-plotraction of these factors was non-significant, the analysis was repeated without the

ctinomyc. Fungi AMF B:F ratio Stress ratio

.44 (0.08) 1.77 (0.08) 1.74 (0.09) 4.76 (0.06) 0.342 (0.005)

.52 (0.38) 1.93 (0.41) 1.64 (0.29) 5.47 (0.69) 0.350 (0.023)

.17 (0.20) 1.52 (0.19) 1.46 (0.25) 5.40 (0.33) 0.366 (0.041)

.67 (0.08) 2.14 (0.07) 1.58 (0.07) 4.45 (0.14) 0.240 (0.010)

.80 (0.17) 1.36 (0.07) 1.04 (0.12) 5.32 (0.22) 0.379 (0.009)

.52 (0.04) 1.80 (0.30) 1.58 (0.06) 4.44 (0.06) 0.282 (0.016)

.13 (0.05) 1.45 (0.03) 1.56 (0.24) 5.63 (0.29) 0.347 (0.071)

.840 0.836 0.324 0.265 0.221

.013 0.003 0.184 0.052 0.027s ns ns ns ns

.75 (0.24) 2.09 (0.11) 1.70 (0.34) 4.53 (0.04) 0.291 (0.021)

.83 (0.32) 1.84 (0.23) 2.09 (0.33) 5.54 (0.52) 0.337 (0.012)

.68 (0.37) 2.14 (0.37) 1.87 (0.29) 4.79 (0.34) 0.326 (0.007)

.38 (0.20) 1.58 (0.16) 1.69 (0.22) 5.77 (0.43) 0.348 (0.019)

.38 (0.21) 1.80 (0.30) 1.35 (0.11) 5.16 (0.60) 0.326 (0.015)

.27 (0.17) 1.53 (0.12) 1.43 (0.20) 5.93 (0.56) 0.281 (0.022)

.201 0.408 0.150 0.503 0.166

.592 0.068 0.628 0.019 0.577s ns ns ns 0.048

Fig. 2. Biplot of the first two principal components of the PCA analysis. The firstprincipal component explained 41.1% of the total variance, the second principal com-ponent a further 17.4%. Data points were averaged by core type and incubation time.

D. Buchan et al. / Soil Biology & Biochemistry 60 (2013) 142e155 147

of nematode due to the experimental treatments used: 20e30% ofall observed nematodes were unidentifiable in reinoculated coresafter 5 days, while after 82 days this proportion had decreased tobelow 10%, as found in the control cores throughout theexperiment (data not shown). The percentage of unidentifiablenematodes at the initial sampling was lower (4%) than in theincubated controls. In non-inoculated samples in which nema-todes were observed, the proportion of unidentifiable nematodesalways exceeded 90%.

Fig. 3. Vector plot of PCA loadings for individual FAMEs, using the same axes as Fig. 2but on a different scale.

Fig. 4. Nematode abundance in undisturbed (a) and disturbed soil (b).

The amount of viable nematodes was calculated from thesepercentages and the nematode counts. Given the abundance ofviable nematodes was lower in reinoculated cores (Table 2), nem-atode composition data are given and discussed as abundance data(individuals g�1 dry soil) instead of percentages (e.g. Sánchez-Moreno et al., 2010; Xiao et al., 2010).

Root-feeding nematodes of the families Tylenchidae (excludingFilenchus), Tylodoridae (Cephalenchus), Dolichodoridae and Praty-lenchidae together represented about a third of the nematodecommunity in all treatments. In control cores, root-feeders main-tained amore or less constant presence throughout the experimentdespite the absence of any visible plant root fragments. In reino-culated cores on the other hand the abundance of root-feedersdecreased by about a third between the two sampling dates(Table 2). When Tylenchidae were excluded from this trophic

Table

2Abu

ndan

ceof

mainnem

atod

etrop

hicgrou

ps(individualsg�

1dry

soil)

andresu

ltsof

split-plot(byco

retype)

two-way

ANOVAan

alysisforthefactorsga

mmaan

dtime.W

heretheinteractionof

thesefactorswas

non

-significant,

thean

alysis

was

repea

tedwithou

ttheinteractioneffect

inthemod

el.S

ignificanteffects(p

<0.05

)areindicated

inbo

ld.

Factors

Total(viable)

Roo

tfeed

ers

Fungivo

res

Bacterivo

res

Omnivores

Total

�Ty

lench

aTo

tal

þTy

lench

aFilenc

hus

Total

Rhab

dit.

Panag

rol.

Diploga

st.

Dau

erCep

halo.

Total

Aporcelai.

Distu

rbed

Control

019

.23(1.64)

6.03

(0.85)

4.50

(0.69)

1.78

(0.13)

3.35

(0.05)

1.55

(0.18)

7.18

(0.97)

1.39

(0.31)

0.91

(0.19)

0.13

(0.10)

3.93

(0.31)

4.56

(0.62)

0.25

(0.01)

0.03

(0.03)

518

.54(0.42)

7.05

(0.35)

4.52

(0.56)

1.80

(0.41)

4.32

(0.96)

1.28

(0.15)

5.18

(0.90)

0.55

(0.19)

0.35

(0.19)

0.26

(0.06)

4.12

(0.52)

3.78

(0.84)

0.36

(0.16)

0.15

(0.09)

8219

.11(2.98)

6.45

(0.75)

3.84

(0.19)

2.24

(0.44)

4.85

(1.01)

1.07

(0.06)

5.77

(1.33)

0.23

(0.08)

0.56

(0.16)

0.12

(0.07)

4.29

(1.05)

4.71

(1.33)

0.34

(0.17)

0.14

(0.02)

Reino

culated

522

.74(2.47)

5.71

(0.40)

3.73

(0.10)

1.44

(0.29)

3.24

(0.53)

1.13

(0.25)

8.53

(1.75)

0.16

(0.03)

0.04

(0.04)

2.62

(1.29)

6.42

(1.54)

5.49

(0.64)

0.65

(0.12)

0.36

(0.12)

8213

.57(0.40)

3.39

(0.39)

2.31

(0.38)

0.79

(0.18)

1.88

(0.28)

0.50

(0.04)

4.91

(0.37)

0.28

(0.22)

0.22

(0.10)

1.24

(0.04)

3.91

(0.19)

3.10

(0.57)

0.51

(0.15)

0.19

(0.08)

P(gam

ma)

0.73

20.00

40.02

20.02

30.02

00.04

90.34

90.40

90.04

40.01

20.32

70.95

50.04

40.11

7P(tim

e)0.12

40.07

70.08

50.82

10.30

70.02

90.31

10.00

50.03

80.45

50.46

70.67

10.43

00.28

0P(gam

ma�

time)

0.02

8ns

ns

ns

ns

ns

ns

ns

ns

ns

ns

Ns

ns

ns

Undistu

rbed

Control

522

.79(1.19)

8.03

(0.92)

6.36

(1.06)

1.46

(0.34)

3.31

(0.69)

1.10

(0.39)

7.79

(0.67)

1.69

(1.05)

0.61

(0.24)

0.10

(0.10)

4.75

(0.29)

5.35

(0.90)

0.54

(0.15)

0.15

(0.11)

8217

.75(2.16)

6.46

(0.90)

3.57

(0.58)

1.89

(0.39)

4.79

(0.82)

1.21

(0.38)

6.26

(0.69)

0.61

(0.01)

0.47

(0.10)

0.05

(0.05)

2.73

(0.28)

4.65

(0.58)

0.35

(0.23)

0.16

(0.09)

Reino

culated

511

.35(1.24)

3.39

(0.44)

2.64

(0.49)

0.66

(0.15)

1.41

(0.28)

0.48

(0.14)

3.34

(0.47)

0.23

(0.06)

0.40

(0.34)

0.77

(0.73)

3.60

(0.99)

1.83

(0.17)

0.37

(0.08)

0.24

(0.05)

8211

.83(0.67)

4.46

(0.42)

2.90

(0.38)

1.31

(0.32)

2.67

(0.30)

1.01

(0.20)

4.22

(0.18)

0.23

(0.05)

0.37

(0.05)

0.09

(0,05)

1.55

(0.20)

3.48

(0.24)

0.45

(0.15)

0.20

(0.04)

P(gam

ma)

<0.000

0.00

20.02

50.03

00.00

50.18

80.00

10.11

70.46

10.35

20.04

90.34

30.52

40.39

6P(tim

e)0.19

40.66

50.15

50.16

80.03

20.29

80.62

80.33

60.69

50.34

80.00

30.01

40.39

10.82

0P(gam

ma�

time)

ns

ns

ns

ns

ns

ns

ns

ns

ns

ns

ns

ns

ns

ns

aTh

etotalab

undan

ceof

root-fee

derswas

also

calculatedwithallTy

lench

idae

excluded

(�Ty

lench

.)an

dthetotalab

undan

ceof

fungivo

resalso

calculatedwithallTy

lench

idae

included

(þTy

lench

.).

