lable at ScienceDirect

Soil Biology & Biochemistry 76 (2014) 45e52

Contents lists avai

Soil Biology & Biochemistry

journal homepage: www.elsevier .com/locate/soi lb io

Exploring carbon flow through the root channel in a temperate forestsoil food web

Kadeem J. Gilbert a,1, Timothy J. Fahey a,*, John C. Maerz b, Ruth E. Sherman a,Patrick Bohlen c, Jason J. Dombroskie d, Peter M. Groffman e, Joseph B. Yavitt a

aDepartment of Natural Resources, Cornell University, Fernow Hall, Ithaca, NY 14853, USAbWarnell School of Forestry and Natural Resources, University of Georgia, 180 East Green Street, Athens, GA 30602, USAcDepartment of Biology, University of Central Florida, Orlando, FL 32816, USAdDepartment of Entomology, Cornell University, Comstock Hall, Ithaca, NY 14853, USAeCary Institute of Ecosystem Studies, Millbrook, NY 12545-0129, USA

a r t i c l e i n f o

Article history:Received 15 January 2014Received in revised form25 April 2014Accepted 3 May 2014Available online 19 May 2014

Keywords:AntsCarbonEarthwormsEnergy flowInvasive speciesSalamander

* Corresponding author. Tel.: þ1 607 255 5470.E-mail addresses: kgilbert@fas.harvard.edu (K.J.

J. Fahey), jcmaerz@uga.edu (J.C. Maerz), res6@cornelBohlen@ucf.edu (P. Bohlen), jjd278@cornell.edu (J.caryinstitute.org (P.M. Groffman), jby1@cornell.edu (J

1 Present address: Department of Organismic and EUniversity, Cambridge, MA, USA.

http://dx.doi.org/10.1016/j.soilbio.2014.05.0050038-0717/� 2014 Elsevier Ltd. All rights reserved.

a b s t r a c t

Soil food webs play an important role in forest ecosystem functions that may be sensitive to globalenvironmental change. We traced a pulse of 13C from the root systems of young sugar maple standsthrough the soil food web to explore energy flow from roots to soil heterotrophs. Invasive earthworms(Lumbricidae) were the most highly-enriched taxa sampled, indicating that they were consuming sig-nificant quantities of live fine roots and mycorrhizae and assimilating significant amounts of root-derivedC. Another invasive invertebrate, a weevil (Barypeithes pellucidus) also appeared to consume significantamounts of roots or root-derived C. High isotope enrichment in ants (Formicidae) may reflect theirfeeding on phloem-sucking coccids in soil. All the predators collected from the litter layer also exhibitedhigh isotope enrichment, including the salamander Plethodon cinereus. Moreover, in the absence ofinvasive earthworms all the taxa of animals collected from surface soil were similarly highly-enriched inroot-derived isotopes. The apparent importance of the root channel in supplying energy to the soil foodweb suggests that forest ecosystem functions could be highly sensitive to global changes that alterproportional tree C allocation belowground.

� 2014 Elsevier Ltd. All rights reserved.

1. Introduction

In the face of global environmental change, the dynamics of soilfood webs need to be better resolved because of their likelyimportance for terrestrial ecosystem responses to changingclimate, pollution loading and invasive species (Wardle et al.,2004). Soil food webs play a major role in soil carbon stabiliza-tion (Brussaard et al., 2007) and nutrient recycling (Bonkowskiet al., 2009), and invasive species can disrupt these keyecosystem functions (Fahey et al., 2012). Energy is supplied to soilfood webs by two principal input channels, aboveground detritus

Gilbert), tjf5@cornell.edu (T.l.edu (R.E. Sherman), Patrick.J. Dombroskie), groffmanp@.B. Yavitt).volutionary Biology, Harvard

and roots, both of which contribute directly or indirectly to a thirdenergy source, older soil organic matter. Evidence suggests that theliving root channel may predominate in energy supply to the soilfood web in temperate forests (Pollierer et al., 2007, 2012). This isnot surprising because roots supply a high proportion of C inputs toforest soil (Kuzyakov and Domanski, 2000; Matamala et al., 2003),and leaf litter on the soil surface is often biochemically recalcitrant(e.g., high C:N ratio and high lignin concentration) and subject toenvironmental extremes (e.g., drying, heating, freezing).

The nature of the root-based energy channel of the soil foodwebhas received limited study because it is difficult to observe.Photosynthetic C supplied to roots can reach soil heterotrophsdirectly via living roots and root hairs (grazing), but also indirectlyvia root detritus, root exudates and mycorrhizal fungi. The feedinghabits of many taxonomic groups of soil animals have been clas-sified in general (e.g., detritivores, fungivores, omnivores, preda-tors), and food web linkages are a topic of great interest (Scheu,2002). The basal resources supplying the soil food web have beenevaluated qualitatively based on fatty acid (FA) signatures that

K.J. Gilbert et al. / Soil Biology & Biochemistry 76 (2014) 45e5246

distinguish among plant, fungal and bacterial sources (Ruess andChamberlain, 2010). Recently, Pollierer et al., (2012) utilized FAsignatures in combination with stable isotopes of C to demonstratethe details of the soil food web in a beech (Fagus sylvatica)-domi-nated forest. They demonstrated the predominant role of ectomy-corrhizal fungi and rhizosphere bacteria as basal resources for thefood web in this forest, as well as the lesser importance of leaf littersources. They also clarified the feeding relationships among Col-lembola, mites, native earthworms and predators. Analogousstudies have not been reported in forests dominated by trees witharbuscular mycorrhizal associations.

Invasive soil invertebrates can profoundly disrupt the normalfunctioning of soil food webs. Most prominent in many northernforest ecosystems in this respect are large earthworms (Anne-lida) (Hale et al., 2005; Eisenhauer et al., 2011; Greiner et al.,2012). For example, earthworm invasion of temperate broadleafforest reduced the bacterial:fungal ratios in soils (Dempsey et al.,2011, 2013), altered and simplified the oribatid mite (Acari) fauna(Burke et al., 2011), and altered the diet and abundance of forestfloor consumers such as salamanders (Maerz et al., 2005, 2009)and birds (Loss and Blair, 2011; Loss et al., 2012). The principalmechanism underlying these earthworm-induced changes wasthe elimination of surface organic horizons by earthwormfeeding and mixing, thereby altering the habitat of other soil-dwelling organisms. Also, Cortez and Bouche (1992) showedconclusively that earthworms graze on live fine roots in a labo-ratory setting, and recent isotopic studies indicate the likelyimportance of this process in natural forest settings (Horowitzet al., 2009; Fahey et al., 2013a). Another important group ofinvasive soil invertebrate taxa is soil-dwelling weevils (Curcu-lionidae); which as larvae may consume significant amounts oflive fine roots of trees (Coyle et al., 2008) and when they emergeen masse constitute an import prey resource for other taxa(Maerz et al., 2005).

The objective of the present study was to quantify C flux fromfine tree roots through dominant taxa in the surface soil food webof the arbuscular mycorrhizal tree species sugar maple (Acer sac-charumMarsh.). We labeled sugar maple saplings with a large pulseof 13C and traced this pulse for several years through roots, rhizo-sphere soil and soil invertebrates, as well as a prominent soilvertebrate, the red-backed salamander (Plethodon cinereus). Wehypothesized high 13C enrichment of soil food web components,reflecting the prominence of the root channel in energy supply tosoil organisms. Further, we expected especially high enrichment ofinvasive soil invertebrates, earthworms andweevils that are knownto feed on live roots. Finally we hypothesized higher enrichment ofsaprotrophic and fungivorous soil invertebrates (Collembola, Dip-loda) than predatory taxa (Araneae, Chilopoda, Coleoptera, sala-manders), reflecting dilution of the isotope signal along the soilfood chain.

2. Methods

2.1. Study site

The research was conducted in a natural northern hardwoodforest in the Arnot Forest, central New York, USA (Fain et al., 1994).Annual precipitation at the site averages 90 cm, and average Juneand December temperatures are 22� and �4 �C, respectively. Soilsare acidic (pH 4.5e5.0) Dystrochrepts formed in glacial till locallyderived from Devonian shales. We chose a 90-yr-old forest domi-nated by sugar maple that had been selectively harvested in 2000leaving about 50 overstory trees per ha. This thinning released adense sugar maple understory which at the time of isotope labelingin 2006 averaged 2.5 m height.