D. Buchan et al. / Soil Biology & Biochemistry 60 (2013) 142e155148

group, the trend of the remaining root feeders was similar but lesspronounced.

Fungivorous nematodes consisted mainly of Filenchus anderratically occurring numbers of Aphalenchus and Aphalenchoides.Reinoculation resulted in significantly lower abundances of fungi-vores than in the controls for both core types (Table 2). This wasmainly due to a lower occurrence of Filenchus in reinoculated cores,whose abundance was significantly and negatively affected byreinoculation and incubation time in disturbed cores only. Theabundance of Filenchus had doubled by the end of incubation inundisturbed reinoculated cores. When the remaining Tylenchidaewere also included in the fungivorous trophic group, reinoculationhad a similarly significant effect in disturbed cores and a stronglysignificant effect in undisturbed cores (Table 2). For undisturbedcores, this inclusion also resulted in a positive effect for the dura-tion of incubation, the population of fungivores tending to begreater after 82 days than after 5 days for both control and reino-culated cores.

Only the abundance of Aphalenchoides (data not shown) waspositively correlated with the fungal PLFA biomarker in disturbedcores (R ¼ 0.66, p ¼ 0.008).

All together bacterivorous nematodes constituted an importantproportion (28e37%) of the nematode population irrespective oftreatments (Table 2). In disturbed cores, the abundance of bacter-ivores was highest in reinoculated cores after 5 days and lowest inreinoculated cores after 82 days, while the abundance in controlcores remained more or less constant (no significant effects). Inundisturbed cores the abundance of bacterivores was significantlylower in reinoculated cores compared to control cores. The generalopportunist nematodes of the family Cephalobidae made upthe bulk of the active bacterivores at all sampling times (17e25%of total nematode abundance), although the significantlydecreased in abundance in undisturbed cores. The differentfamilies of enrichment opportunists encountered (Rhabditidae,Panagrolaimidae and Diplogasteridae) all responded differently toreinoculation and incubation conditions depending on core type.Rhabditidae strongly decreased in abundance in both control andreinoculated disturbed cores compared to the initial conditions.In undisturbed cores they remained more strongly present in thecontrols only, although this was not significant. Panagrolaimidaein disturbed cores were also decreased in abundance with timeand due to reinoculation, while in undisturbed cores theymaintained a more stable but low presence. Diplogasteridae onthe other hand increased in abundance due to reinoculation indisturbed cores by a factor of 10 compared to control cores,although their abundance had decreased by about 50% after 82days. Reinoculation in undisturbed cores did not significantlyaffect Diplogasteridae, although their abundance after 5 daysincreased several fold in some cores. The high variability reportedfor several of the nematode taxa (see magnitude of standarderrors compared to that of the mean in Table 2) was due todiverging population developments in individual cores. Amongboth undisturbed and disturbed reinoculated cores, only someshowed a substantial increase in Diplogasteridae by the end of theexperiment, while in other cores they were hardly detectable (in-dividual core data not shown). Overall, Rhabditidaewere negativelycorrelatedwith Diplogasteridae after 5 days of incubation (R¼ 0.65,p ¼ 0.023) but not at the end of the experiment. The abundances ofbacterivorous nematodes were only correlated with Cmic after 5days of incubation when microbial biomass was at its highest. Inundisturbed cores, the total abundance of c-p 1 bacterivores wascorrelated with Cmic (R ¼ 0.60, p ¼ 0.039). In disturbed cores, theabundance of Rhabditidae was correlated with Cmic in control coresonly (R ¼ 0.75, p ¼ 0.005). Overall, Panagrolaimidae were weaklypositively correlated with Cmic (R ¼ 0.62, p ¼ 0.032), while

D. Buchan et al. / Soil Biology & Biochemistry 60 (2013) 142e155 149

Diplogasteridae were negatively correlated with Cmic (R ¼ �0.74,p ¼ 0.006) due to their low presence at the beginning of theexperiment.

Dauer larvae were aggregated irrespective of the family towhich they belonged, and were present in proportions of 13e30%compared to 6e13% for the sum of the active enrichment oppor-tunists. In disturbed cores their abundance did not significantlyvary between the different treatments, but was greatest in reino-culated cores after 5 days. In undisturbed cores both reinoculationand time significantly affected the abundance of dauer larvae: theywere less abundant in reinoculated cores and had decreased byabout half in both treatments (Table 2). Dauer larvae were pos-itively correlated to the sum of all c-p 1 bacterivores in undisturbed(R ¼ 0.75, p ¼ 0.005) and reinoculated cores (R ¼ 0.81, p ¼ 0.001),and alsowith Cmic (R¼ 0.76, p¼ 0.005) in undisturbed cores only. Inreinoculated cores dauer larvae were positively correlated withDiplogasteridae (R ¼ 0.83, p ¼ 0.001) and negatively correlatedwith Panagrolaimidae (R ¼ �0.74, p ¼ 0.006).

Omnivorous dorylaimid nematodes primarily consisted of in-dividuals from the families Qudsianematidae and Aporcelaimidaeand occurred in proportions of 1e4%. Reinoculation only hada positively significant effect on omnivores in disturbed cores, be-ing twice as abundant as at the initial sampling. Omnivore abun-dance in undisturbed cores was comparable between the controland reinoculated treatments. Numbers for individual families weretoo low for statistical analysis, but seem to indicate that Aporce-laimidae in particular responded positively to incubation condi-tions and reinoculation (Table 2).

The percentage compositions of the nematode communitieswere used to calculate maturity and other community indices. Thematurity index (MI) increased with incubation time in the dis-turbed cores only due to an increase in the proportion of omnivo-rous nematodes with c-p values of 4 and 5 (Table 3). TheMI 2e5 didnot display any significant trend. The plant-parasitic index (PPI)decreased with time in undisturbed cores only, reflecting the near-disappearance of Trichodorus (c-p 5) after 82 days of incubation(not shown in Table 2 due to low abundance). The enrichment in-dex (EI) in disturbed cores was significantly higher in the reino-culated compared to the control treatment; in undisturbed coresthere was no consistent response. The generally increased EI in

Table 3Nematode maturity and community indices and results of split-plot (by core type) two-wfactors was non-significant, the analysis was repeated without the interaction effect in t

Factors Maturity indices

MI MI 2e5 PPI

DisturbedControl0 1.79 (0.06) 2.17 (0.02) 2.61 (0.03)5 1.98 (0.06) 2.25 (0.05) 2.51 (0.06)82 2.02 (0.05) 2.24 (0.14) 2.57 (0.08)Reinoculated5 1.91 (0.10) 2.28 (0.03) 2.64 (0.02)82 1.94 (0.30) 2.42 (0.15) 2.57 (0.06)P (gamma) 0.236 0.287 0.299P (time) 0.042 0.627 0.611P (gamma � time) ns ns nsUndisturbedControl5 1.88 (0.16) 2.22 (0.02) 2.71 (0.04)82 2.05 (0.08) 2.26 (0.08) 2.47 (0.05)Reinoculated5 1.95 (0.19) 2.45 (0.06) 2.83 (0.07)82 2.12 (0.09) 2.31 (0.09) 2.56 (0.08)P (gamma) 0.604 0.082 0.108P (time) 0.233 0.483 0.002P (gamma � time) ns ns ns

reinoculated cores was brought about by a greater proportion of c-p1 bacterivores (17.77 � 3.21% in reinoculated cores versus9.99 � 1.55% in the control cores, p ¼ 0.007 e data not shown). Thestructure index (SI) tended to be higher in reinoculated cores (nosignificant effects) due to a greater proportion of omnivorousnematodes. The basal index (BI) and channel index (CI) were bothsignificantly lower following reinoculation in disturbed cores, dueto a lower proportion of fungivores in reinoculated cores comparedto controls (6.27 � 1.07% and 10.27 � 0.91% respectively excludingTylenchidae, p ¼ 0.023 e data not shown).