2.2. Isotope labeling

The sugar maple saplings were labeled with 13C following pro-cedures detailed elsewhere (Horowitz et al., 2009; Fahey et al.,2011). Briefly, seven aluminum-frame chambers (2.5 m tall � 3 mdiameter) were positioned around nearly pure groves of sugarmaple saplings (ca. 10 stems/m2) in spring 2006. The soils in eachchamber were isolated by trenching around the chambers to 0.5 mdepth, lining with 6-mil polyethylene and back-filling with soil. Thesaplings were labeled with 13C by enclosing the chambers inpolyethylene sheeting and injecting 40% atom-enriched 13CO2 intothe chambers on 13 sunny mornings between 1 September and 20September 2006. This procedurewas successful in enriching foliageto d13C over 300& (Horowitz et al., 2009). All leaf litterfall wasremoved from the chambers in October and early November 2006e2008 and replaced with equivalent amounts of unlabeled leaf litterso that 13C isotope enrichment of the soil food web was derivedalmost entirely from belowground inputs (i.e., 13C translocated toroot systems).

2.3. Field sampling

Roots and rhizosphere soil were collected from 0 to 5 cm soildepth in four chambers in October 2006 (after leaf senescence),prior to leaf out in May 2007, and in late summer (15 August) 2007.For these samples root fanswere gently excavated by loosening soil,and fine roots (<1 mm diameter) and adhering rhizosphere soilwere manually separated (Phillips and Fahey, 2006), dried andstored for isotope analysis. Roots and soil also were sampled inearly June and mid-October 2008 and mid-October 2009 by col-lecting four soil cores (5 cm diameter, 0e10 cm depth) from eachchamber. Fine roots (<1 mm) were manually separated from soil,pooled within chambers, dried and stored for isotope analysis.Subsamples of soil from the October 2008 collection were used tomeasure microbial biomass C and 13C. Roots and soil for isotopenatural abundance measurements were collected in fall 2006 fromadjacent locations outside the chambers and processed as above. Inlate October of each year (2006e2008) we added unlabeled freshsugar maple leaf litter collected from adjacent forest in an amountdesigned to replace that removed from the chambers (300 g/m2).

Soil invertebrates were obtained from each chamber by col-lecting surface organic horizons in two 0.25m2 sub-plots on each offour dates: 4 June and 15 October 2008, 1 June and 15 October2009; also in June 2011 only two chambers were sampled as theothers had received an additional 13C labeling. These samples werereturned to the laboratory for invertebrate extractions the sameday. Earthworms were collected from the chambers on 26 April2007, 10e25 October 2007, 4 June 2008 and 15 May 2009. When-ever available the earthworms were obtained from beneath two30 � 15 cm cover boards in each chamber; however, in 2008 and2009 we applied a mustard suspension (Lawrence and Bowers,2002) to extract earthworms from the chambers. Finally, wecollected 2e4 salamanders (P. cinereus) from beneath cover boardsin each of the chambers on 9e11 October 2008 and from twochambers on 25 July 2009. Reference collections of earthworms,other soil invertebrates and salamanders were obtained from anearby study site (Fahey et al., 2013a) for isotope natural abundancemeasurements.

2.4. Laboratory processing

Fine root samples were finely ground and homogenized forisotope analysis using a ball mill. Soil samples were sieved to 2 mmand finely ground in a ball mill for isotope analysis. The isotopiccomposition of these samples was measured on a Finnegan IRMS at

-50

0

50

100

150

∂13 C

(0/ 00

)

Oct '06 May '07 Aug-Oct '07 Jun '08 Oct '08 Oct '09

Fine Roots

Rhizosphere soil

Fig. 1. Isotopic enrichment (d13C) of sugar maple fine roots (<1 mm) and rhizospheresoil collected from chambers following labeling in September 2006. Error bars indicatestandard errors (n ¼ 4) and baseline on y axis is natural abundance.

Fig. 2. Isotopic enrichment (d13C) in earthworms collected from chambers on fourdates following labeling in September 2006. Error bars indicate standard errors (n ¼ 5)and baseline on y axis is natural abundance.

K.J. Gilbert et al. / Soil Biology & Biochemistry 76 (2014) 45e52 47

the Cornell Stable Isotope Laboratory. Appropriate NIST standardswere used for normalization correction, instrument linearity andprecision purposes.

Microbial biomass C and 13C were measured on cold-stored,fresh samples within 48 h of collection (mid-October 2008) usingthe chloroform fumigation-incubation method (Jenkinson andPowlson, 1976). Soils were fumigated to kill and lyse microbialcells and fresh soil was used to inoculate samples with microbesthat used lysed cells as substrate. No correction for unfumigatedcontrols was applied as Horwath et al. (1996) indicated that littlebackground C is respired in fumigated incubations. Flushes of CO2over 10 day incubation at field moisture content were assumed tobe proportional to C in microbial biomass after applying a standard(0.45) proportionality constant. Trapped CO2 was stored in Exe-tainer gas sampling tubes, and CO2 was measured by gas chroma-tography. Samples were shipped to the Stable Isotope Laboratory atthe University of California-Davis to measure 13C on a EuropaIntegra IRMS.

Invertebrates were extracted from samples of surface organichorizons using Berlese funnels (MacFadyen, 1955). Samples wereextracted over a four-day period, and samples were stored in 70%ethanol until processed. By examining under a dissecting micro-scope, the invertebrates in each sample were sorted into thefollowing broad taxonomic groups on the basis of grossmorphology: Acari (mites), Araneae (spiders), Chilopoda (centi-pedes), Coleoptera (non-weevil beetles), Collembola (springtails),Curculionidae (weevils), Diplopoda (millipedes), Diptera (trueflies), Formicidae (ants), and Lepidoptera (moths). Selected sub-samples were stored for taxonomic identification by the CornellUniversity Insect Diagnostic Lab and the remainder of the sampleswas dried to constant mass and weighed. A subsample of eachtaxon from each date and chamber was ground in a ball mill andanalyzed for isotopes at the Cornell isotope laboratory as describedabove.

Earthworms collected from the chambers were sorted into twogroups for isotope analysis: saprophagous species (Lumbricusrubellus, L. terrestris and Lumbricus immatures) and geophagousspecies (Aporrectodea spp., Octolasion tyrtaeum). The anterior ends(15 segments) were dissected and cleaned of gut contents, groundand analyzed for isotopes, as above. Finally, salamander tissueisotope concentrations were measured by IRMS on ground samplesof whole salamanders at the University of Georgia stable isotopelaboratory.

2.5. Data analysis

We compared isotopic enrichment (d13C) across invertebratetaxa and collection date using two-way analysis of variance.Because of their opportunistic nature, collection dates for earth-worms sometimes differed from other invertebrates, and wecompared isotopic enrichment between saprophagous and geo-phagous earthworms with a separate ANOVA. We calculated thepool of excess 13C (i.e., above natural abundance) in various inver-tebrate taxa based upon estimated biomass and atom % enrichmentabove natural abundance for each collection date in 2008 and 2009.

3. Results

3.1. 13C enrichment of roots and rhizosphere soil

Fine roots collected from surface soil in the chambers werehighly enriched in 13C by October 2006, one month after labeling(Fig. 1). Isotopic enrichment of fine roots declined gradually overthe next three years. The most rapid decline of d13C in rootsoccurred during the 2nd growing season after labeling (2008), and

d13C values were close to natural abundance by October 2009(Fig. 1) illustrating rapid turnover of fine roots of sugar maple.Rhizosphere soil d13C increased to a peak in late-summer of 2007(one year after labeling; Fig. 1) and gradually declined thereafter(Fahey et al., 2013b).

3.2. 13C enrichment of soil heterotrophs

Earthworm tissues were more highly enriched in 13C than mostother soil invertebrate taxa (Fig. 2). Levels of d13C in saprophagousearthworms were much higher than for geophagous species, andvalues for both groups peaked in May 2008, 1.5 yr after labeling.Thereafter, d13C of earthworm tissues declined sharply, thoughvalues still remained highly enriched in May 2009 (5.73 � 8.86&)compared with natural abundance (about �25&; Fahey et al.,2013a). In October 2008, two years after labeling, microbialbiomass was highly enriched in 13C (�1 � 3&), much lower thanthe earthworms but comparable to many of the soil invertebrates.

The 13C label was detected in all eight taxonomic groups of ar-thropods on most of the sampling dates through October 2009(Fig. 3). The principal taxa identified for each of these broad groupsare presented in Supplementary Table 1. Differences in d13C acrossdates were highly significant (F ¼ 5.85, p < 0.001). Differencesacross taxa also were significant (F ¼ 6.95, p < 0.001), but post-hoccomparisons indicated that only one taxonomic group (ants; For-micidae) differed significantly from the others: the ants exhibitedby far the highest d13C enrichment while other taxa were not sta-tistically different from one another. High variation across cham-bers within dates was observed for most taxa. In general, d13Cenrichment in most taxa peaked in 2008 and declined thereafter(Fig. 3); however, values in Formicidae remained highly enrichedthrough 2009, returning to natural abundance levels in 2011.