3.4. Mineral nitrogen

Nitrate and ammonium concentrations were not grouped foranalysis due to the effect of irradiation on ammonium release. Fordisturbed cores the factors gamma and time had a highly significanteffect on nitrate concentrations (p < 0.001) and their interactionwas also significant (p ¼ 0.023). In undisturbed cores time andgamma also had a significant effect (p < 0.001 and p ¼ 0.017respectively) but their interaction was not significant (Fig. 5).Fitting a simple linear model over the 72-day period betweenthe first and last samplings, nitrate concentrations in disturbedcores increased at average rates of 0.247, 0.221 and0.286 mg N g�1 dry soil day�1 in control, defaunated and reinocu-lated cores respectively. In undisturbed cores these rates were0.167, 0.156 and 0.325 mg N g�1 dry soil day�1 respectively. Thesegreater calculated mineralisation rates in reinoculated cores weredue to significantly higher nitrate concentrations at the last sam-pling date (82 days). In undisturbed cores irradiated soil had mar-ginally higher concentrations than the control from the start, but itwas not until the last sampling that reinoculated cores containedalmost twice as much nitrate than control and non-inoculatedcores (Fig. 5a). In disturbed cores considerably more nitrate wasformed in the control and non-inoculated samples, meaning thereinoculated cores contained only ca. 20% more nitrate by the endof the experiment (Fig. 5b).

The effect of incubation time and gamma on ammonium con-centrations could not be tested statistically using ANOVA due tohighly unequal variances between control and irradiated cores,even after log transformation (Fig. 6). Irradiation caused an

ay ANOVA analysis for the factors gamma and time. Where the interaction of thesehe model. Significant effects (p < 0.05) are indicated in bold.

Community indices

EI SI BI CI

78.6 (1.4) 21.2 (3.0) 20.2 (1.3) 15.7 (3.2)77.5 (1.1) 27.8 (5.2) 20.7 (1.3) 27.3 (5.8)70.1 (5.9) 22.5 (9.6) 28.0 (6.1) 37.9 (2.2)

81.3 (3.3) 33.9 (4.1) 17.0 (3.0) 16.0 (7.8)84.6 (3.3) 42.3 (9.7) 14.0 (3.2) 10.0 (1.7)0.030 0.087 0.033 0.0030.502 0.876 0.558 0.061ns ns ns ns

78.2 (4.9) 27.8 (1.7) 20.0 (4.3) 20.8 (8.3)78.4 (0.4) 30.5 (6.6) 19.7 (0.4) 27.9 (4.0)

86.4 (2.0) 45.6 (3.1) 12.2 (1.6) 15.6 (9.3)76.3 (3.6) 33.9 (7.6) 21.4 (3.7) 30.8 (6.9)0.404 0.088 0.364 0.8770.186 0.439 0.198 0.152ns ns ns ns

Fig. 5. Nitrate concentrations in undisturbed (a) and disturbed (b) soil cores.

Fig. 6. Ammonium concentrations in undisturbed (a) and disturbed (b) soil cores.

D. Buchan et al. / Soil Biology & Biochemistry 60 (2013) 142e155150

increased release of ammonium, which was much more pro-nounced in undisturbed cores where concentrations reached sim-ilar levels to those of nitrate. In disturbed cores, ammoniumconcentrations reached a plateau after 53 days around 10 mg N g�1

dry soil; due to nematode reinoculation this amount decreased toabout 50% of that level by the end of the experiment (Fig. 6b). Inundisturbed cores, levels in reinoculated soil similarly decreased toabout 10 mg N g�1 dry soil or 70% of those in non-inoculated soil, butonly after a strong decrease from a maximum around 25 mg N g�1

dry soil at 53 days of incubation (Fig. 6a).The net effect of nematode addition onmineral N was calculated

by subtracting mean concentrations of nitrate or ammonium indefaunated cores from those in the reinoculated cores at the initialand final sampling dates (5 and 82 days respectively). The balance isexpressed as the difference between the final and initial date. Fornitrate the balance was 13 mg N g�1 dry soil in undisturbed coresand 5 mg N g�1 dry soil in disturbed cores. For ammoniumthe balance was �9 mg N g�1 dry soil in undisturbed coresand �5 mg N g�1 dry soil in disturbed cores. Thus in terms of total

mineral nitrogen over the entire incubation period, the presence ofnematodes caused a positive balance of only 4 mg N g�1 dry soil inundisturbed cores, and no difference at all in disturbed cores.

4. Discussion

4.1. Microbial biomass carbon and microbial community structure

The aim of this experiment was not to increase or necessarilysustain a high microbial biomass, but rather to investigate whethermixing and disturbing soil for microcosm preparation, gammairradiating soil to remove soil fauna, and reinoculating irradiatedsoil with nematodes resulted in a modified microbial biomass.Steady state microbial of faunal biomasses allows a more reliableassessment of trophic effects on nutrient dynamics (Hunt et al.,1987). Nevertheless, incubation conditions and the probable rapidconsumption of any easily available resources combined todecrease the Cmic so strongly. While irradiation clearly reducedCmic, the difference in mean values is far smaller than commonlyoccurs between treatments and sampling times in numerous ex-periments (e.g. Djigal et al., 2004). The decrease in Cmic caused bya 5 kGy irradiation dose is considerably smaller than losses

D. Buchan et al. / Soil Biology & Biochemistry 60 (2013) 142e155 151

reported for the effect of fumigation onmicrobial biomass (Kemmitet al., 2008).

Reinoculation of nematodes negatively affected Cmic at thebeginning of the incubation in disturbed cores but was notdetectable in undisturbed cores. Previous incubation experimentshave reported the effects of nematode grazers onmicrobial biomassto range from positive (Bardgett et al., 1998; Standing et al., 2006) tonegative (Mikola and Setälä, 1999; Djigal et al., 2004); while othersreport little or no effect (Mikola, 1998; Villenave et al., 2004). Giventhat greatly differing conditions were used in these differentstudies, we can hardly expect to find a single generalised con-clusion concerning the impact of nematode grazing on microbialbiomass. Fu et al. (2005) demonstrated the positive feedback effectof nematodes on bacterial biomass and activity is very muchdensity-dependent, however only two nematode species anda single bacterium were used in their study. In the disturbed coresin this experiment, in which both the soil pore network and thedistribution of bio-available organic matter and microbial com-munities is likely to have been more homogenous, nematodes mayhave been brought into closer contact with their prey. In contrast, inundisturbed cores, microbial communities may have been enclosedwith existing aggregates separated by larger pores and thereforenot accessible to reinoculated nematodes. Such a mechanism isanalogous to the enclosure hypothesis proposed by Neher et al.(1999) and Savin et al. (2001) to explain differing nematode pop-ulation developments at different matric potentials. Given wefound surprisingly fewmicrocosm experiments involving soil faunawhere the soil microbial biomass was estimated by thefumigationeextraction method, we are hesitant to attribute thesignificantly higher microbial biomass in reinoculated undisturbedcores to microbial stimulation by nematode grazing.