No earthworms were observed in two of the seven chambers,and well-developed surface organic horizons in these chamberscontrasted with the mull surface soils in the other chambers whereearthworms were observed and collected. No significant

Fig. 3. Isotopic enrichment (d13C) of soil microbial biomass and eight groups ofarthropod taxa collected from litter layers of chambers following labeling in September2006. Error bars indicate standard errors (n ¼ 7) and baseline on y axis is naturalabundance.

Fig. 4. Excess 13C pool in eight groups of litter arthropod taxa collected from chamberson four dates following labeling in September 2006. Error bars indicate standard errors(n ¼ 7).

K.J. Gilbert et al. / Soil Biology & Biochemistry 76 (2014) 45e5248

differences were observed in d13C for any of the invertebrate taxabetween the no worm and worm-invaded chambers; in fact,average isotope enrichment values were generally very similarbetween these two sets of labeling chambers (data not shown).

Tissues of the woodland salamander, P. cinereus, collected fromthe chambers in October 2008, were significantly enriched in 13C(d13C ¼ �15.9 � 1.77) compared with natural abundance values(13C ¼ �23.14 � 0.15). Again, there were no differences in isotopeenrichment of salamanders between the chambers with earth-worms (d13C ¼ �15.73 � 2.12) and those lacking earthworms(�16.53 � 3.64).

In summary, during the second growing season after isotopelabeling of tree root systems by far the highest d13C enrichment wasobserved for fine roots, earthworms and ants. Isotopic enrichmentof microbial biomass and two groups of litter-dwelling arthropodseweevils and mites ewas intermediate. Lower d13C enrichment ofrhizosphere soil reflected in part the much larger pool size intowhich the tracer was incorporated (see Section 3.3), while lowerenrichment of various predators and saprotroph/fungivores pre-sumably resulted fromproportionately higher feeding on unlabeledsubstrates.

3.3. 13C pools in fine roots, rhizosphere soil and heterotrophs

We calculated the excess 13C pools in fine roots, rhizosphere soil,microbial biomass and arthropods on the basis of biomass esti-mates and the atom % enrichment above natural abundance values(excess 13C). Fine root excess 13C averaged about 300 mg 13C/m2

during the first 1.5 yr following isotope labeling. The rhizospheresoil 13C pool gradually increased from <50 mg 13C/m2 in late fallafter labeling, to 242 mg 13C/m2 the following May and a peak of530 mg 13C/m2 one year after labeling (Fahey et al., 2013b). Incomparison, the excess 13C pool in microbial biomass two yearsafter labeling was estimated at 3.9 mg 13C/m2. Excess 13C pools inlitter arthropods collected in the chambers were much lower thanfor microbes, ranging from 0.05 to 0.09 mg 13C/m2 during thesecond growing season after labeling (Fig. 4). The total excess 13Cpool in litter arthropods declined exponentially from May 2008onward. However, temporal patterns in arthropod excess 13C poolsdiffered among taxa, reflecting both patterns in arthropod biomassand d13C enrichment (Fig. 4). Excess 13C pools declined throughtime in Araneae, Chilopoda and non-weevil Coleoptera, whereasdistinct later peaks in 13C pools were observed in Diplopoda andHymenoptera in October 2008 and in Collembola and Curculioni-dae in May 2009.

4. Discussion

Assimilation of 13CO2 by the trees in September 2006 greatlyenriched foliage, and significant amounts of 13C were translocatedto the roots (Horowitz et al., 2009). Because we removed the 13Clabeled aboveground litterfall (replaced with unlabeled litter),nearly all of the 13C supplied to the soil was derived from the treeroot systems. We observed peak concentrations of 13C in fine rootsin late fall 2006 and spring 2007 (Fig. 1), probably largely in theform of stored carbohydrates because little root growth occursoutside the growing season (Tierney et al., 2003). By late summerand fall 2007 (one year after labeling) these labile pools wereprobably largely depleted, and high enrichment of structural tis-sues of fine roots was observed (Fahey et al., 2012). Over the nexttwo years this signal was diluted by turnover of 13C-enriched rootsand new root growth supplied from unenriched assimilates (Fig. 1).Rhizosphere soil and microbial biomass also were enriched in 13C(Figs. 1 and 3) but to a much lower enrichment than the fine roots;however, because of the greater mass of rhizosphere soil, itconstituted a somewhat larger excess 13C pool than the fine roots(Fahey et al., 2012).

The high 13C enrichment of all the components of the litter andsoil food web when the 13C was supplied via the root channelstrongly supports the importance of root-derived C in poweringheterotrophic activity in temperate forest soils. Roots are known tosupply a high proportion of C inputs to forest soil organic matter(Kuzyakov and Domanski, 2000; Matamala et al., 2003), and iso-topic evidence suggests their predominance in energy supply tomany soil heterotrophs in both forests (Pollierer et al., 2007, 2009)and agroecosystems (Kramer et al., 2012). The reasons for thepredominance of the root over the litter channel in powering thesoil food web are not clear, but biochemical recalcitrance of leaflitter (high C:N ratio, lignin concentration) and extremes of thesurface litter environment (i.e. drying, heating, freezing) couldcontribute. The importance of gaining a better understanding ofthis problem is emphasized by probable effects of global changephenomena in altering the proportional allocation of C to above-ground vs belowground plant tissues (Hyvönen et al., 2007).

The importance of various basal resources e fine roots, mycor-rhizal and saprotrophic fungi, bacteria, soil organic matter e inenergy supply to soil food webs is poorly understood. Pollierer et al.(2012) demonstrated that ectomycorrhizal fungi were a predomi-nant source of C input to a beech-dominated forest soil food web.Our observations suggest that the root channel also predominates

K.J. Gilbert et al. / Soil Biology & Biochemistry 76 (2014) 45e52 49

in an arbuscular mycorrhizal tree species, but we cannot conclu-sively evaluate the relative importance of AM fungi vs other root-derived basal resources (live roots, root exudates, bacteria, sapro-trophic fungi). High enrichment of the rapid turnover microbialbiomass pool indicates its likely significance as a basal resource,supporting observations for a beech forest (Pollierer et al., 2012).Moreover, as detailed below, high enrichment of invasive earth-worms and weevils suggests a significant role of invertebrategrazing on living roots.

4.1. Earthworms

A most striking observation in the soil food web was theexceptionally high 13C enrichment in earthworms (Fig. 2). Europeanearthworms (Lumbricidae) have colonized widely in North Amer-ican forests previously devoid of large, soil-mixing earthworms(James, 1995). These earthworms can profoundly alter soil proper-ties by consuming surface litter and mixing with mineral soil(Alban and Berry, 1994; Bohlen et al., 2004). Although it hasgenerally been observed that earthworms derive most of theirnutrition from soil microbes, soil organic matter and plant detritus(Curry and Schmidt, 2007) including dead roots (Lee, 1985), there isclear evidence that they also consume live roots, root hairs andmycorrhizal fungi (Baylis et al., 1986; Cortez and Bouche, 1992;Gunn and Cherrett, 1993). In a previous report from the presentstudy, Horowitz et al. (2009) noted that earthworms, collected fromthe chambers only one month after completion of isotope labeling,already were highly enriched in 13C (d13C ¼ 10&e33&). For thoseanalyses whole earthworms, including gut contents, wereanalyzed, and the high enrichment probably reflected root frag-ments in soil within earthworm guts. However, by the followingApril we observed still higher d13C enrichment in earthworm tis-sues (gut contents removed, Fig. 2) indicating a high degree ofassimilation of root-associated carbon.

The sources of root-derived C in earthworm tissues couldinclude root tissues, mycorrhizal hyphae and rhizosphere C.Given the brief transit time through the earthworm gut (severalhours; Edwards and Bohlen, 1996), digestion of root structuraltissues probably would be minimal; however, earthworms couldassimilate C from root tissues that had previously passed throughtheir gut, was partially decayed by soil microbes and then rein-gested. Also, root carbohydrate storage pools could be morereadily assimilated during gut passage. Notably, the timing of thefirst earthworm collections preceded most fine root growthwhich commences in late April and early May in sugar maple(Tierney et al., 2003). Rapid flux of 13C from roots to arbuscularmycorrhizal hyphae and soil saprotrophs (De Deyn et al., 2011;Bradford et al., 2012) also could have supplied significantamounts of root-derived 13C to the earthworms, but given thetiming and high level of enrichment (Fig. 2), root detritus waslikely a minor source.