4.2. PLFA profiles

Several mechanisms associated with the methodology we usedmay have caused changes in the structure of the microbial com-munity. Microbial taxa may differ in their sensitivity to low-doseirradiation, and some microbes may subsequently benefit fromreduced competition and increase in abundance (McNamara et al.,2007). Irradiation also caused a 55e75% increase in theconcentration of soluble organic carbon extracted during thedetermination of microbial biomass (data not shown) andinduced abiotic nitrogen mineralisation, thereby providinga greater pool of easily accessible organic matter sources formicro-organisms (Marschner and Bredow, 2002; Kemmit et al.,2008). Our PLFA results were not able to detect any significantchanges in microbial community structure other than a generaldecrease in most microbial markers throughout the incubationperiods, especially in disturbed cores. The PLFA profiles suggest thatfungi may be more sensitive to an irradiation dose of 5 kGy thanbacteria, a general tendency reported by McNamara et al. (2003)and Buchan et al. (2012). Furthermore, in disturbed cores Gram-positive bacteria and actinomycetes appear to have been neg-atively affected by the irradiation process. This was surprising giventhe radiation tolerances of these taxa were expected to lie muchhigher than 5 kGy based on the data reported by McNamara et al.(2003). The lack of significant differences in undisturbed coressuggests the mixing of soil might render physically occludedmicro-organisms more susceptible to irradiation.

The concentrations of PLFA biomarkers did not correlate verystrongly with Cmic (data not shown), in contrast to the confirmedrelationship derived from compiled data across ecosystem types(Bailey et al., 2002). It is known that in biologically suppressed soil,PLFAs of microbial origin persist even in the absence of anydetectable microbial biomass (Ranneklev and Bååth, 2003; Buchan

et al., 2012). PLFA analysis has been suggested to be more suited totrack changes in the composition of the microbial community inexperiments where organic amendments are added to soil than inthose where the soil biota is suppressed (Frostegård et al., 2011).We measured a relatively narrow range of PLFA concentrationscompared to other studies (see Bailey et al., 2002), which may alsoexplain the lack of a significant correlation between PLFA concen-tration and Cmic.

Generally, we were not able to detect any effects of nematodegrazing or its absence on the composition of the microbial com-munity as revealed by the multivariate analysis of PLFA profiles.Although PLFA profiles tend to give a more coarse level of resolu-tion of microbial diversity compared to what can be achieved byother methods such as DGGE (Djigal et al., 2004), Griffiths et al.(1999) did detect change in the microbial structure based on PLFAanalysis as well as DGGE and Biolog analyses. It is possible that inthe presence of a greater diversity of microbivorous nematodes inour study, a wider range of microbes was grazed upon, as opposedto a pronounced reduction in a narrow range of microbes by a sin-gle or small number of nematode species in most studies. The lackof change in the concentration of bacterial biomarkers in thepresence of bacterivorous nematodes may also be explained bysupplemental growth stimulated by either the availability of neworganic matter sources from recently deceased organisms, or by anextended habitat due to decreased competition from fungi. Theapparent decrease in the biomass of fungi is most probably due totheir greater sensitivity to gamma irradiation, after which fungalPLFA biomarkers may have been additionally degraded by bacteria.Compiled data from the literature (McNamara et al., 2003) suggeststhat the threshold of gamma irradiation resistance for fungi liesaround 8e10 kGy, however the studies cited did not include anywhich made use of PLFAs as a proxy for the biomass of fungi. Ameta-analysis of the most reliable methods available to estimatethe contribution of fungi to soil microbial biomass (Joergensen andWichern, 2008) convincingly demonstrated that the biomass ofphysiologically active fungi is best estimated using PLFA bio-markers. When using a partial sterilisation technique such asgamma irradiation, it is perhaps even more crucial to base esti-mation of fungal biomass on a sensitive indicator. Even so, it isprobable that PLFA biomarkers incorrectly estimate the change inbiomass of several if not most microbial taxa in soil subjected togamma irradiation.

4.3. Nematode population dynamics

4.3.1. Irradiation and extraction efficiencyThe zonal centrifuge extraction method we employed did not

segregate between living and dead nematodes, thus nematodeswere also extracted from defaunated cores throughout the durationof the experiment, although their abundance gradually decreasedas they were presumably decomposed. A previous study hadalready indicated that nematodes extracted from the defaunatedmicrocosms were indeed immobile, dead and devoid of inner bodyparts (Buchan et al., 2012). In this study it was also observed duringcounting that more than 90% of the nematodes in the defaunatedcores were not moving, while in reinoculated cores at least 60e70%of the nematodes were clearly mobile. If we assume that irradiatednematodes were decomposed to an equal extent in defaunated andreinoculated microcosms, then by the end of the experiment theabundance of living nematodes in the reinoculated microcosmswould be ca. 10 individuals g�1 soil, compared to 20 individuals g�1

soil in the control cores, resulting in a final 50% inoculation effi-ciency. Although considerably lower than the intended efficiency of70%, we purposefully chose to err on the side of caution, as over-inoculating the defaunated cores might have resulted in over-

D. Buchan et al. / Soil Biology & Biochemistry 60 (2013) 142e155152

grazing of microbial biomass and unrealistically high mineralisa-tion rates. Although actual populations in reinoculated cores wereonly half the size of that in control cores, similar abundances havebeen reported from several other incubation experiments (Ferriset al., 2004; Georgieva et al., 2005). Considering no nutrientswere added to microcosms to stimulate microbial activity, thereinoculation procedure may be considered to have successfullyestablished viable nematode populations in all cores, and wouldmost probably result in similarly sized populations if given theopportunity to grow.

The considerable percentage of unidentifiable nematodes in thereinoculated cores at the beginning of the experiment correspondsto the numerous nematodes recently killed by gamma irradiationthat were extracted by zonal centrifugation. The background levelof ca. 10% found in control cores and also in reinoculated cores bythe end of the incubation probably reflects recently deceasednematodes or individuals otherwise degraded and thereby diffi-cultly identified. The very high percentage observed in non-inoculated cores confirms that irradiated nematodes, althoughextracted together with viable living nematodes, are readily dif-ferentiated during standard microscopic identification.

4.3.2. Nematode community structureRoot-feeders and herbivorous nematodes persisted throughout

the entire duration of the experiment, although they decreased byabout a third in abundance in the reinoculated cores. While in mostcores the majority of these were Tylenchidae, which may beregarded as fungivorous (see below), Cephalenchus, Pratylenchusand Dolichodoridae were still present at the end of the experiment.Although we found one other experiment where such a phenome-non was observed (Villenave et al., 2004), we can only speculate asto the reason of their persistence. Either they were able to surviveby feeding on remaining root (hair) fragments in the soil or facul-tatively on fungal hyphae, or their thick cuticles were protectedfrom microbial decomposition and the individuals observed werein fact already dead. In the latter case, we may ignore root-feedingnematodes in our discussion of nitrogen mineralisation.

The abundance of fungivorous nematodes was correlated withthe fungal PLFA biomarker in disturbed cores only (R ¼ 0.78,p¼ 0.001), due in large part to Filenchus (R¼ 0.52, p¼ 0.048).Whenthe Tylenchidae were collectively considered as root-feeders, thetotal abundance of root feeders was weakly correlated with thefungal biomarker (R ¼ 0.55, p ¼ 0.035), when removed no signifi-cant correlations were observed, while the Tylenchidae by them-selves were positively correlated with the fungal biomarker(R ¼ 0.66, p ¼ 0.008 e disturbed cores only). These relationshipsstrongly suggest that the majority of Tylenchidae genera other thanFilenchus found in this study are capable of feeding on fungi.Despite the fact that several Filenchus species have convincinglybeen shown to be fungivorous (Okada and Kadota, 2003), a numberof authors still entirely classify Tylenchidae as root-feeders (e.g.Zelenev et al., 2004; Blanc et al., 2006; Leroy et al., 2007). While theuncertainty of the feeding mechanisms of Tylenchidae as awhole isstill acknowledged (Forge et al., 2003; Christensen et al., 2007),there is a growing tendency to consider them as being fungivorous(Wang et al., 2006; Sánchez-Moreno et al., 2010; Carrillo et al.,2011). Until the urgent need to study the trophic preferenceTylenchidae has been met (Yeates, 2003), it is probably mostappropriate to consider to them as ‘facultative root feeders’ (Okadaand Harada, 2007) and carefully note their dynamics separately inlaboratory experiments.

As reported in other studies (Laakso and Setälä, 1999; Mikolaand Sulkava, 2001; Ferris et al., 2004; Leroy et al., 2009), the fun-givorous nematodes Aphalenchus and Aphalenchoides were presentin very low abundances compared to bacterivorous nematodes.

Their generally low abundance may explain a lack of correlationwith fungal PLFA biomarkers, and it is doubtful that in a realisticmicrocosm experiment such as this one, they may have affectedfungal biomass in any significant way.

The numerical dominance of the bacterivorous Cephalobidaeacross numerous soil types is a well-established fact (Yeates, 2003)that reflects their adaptability to a wide range of environmentalconditions. In the control cores, Cephalobidae appeared to remainmore abundant in undisturbed than in disturbed soil. This couldpartly be explained by the sensitivity of several genera of Cepha-lobidae to disturbance (Fiscus and Neher, 2002). Given the lowabundance of the enrichment opportunists and (known) fungivoresin this experiment, it is likely that most microbial grazing andconsequent nitrogen mineralisation was carried out by the ubiq-uitous Cephalobidae. The generally low microbial biomass limitedthe development of significant bacterivorous nematode pop-ulations and probably explains the lack of a strong effect of nem-atode addition on nitrogen mineralisation (see below).

Our data suggest that in control cores Rhabditidae were able tomake best use of the highmicrobial biomass at the beginning of theexperiment. As the microbial biomass decreased due to a lack oforganic resources, the abundance of Rhabditidae dropped by half toreach similar levels as in the reinoculated cores. Panagrolaimidaemaintained a greater background presence in undisturbed cores, inwhich Diplogasteridae tended to be rare. In contrast, Diplogaster-idae greatly increased in abundance only in the reinoculated dis-turbed cores, in which Panagrolaimidae in turn were present in thelowest amounts. These relationships suggest that in all reinocu-lated cores Rhabditidae could not establish themselves to the sameextent as in control cores, while Panagrolaimidae and Dip-logasteridae were able to establish themselves better in undis-turbed and disturbed cores respectively. These apparent shifts inthe population structure of the enrichment opportunists couldhave been brought about either directly by differing abilities ofthese nematode taxa to survive the reinoculation process and adaptto changing physico-chemical conditions, or indirectly by shifts incommunity structure of the consumed bacterial populations due tomicrocosm preparation and irradiation. Due to high variability andoverall low nematode abundances, our data is not a conclusivedemonstration of inter-specific effects between bacterivoresbelonging to a same trophic guild (sensu Brussaard,1998). However,given the commonly employed ecological indices based on thecoloniserepersister classification (Bongers, 1990, 1999) group all c-p 1 bacterivores into a single functional group, the possibility ofsuch interactions occurring in microcosm conditions certainlywarrants further study as well as caution with the interpretation ofthese indices. Despite these overall trends, populations sometimesappeared to develop erratically in individual cores. Unexplainablevariability in the occurrence of these families taxa of bacterivorousnematodes and their dauer larvae forms has been noted before(Griffiths, 1994) and may be due to chaotic population dynamics ofisolated nematode populations in individual and separate soil mi-crocosms (Sohlenius, 1993). The high proportion of dauer larvaeversus that of their active counterparts confirms the pattern of Cmic,indicating that conditions of nutrient limitation prevailedthroughout the length of the experiment.

As nematodes with a high c-p rank (sensu Bongers, 1990),(dorylamid) omnivores are usually the first to be affected byadverse environmental conditions and tend to be absent fromstrongly disturbed soil (Ferris et al., 2001). However Ettema andBongers (1993) found Aporcelaimellus responded positively andrapidly to unicellular algal blooms, while Ferris et al. (2001) alsoreported an increase in the abundance of c-p 5 carnivores indefaunated soil. Their maintained or increased presence in thisstudy indicates that stable populations were established during the

D. Buchan et al. / Soil Biology & Biochemistry 60 (2013) 142e155 153

incubation time frame and within the experimental conditionscreated. Few incubation experiments to date have attempted toinclude higher-level omnivorous and predatory nematodes in theirdesign. Where nematodes other than the most commonlyemployed bacterivores and fungivores have been inoculated, spe-cies extinction has been reported (Mikola and Setälä, 1999; Laaksoet al., 2000). If inoculations of single species prove to be unsuc-cessful for introducing higher-level predatory and omnivorousnematodes into controlled experiments, then a bulk inoculationtechnique as used in this study may be the only way to create a full-spectrum nematode community.

Although our results do not permit us to asses whetheromnivorous and predatory nematodes had an impact on thenumber of other nematodes, their increase in the reinoculatedcores could explain the lower total nematode abundance comparedto the control, as was found by Wardle et al. (1995). It is also pos-sible that the lower abundance of omnivorous nematodes in con-trol samples was due to the presence of higher predators such asmites, which despite not being sampled in this experiment, werenever observed in any of the nematode extracts used to reinocu-lated irradiated cores. We did not verify the presence of micro-arthropods in the control cores but regularly detected bothcollembola and mites in the nematode extracts. Results from somefood web models have shown predation by predatory nematodesmay affect the mineralisation potential of bacterivorous nematodes(De Ruiter et al., 1993).

The maturity and community indices failed to fully describe thechanges in the composition of the nematode community structureas discussed above. Aggregation of nematodes of different familieswith similar life histories classes into a single c-p class assumesthey have equal responses to changing soil ecological conditions.Our data suggests this may especially not be the case for bacter-ivorous enrichment opportunists, which in agricultural soil areoften the most dominant taxa (Neher, 2010). Inconsistencies be-tween observed and predicted responses of nematode taxa fol-lowing their prescribed c-p classification have been highlighted byYeates (2003). On the other hand, Ferris et al. (2004) found thatenrichment and channel indices were suitable predictors of meanyield and nitrogen mineralisation over an entire season in a fieldtrial. The use of the maturity and associated indices is perhaps bestused to compare soil nematode populations between different sitesor agronomic management practices (Berkelmans et al., 2003).

4.4. Nitrogen dynamics

As expected, the irradiation procedure caused a flush ofammonium, however this was rather modest compared to thosereported from other studies using other methods to sterilise ordefaunate microcosms (e.g. Anderson et al., 1979; Mikola andSetälä, 1999). In disturbed cores, peak ammonium concentrationswere considerably lower than in undisturbed cores, which may bedue to a greater microbial re-assimilation as a result of mixing thesoil (Woods et al., 1982). Alternatively, nitrifiers could have rees-tablished themselves more effectively in disturbed than in undis-turbed soil cores following irradiation, which is supported by thegenerally greater nitrate concentrations in disturbed cores. Whileincubation studies involving only selected heterotrophic bacteriahave mostly only considered the effect of nematode grazing onammonium concentrations (Anderson and Coleman, 1981; Inghamet al., 1985), in our case the presence of autotrophic nitrifiers in-validates such an approach. In both core types, nematode reino-culation clearly resulted in a reduction of ammonium levels and anincrease of nitrate levels, suggesting nematode microbial grazersstimulate the nitrifier community in soil. This was fist suggested byGriffiths (1986) and subsequently demonstrated by Xiao et al.