By October 2007, 13C enrichment in the tissues of saprophagousearthworms exceeded that in any of the presumed 13C source poolsthat were measured (live and dead fine roots, rhizosphere soil,microbial biomass). Assuming minimal biochemical fractionationof C isotopes in their physiological processes, this observationsuggests that earthworms in these soils preferentially assimilated Cfrom sub-fractions of the measured pools that were more highlyenriched (e.g., labile C in fine roots). Undoubtedly, 13C enrichmentvaried among fine root sizes, age classes and tissues, includingmycorrhizal hyphae. Fahey et al. (2012) reported that cell walls oflower order fine roots (order 1e3) collected from these chamberswere more highly enriched in 13C than higher order roots (order 4e5), but measurements of other tissues or arbuscular mycorrhizalhyphae were not conducted.

Geophagous earthworms (Aporrectodea spp., O. tyrtaeum) thatfeed belowground and consume a higher proportion of mineral soilwere significantly less enriched in 13C than the saprophagousspecies that feed on surface litter and on a higher proportion oforganic material. Also, 13C enrichment in geophagous earthwormswas delayed in comparison to saprophagous species, and theirenrichment declined more gradually following the peak in the 2ndyear after labeling (Fig. 2). A similar pattern was observed in arelated study in which 13C labeled leaf litter was the source ofearthworm isotopic enrichment; Fahey et al. (2013a) attributed thelower and delayed enrichment of geophagous earthworms in thatstudy to gradual mixing of labeled litter into mineral soil bysaprophagous earthworms, making it available to the geophagousspecies. It is possible that a similar process occurred in this study:perhaps enrichment of geophagous species was delayed due togradual mixing of 13C label into fractions fed upon by these species.Later the remaining soil 13C may have been in fractions moreavailable to geophages than to saprophages due to differences infeeding strategies. Metabolic differences and differential incorpo-ration of C into different earthworm tissues may also have been inplay as young earthworms are likely to incorporate C into tissuesthat are retained whereas more mature earthworms are likely torespire incorporated C or shed it in labile fractions such as mucous(Whalen and Janzen, 2002).

The levels of 13C enrichment observed in earthworms in thechambers exceeded the peak levels in the leaf litter study describedabove even though the leaf litter had much higher d13C enrichment(300&; Fahey et al., 2013a) than peak values in the fine roots(100&, Fig.1); this strongly supports the importance of fine roots tothe nutrition of these invasive earthworms. The pattern of changingd13C in earthworm tissues mirrored the pattern in live fine roots,further supporting the argument that roots and possibly mycor-rhizal fungi and rhizosphere C were a principal source of C in theearthworms (Figs. 1 and 2). Fisk et al. (2004) observed that fine rootbiomass was significantly lower in earthworm-invaded sugar ma-ple forest than adjacent earthworm-free sites, reinforcing thesuggestion that earthworms consume significant amounts of fineroots. Also, Lawrence et al., (2003) observed lower AM colonizationof sugar maple roots in earthworm invaded than adjacent no-earthworm sites, possibly because of earthworm grazing.

To avoid excessive soil disturbance, we did not attempt toquantify earthworm biomass in the chambers; however, mea-surements in nearby sugar maple forests provide a basis for esti-mating the pool size of excess 13C in the earthworm communities.Fahey et al. (2013c) reported 2-yr average values for saprophagousspecies (4.95 g/m2) and geophagous species (1.03 g/m2) for threenearby maple forest sites. Applying the peak values of earthworm13C enrichment in May 2008 (Fig. 2) and natural abundance (Faheyet al., 2013a), we estimated the excess 13C pool for saprophagousearthworms (3.65 mg 13C/m2) and for geophagous species(0.43 mg 13C/m2) for a total of 4.08 mg 13C/m2. As noted, the peakpool of excess 13C in fine roots was 300 mg 13C/m2; hence, the peakpool of excess 13C in earthworm biomass was about 1.4% as large asthat in fine roots. For illustrative purposes, if we assume that allearthworm excess 13C was derived from fine roots, and an assimi-lation efficiency of 10%, then earthworms might be consuming asmuch as 14% of fine root biomass. Clearly this potentially largepathway of energy flow and nutrient cycling in temperate forestscolonized by invasive earthworms deserves further study.