(2010) in a similar experiment where sterile microcosm were ino-culated with a bacterial suspension and entire nematodecommunities.

The bulk of the evidence accumulated so far from microcosmexperiments with limited species assemblages suggests microbialgrazing by nematodes substantially increases total nitrogen min-eralisation (Verhoef and Brussaard, 1990; Griffiths, 1994; Ferriset al., 1998). However our results do not indicate that the inclu-sion of nematodes in realistic microcosms results in greater totalnitrogen mineralisation. However microbivorous nematodesclearly influence the abundance and activity of nitrifiers, and maythereby exert strong controls on the fate of plant-available forms ofmineral nitrogen in agricultural soil.

Although we were able to fully control the presence or absenceof viable nematode populations, we did not consider the popu-lation dynamics of protozoa in response to irradiation and rein-oculation. Protozoa, due to their higher efficiency as bacterialgrazers, have been shown to be the main contributor to nitrogenmineralisation amongst soil fauna (Hunt et al., 1987; Verhoef andBrussaard, 1990). Griffiths (1986) found that nematodes miner-alized less when in presence of protozoa, due to the greater abilityof protozoa to forage for bacteria in smaller pores. Although, wefound no studies reporting the effect of gamma irradiation onentire protozoan communities in soil, Fuma et al. (2010) foundthat a species of ciliate in an aquatic microcosm was entirelyeliminated in a matter of days following a 5 kGy irradiation dose,while two species of rotifer survived for several weeks. Theseresults from an aquatic environment with a single species cannotpossibly be extrapolated to soil, however it is likely that gamma-induced radiolysis for a given dose is far more extensive in a ho-mogenous aqueous medium than in a highly heterogeneous andcomplex soil environment. Given that as a general rule the sen-sitivity to irradiation tends to be positively correlated with thecomplexity of an organism (McNamara et al., 2003), it is probablethat at least a portion of the protozoan community survivedirradiation in our study. If it is assumed that a viable population ofprotozoa survived irradiation, then the mineralisation observed inthe ‘defaunated’ cores may be the result of protozoan grazing onbacteria rather than only mineralisation caused by primary con-sumers themselves. However in the present study we cannotassess the magnitude of this process, as it cannot be distinguishedfrom abiotic mineralisation mechanisms resulting from gammairradiation (Buchan et al., 2012) or mineralisation due to microbialturnover. The inclusion of representative populations of protozoain microcosm experiments and their reliable quantificationalongside that of free-living nematodes is an acute research need,which needs to be addressed if we are to develop an in-depthunderstanding of the effect of microbial grazing on nitrogendynamics.

5. Conclusions

This study demonstrated that defaunation by gamma irradiationfollowed by selective reinoculation of nematodes allows the prep-aration of microcosms whose microbial community and a compo-nent of higher soil biota more closely resemble those of the originalsoil than is characteristic of most experimental setups to date. Suchan approach could be applied to other groups of soil fauna,although the possible bias of extraction and inoculation methodsshould be investigated beforehand. The methodology developed isequally applicable to undisturbed soil cores, making it attractive forexperiments involving interactions between the soil pore spacenetwork and the soil biota. This is of potential significance, as ourresults clearly show that the preparation method of soil micro-cosms exerts a strong influence on nitrogen dynamics.

D. Buchan et al. / Soil Biology & Biochemistry 60 (2013) 142e155154

The reinoculation of nematode populations into defaunated soilcores clearly affected the forms of mineral nitrogen but did notincrease the total amount of nitrogen mineralised. Nitrate levelswere increased at the expense of ammonium concentrations inreinoculated microcosms, suggesting a strong stimulation of thesoil nitrifier community by nematode grazing.

Our results have two important implications: (1) using undis-turbed cores instead of disturbed ones strongly determines theresponse of nitrogen mineralisation to defaunation and/or reino-culation practices; and (2) the removal and reinoculation of entirefree-living nematode populations is perfectly possible while usingundisturbed cores that preserve the original field habitat of soilorganisms.

Acknowledgements

This research was conducted within a project financed by theSpecial Research Fund from Ghent University (BOF e UGent).

References

Anderson, R.V., Coleman, D.C., 1981. Population development and interactions be-tween two species of bacteriophagic nematodes. Nematologica 27, 6e19.

Anderson, J.M., Ineson, P., 1982. A soil microcosm system and its application tomeasurements of respiration and nutrient leaching. Soil Biology & Biochemistry14, 415e416.

Anderson, R.V., Coleman, D.C., Cole, C.V., Elliott, E.T., McClellan, J.F., 1979. The use ofsoil microcosms in evaluating bacteriophagic nematode responses to otherorganisms and effects on nutrient cycling. International Journal of Environ-mental Studies 13, 175e182.

Anderson, R.V., Coleman, D.C., Cole, C.V., Elliott, E.T., 1981. Effect of the nematodesAcrobeloides sp. and Mesodiplogaster lheritieri on substrate utilization and ni-trogen and phosphorous mineralization in soil. Ecology 62, 549e555.

Bailey, V.L., Peacock, A.D., Smith, J.L., Bolton, H., 2002. Relationships between soilmicrobial biomass determined by chloroform fumigationeextraction,substrate-induced respiration, and phospholipid fatty acid analysis. Soil Biology& Biochemistry 34, 1385e1389.

Bardgett, R.D., Chan, K.F., 1999. Experimental evidence that soil fauna enhancenutrient mineralization and plant nutrient uptake in montane grassland eco-systems. Soil Biology & Biochemistry 31, 1007e1014.

Bardgett, R.D., Keiller, S., Cook, R., Gilburn, A.S., 1998. Dynamic interactions betweensoil animals and microorganisms in upland grassland soils amended with sheepdung: a microcosm experiment. Soil Biology & Biochemistry 30, 531e539.

Beare, M.H., Parmelee, R.W., Hendrix, P.F., Cheng, W., 1992. Microbial and faunalinteractions and effects on litter nitrogen and decomposition in agro-ecosystems. Ecological Monographs 62, 569e591.

Berkelmans, R., Ferris, H., Tenuta, M., van Bruggen, A.H.C., 2003. Effects of long-termcrop management on nematode trophic levels other than plant feeders dis-appear after 1 year of disruptive soil management. Applied Soil Ecology 23,223e235.

Blanc, C., Sy, M., Djigal, D., Brauman, A., Normand, P., Villenave, C., 2006.Nutrition on bacteria by bacterial-feeding nematodes and consequences onthe structure of soil bacterial community. European Journal of Soil Biology42, S70eS78.

Bligh, E.G., Dyer, W.J., 1959. A rapid method of total lipid extraction and purification.Canadian Journal of Biochemistry and Physiology 37, 911e917.

Bongers, T., 1988. De Nematoden van Nederland. Stichting Uitgeverij KoninklijkeNederlandse Natuurhistorische Vereniging, Utrecht.

Bongers, T., 1990. The maturity index: an ecological measure of environmentaldisturbance based on nematode species composition. Oecologia 83, 14e19.

Bongers, T., 1999. The Maturity Index, the evolution of nematode life history traits,adaptive radiation and cp-scaling. Plant & Soil 212, 13e22.

Bongers, T., Ferris, H., 1999. Nematode community structure as a bioindicator inenvironmental monitoring. Trends in Ecology & Evolution 14, 224e228.

Bruckner, A., Wright, J., Kampichler, C., Bauer, R., Kandeler, E., 1995. A method ofpreparing mesocosms for assessing complex biotic processes in soils. Biologyand Fertility of Soils 19, 257e262.

Brussaard, L., 1998. Soil fauna, guilds, functional groups and ecosystem processes.Applied Soil Ecology 9, 123e135.

Brussaard, L., Bouwman, L.A., Geurs, M., Hassink, J., Zwart, K.B., 1990. Biomass,composition and temporal dynamics of soil organism of a silt loam soil underconventional and integrated management. Netherlands Journal of AgriculturalScience 38, 283e302.

Buchan, D., Moeskops, B., Ameloot, N., De Neve, S., Sleutel, S., 2012. Selective ster-ilisation of undisturbed soil cores by gamma irradiation: effects on free-livingnematodes, microbial community and nitrogen dynamics. Soil Biology & Bio-chemistry 47, 10e13.

Carrillo, Y., Ball, B.A., Bradford, M.A., Jordan, C.F., Molina, M., 2011. Soil fauna altersthe effects of litter composition on nitrogen cycling in a mineral soil. SoilBiology & Biochemistry 43, 1440e1449.

Chen, J., Ferris, H., 1999. The effects of nematode grazing on nitrogen mineralizationduring fungal decomposition of organic matter. Soil Biology & Biochemistry 31,1265e1279.

Christensen, S., Alphei, J., Vestergard, M., Vestergaard, P., 2007. Nematode migrationand nutrient diffusion between vetch and barley material in soil. Soil Biology &Biochemistry 39, 1410e1417.

Clarholm, M., 1985. Interactions of bacteria, protozoa and plants leading to min-eralization of soil nitrogen. Soil Biology & Biochemistry 17, 181e187.

Coleman, D.C., Cole, V.C., Hunt, H.W., Klein, D.A., 1978. Trophic interaction in soils asthey affect energy and nutrient dynamics. I. Introduction. Microbial Ecology 4,345e349.

Crawley, J.W., Harris, J.A., Ritz, K., Young, I.M., 2005. Towards an evolutionaryecology of life in soil. Trends in Ecology & Evolution 20, 81e87.

De Ruiter, P.C., van Veen, J.A., Moore, J.C., Brussaard, L., Hunt, H.W., 1993. Calculationof nitrogen mineralisation in soil food webs. Plant & Soil 157, 263e273.

Djigal, D., Brauman, A., Diop, T.A., Chotte, J.L., Villenave, C., 2004. Influence ofbacterial-feeding nematodes (Cephalobidae) on soil microbial communitiesduring maize growth. Soil Biology & Biochemistry 36, 323e331.

Ettema, C.H., Bongers, T., 1993. Characterization of nematode colonization andsuccession in disturbed soil using the Maturity Index. Biology and Fertility ofSoils 16, 79e85.

Ferris, H., Venette, R.C., van der Meulen, H.R., Lau, S.S., 1998. Nitrogen mineraliza-tion by bacterial-feeding nematodes: verification and measurement. Plant &Soil 203, 159e171.

Ferris, H., Bongers, T., de Goede, R.G.M., 2001. A framework for soil food web di-agnostics: extension of the nematode faunal analysis concept. Applied SoilEcology 18, 13e29.

Ferris, H., Venette, R.C., Scow, K.M., 2004. Soil management to enhance bacterivoreand fungivore nematode populations and their nitrogen mineralisation func-tion. Applied Soil Ecology 25, 19e35.

Fiscus, D., Neher, D., 2002. Distinguishing sensitivity of free-living soil nematodegenera to physical and chemical disturbances. Ecological Applications 12,565e575.

Forge, T.A., Hogue, E., Neilsen, G., Neilsen, D., 2003. Effects of organic mulches onsoil microfauna in the root zone of apple: implications for nutrient fluxes andfunctional diversity of the soil food web. Applied Soil Ecology 22, 39e54.

Frostegård, Å., Tunlid, A., Bååth, E., 2011. Use and misuse of PLFA measurements insoils. Soil Biology & Biochemistry 43, 1621e1625.

Fu, S., Ferris, H., Bown, D., Plant, R., 2005. Does the positive feedback effect ofnematodes on the biomass and activity of their bacteria prey vary with nem-atode species and population size? Soil Biology & Biochemistry 37, 1979e1987.

Fuma, S., Ishii, N., Takeda, H., Doi, K., Kawaguchi, I., Tanaka, N., Inamori, Y., 2010.Effects of acute g-irradiation on community structure of the aquatic microbialmicrocosm. Journal of Environmental Radioactivity 101, 915e922.

Georgieva, S., Christensen, S., Petersen, H., Gjelstrup, P., Thorup-Kristensen, K., 2005.Early decomposer assemblages of soil organisms in litterbags with vetch andrye roots. Soil Biology & Biochemistry 37, 1145e1155.

Görres, J.H., Savin, M.C., Neher, D.A., Weicht, T.R., Amador, J.A., 1999. Grazing ina porous environment: 1. The effect of soil pore structure on C and N miner-alization. Plant & Soil 212, 75e83.

Griffiths, B.S., 1986. Mineralization of nitrogen and phosphorus by mixed cultures ofthe ciliate protozoan Colpoda steinii, the nematode Rhabditis sp. and the bac-terium Pseudomonas fluorescens. Soil Biology & Biochemistry 18, 637e641.

Griffiths, B.S., 1994. Microbial-feeding nematodes and protozoa in soil: their effectsof microbial activity and nitrogen mineralization in decomposition hotspotsand the rhizosphere. Plant & Soil 164, 25e33.

Griffiths, B.S., Bonkowski, M., Dobson, G., Caul, S., 1999. Changes in soil microbialcommunity structure in the presence of microbial-feeding nematodes andprotozoa. Pedobiologia 43, 297e304.

Griffiths, B.S., Hallett, P.D., Kuan, H.L., Gregory, A.S., Watts, C.W., Whitmore, A.P.,2008. Functional resilience of soil microbial communities depends on both soilstructure and microbial community composition. Biology and Fertility of Soils44, 745e754.

Hendrickx, G., 1995. An automatic apparatus for extracting free-living nematodestages from soil. Nematologica 41, 308 (abstract).

Huhta, V., 2007. The role of soil fauna in ecosystems: a historical review. Pedo-biologia 50, 489e495.

Hunt, H.W., Coleman, D.C., Ingham, E.R., Ingham, R.E., Elliott, E.T., Moore, J.C.,Rose, S.L., Reid, C.P.P., Morley, C.R., 1987. The detrital food web in a shortgrassprairie. Biology and Fertility of Soils 3, 57e68.

Ingham, R.E., Trofymow, J.A., Ingham, E.R., Coleman, D.C., 1985. Interactions ofbacteria, fungi, and their nematode grazers: effects on nutrient cycling andplant growth. Ecological Monographs 55, 119e140.

Joergensen, R.G., 1996. The fumigationeextraction method to estimate soil microbialbiomass: calibration of the kEC value. Soil Biology & Biochemistry 28, 25e31.

Joergensen, R.G., Wichern, F., 2008. Quantitative assessment of the fungal con-tribution to microbial tissue in soil. Soil Biology & Biochemistry 40, 2977e2991.

Jones, T.H., Thompson, L.J., Lawton, J.H., Bezemer, T.M., Bardgett, R.D.,Blackburn, T.M., Bruce, K.D., Cannon, P.F., Hall, G.S., Hartley, S.E., Howson, G.,Jones, C.G., Kampichler, C., Kandeler, E., Ritchie, D.A., 1998. Impacts of risingatmospheric carbon dioxide on model terrestrial ecosystems. Science 280,441e443.

D. Buchan et al. / Soil Biology & Biochemistry 60 (2013) 142e155 155

Kampichler, C., Bruckner, A., Baumgarten, A., Berthold, A., Zechmeister-Boltenstern, S., 1999. Field mesocosms for assessing biotic processes insoils: how to avoid side effects. European Journal of Soil Biology 35,135e143.

Kampichler, C., Bruckner, A., Kandeler, E., 2001. Use of enclosed model ecosystemsin soil ecology: a bias towards laboratory research. Soil Biology & Biochemistry33, 269e275.

Kemmit, S.J., Lanyon, C.V., Waite, I.S., Wen, Q., Addiscott, T.M., Bird, N.R.A.,O’Donnell, A.G., Brookes, P.C., 2008. Mineralization of native organic matter isnot regulated by the size, activity or composition of the soil microbial biomasse a new perspective. Soil Biology & Biochemistry 40, 61e73.

Laakso, J., Setälä, H., 1999. Population- and ecosystem-level effects of predation onmicrobial-feeding nematodes. Oecologia 120, 279e286.

Laakso, J., Setälä, H., Palojärvi, A., 2000. Influence of decomposer food web structureand nitrogen availability on plant growth. Plant & Soil 225, 153e165.

Leroy, B.L.M.M., Bommele, L., Reheul, D., Moens, M., De Neve, S., 2007. The appli-cation of vegetable, fruit and garden waste (VFG) compost in addition to cattleslurry in a silage maize monoculture: effects on soil fauna and yield. EuropeanJournal of Soil Biology 43, 91e100.

Leroy, B.L.M., De Sutter, N., Ferris, H., Moens, M., Reheul, D., 2009. Short-termnematode population dynamics as influenced by the quality of exogenousorganic matter. Nematology 11, 23e38.

Marschner, B., Bredow, A., 2002. Temperature effects on release and ecologicallyrelevant properties of dissolved organic carbon in sterilised and biologicallyactive soil samples. Soil Biology & Biochemistry 34, 459e466.

McNamara, N.P., Black, H.I.J., Beresford, N.A., Parekh, N.R., 2003. Effects of acutegamma irradiation on chemical, physical and biological properties of soils.Applied Soil Ecology 24, 117e132.

McNamara, N.P., Griffiths, R.I., Tabouret, A., Beresford, N.A., Bailey, M.J.,Whiteley, A.S., 2007. The sensitivity of a forest soil microbial community toacute gamma-irradiation. Applied Soil Ecology 37, 1e9.

Mikola, J., 1998. On the role of varying species combinations in microcosm exper-iments: how to test ecological theories with soil food webs. Annal ZoologicaFennici 35, 63e66.

Mikola, J., Setälä, H., 1999. Interplay of omnivory, energy channels and C availabilityin a microbial-based soil food web. Biology and Fertility of Soils 28, 212e218.

Mikola, J., Sulkava, P., 2001. Response of microbial-feeding nematodes to organicmatter distribution and predation in experimental soil habitat. Soil Biology &Biochemistry 33, 811e817.

Moeskops, B., Sukristiyonubowo, Buchan, D., Sleutel, S., Herawaty, L., Husen, E.,Saraswati, R., Setyorini, D., De Neve, S., 2010. Soil microbial communities andactivities under intensive organic and conventional vegetable farming in WestJava, Indonesia. Applied Soil Ecology 45, 112e120.

Moore, J.C., de Ruiter, P.C., 1991. Temporal and spatial heterogeneity of trophic in-teractions within below-ground food webs. Agriculture, Ecosystems & Envi-ronment 34, 371e397.

Neher, D., 2010. Ecology of plant and free-living nematodes in natural and agri-cultural soil. Annual Review of Phytopathology 48, 371e394.

Neher, D.A., Weicht, T.R., Savin, M., Görres, J.H., Amador, J.A., 1999. Grazing ina porous environment. 2. Nematode community structure. Plant & Soil 212,85e99.

Okada, H., Harada, H., 2007. Effects of tillage and fertilizer on nematode commu-nities in a Japanese soybean field. Applied Soil Ecology 35, 582e598.

Okada, H., Kadota, I., 2003. Host status of 10 fungal isolates for two nematodespecies, Filenchus misellus and Aphelenchus avenae. Soil Biology & Biochemistry35, 160e1607.

Peterson, S.O., Debosz, K., Schjønning, P., Christensen, B.T., Elmholt, S., 1997. Phos-pholipid fatty acid profiles and C availability in wet-stable macro-aggregatesfrom conventionally and organically farmed soils. Geoderma 78, 181e196.

Ranneklev, S.B., Bååth, E., 2003. Use of phospholipid fatty acids to detect previousself-heating events in stored peat. Applied and Environmental Microbiology 69,3532e3539.

Sánchez-Moreno, S., Ferris, H., Young-Mathews, A., Culman, S.W., Jackson, L.E., 2010.Abundance, diversity and connectance of soil food web channels along envi-ronmental gradients in an agricultural landscape. Soil Biology & Biochemistry43, 2374e2383.

Savin, M.C., Görres, J.H., Neher, D.A., Amador, J.A., 2001. Uncoupling of carbon andnitrogen mineralization: role of microbivorous nematodes. Soil Biology & Bio-chemistry 33, 1463e1472.

Sohlenius, B., 1993. Chaotic or deterministic development of nematode populationsin pine forest humus incubated in the laboratory. Biology and Fertility of Soils16, 262e268.

Standing, D., Knox, O.G.G., Mullins, C.E., Killham, K.K., Wilson, M.J., 2006. Influenceof nematodes on resource utilization by bacteria d an in vitro study. MicrobialEcology 52, 444e450.

Sulkava, P., Huhta, V., 1998. Habitat patchiness affects decomposition and faunaldiversity: a microcosm experiment on forest floor. Oecologia 116, 390e396.

Vance, E.D., Brookes, P.C., Jenkinson, D.S., 1987. An extraction method for measuringsoil microbial biomass C. Soil Biology & Biochemistry 19, 703e707.

Verhoef, H.A., Brussaard, L., 1990. Decomposition and nitrogen mineralization innatural and agro-ecosystems: the contribution of soil animals. Biogeochemistry11, 175e211.

Villenave, C., Ekschmitt, K., Nazaret, S., Bongers, T., 2004. Interactions betweennematodes and microbial communities in a tropical soil following manipulationof the soil food web. Soil Biology & Biochemistry 36, 2033e2043.

Wang, K., McSorley, R., Marshall, A., Gallaher, R., 2006. Influence of organic Crota-laria juncea hay and ammonium nitrate fertilizers on soil nematode commu-nities. Applied Soil Ecology 31, 186e198.

Wardle, D.A., Yeates, G.W., Watson, R.N., Nicholson, K.S., 1995. The detritus food-web and the diversity of soil fauna as indicators of disturbance regimes inagro-ecosytems. Plant & Soil 170, 35e43.

Woods, L.E., Cole, C.V., Elliott, E.T., Anderson, R.V., Coleman, D.C., 1982. Nitrogentransformations in soil as affected by bacterial-macrofaunal interactions. SoilBiology & Biochemistry 14, 93e98.

Xiao, H., Griffiths, B., Chen, X., Liu, M., Jiao, J., Hu, F., Li, H., 2010. Influence ofbacterial-feeding nematodes on nitrification and the ammonia-oxidizing bac-teria (AOB) community composition. Applied Soil Ecology 45, 131e137.

Yeates, G.W., 2003. Nematodes as soil indicators: functional and biodiversity as-pects. Biology and Fertility of Soils 37, 199e210.

Yeates, G.W., Bongers, T., de Goede, R.G.M., Freckman, D.W., Georgieva, S.S., 1993.Feeding habits in soil nematode families and genera e an outline for soilecologists. Journal of Nematology 25, 315e331.

Yeates, G.W., Dando, J.L., Shepherd, T.G., 2002. Pressure plate studies to determinehow moisture affects access of bacterial-feeding nematodes to food in soil.European Journal of Soil Science 53, 355e365.

Zelenev, V.V., Berkelmans, R., van Bruggen, A.H.C., Bongers, T., Semenov, A.M., 2004.Daily changes in bacterial-feeding nematode populations oscillate with similarperiods as bacterial populations after a nutrient impulse in soil. Applied SoilEcology 26, 93e106.

Zelles, L., 1999. Fatty acid patterns of phospholipids and lipopolysaccharides in thecharacterization of microbial communities in soil: a review. Biology and Fer-tility of Soils 29, 111e129.