4.2. Other litter invertebrates

Another invasive invertebrate group that has been suggested toinfluence root dynamics in northern hardwood forests is root-feeding weevils (Curculionidae). For example, Coyle et al. (2008)

K.J. Gilbert et al. / Soil Biology & Biochemistry 76 (2014) 45e5250

estimated that as much as 15% of fine root production might beconsumed by larval weevils. We observed relatively high d13Cenrichment in adult weevils, mostly Barypeithes pellucidus,collected from the litter layers (Fig. 3). Based on the biomass ofthese weevils recovered from litter, we calculate only a very smallpool of excess 13C in their tissues (Fig. 4); however, much of the soil-dwelling weevil biomass probably would not be recovered fromlitter layers. Assuming that the excess 13C values in adult weevilsapplies to larvae, and using the 10-year average larval populationsize of root feeding weevils measured by Coyle et al. (2008) innorthern hardwood forest (3.6 g FW/m2), we estimate about0.2 mg excess 13C/m2. At 5% efficiency of assimilation, this com-prises 1.3% consumption of fine roots, a value substantially lowerthan the estimate of Coyle et al. (2008). However, our estimatewould increase to the extent that 13C is lost during maturation ofweevils from larvae to adult, and our observations reinforce thelikely significance of invasive root-feeding weevils in fine root dy-namics of northern hardwood forests.

A third group of litter invertebrates that may be influencing fineroot dynamics in sugar maple is Formicidae which exhibited by farthe highest d13C among the litter arthropods collected (Fig. 3). Themost abundant taxon identified in these samples was Lasius spp., aground-nesting genus of ants that is known to tend root-feedingaphids and coccids (Wheeler, 1911). This high enrichment is sug-gestive of an important role of these phloem-sucking invertebratesin root C dynamics. It is also possible that the high 13C abundance innon-weevil Coleoptera in June 2008 (Fig. 4) was connected withthis C flow channel; Brachyacantha quadripunctata is a coccinellidthat has been observed in ant nests and feeding on coccid eggs(Wheeler, 1911). This food web channel also deserves furtherdetailed study.

In sum, the foregoing observations provide additional evidencefor a significant role of feeding on fine roots and root-derived C bysoil invertebrates (Wells et al., 2002; Stevens and Jones, 2006;Hunter, 2008). Our results suggest that this key process may beparticularly sensitive to exotic species invasions (Annelida, Curcu-lionidae). Given the prominent role of fine roots as regulators ofecosystem response to global environmental change (Phillips et al.,2011), further analysis of the structure and function of root-feedinginvertebrate communities is needed.

Patterns of 13C enrichment in other litter arthropods providedfurther insights about the role of the root channel in supplyingenergy to the soil food web. Fungal-feeding Collembola weremoderately enriched in 13C during the 2nd growing season afterlabeling (d13C ¼ �5 to 0&; Fig. 3). This suggests either that theymove between mineral soil and the litter layer or that fungal Cderived from roots (via mycorrhizal or saprophytic fungi) wastransported into the litter. It is well known that soil fungi transportC to decaying litter (Tllalka et al., 2003) possibly contributing to thisresult. Notably, the pool of excess 13C in Collembola increased insummer 2009 even as tissue enrichment declined (compare Figs. 3and 4), suggesting they were feeding on fungi involved in fine rootdecomposition. Detrital-feeding Diplopoda were moderatelyenriched in 13C, but they exhibited the largest isotope pool excess(Fig. 4), suggesting that they obtain much of their nutrition fromfine root detritus.

We also observed unexpectedly high 13C enrichment in severalpredatory invertebrate groups collected in the litter layer e carabidbeetles, spiders and Chilopoda. Moreover, in June 2008 beetles andChilopoda had high 13C pool excess, exceeding that in most of theirpotential prey (Fig. 4). What were they eating that fed in the soil onroot-derived C? Two possible sources are mites and the highly-enriched ants. They might also be feeding on highly-enrichedearthworms and Diplopoda. Carabid beetles are generalist preda-tors known to consume earthworms in agricultural systems,

particularly during periods of high activity or reduced resources(Symondson et al., 2000). However, Schmidt et al. (2004) suggestedthat earthworms, as well as diplopods, may constitute an energyflow dead end in the soil faunal food web because they are too largefor other soil fauna to consume.

4.3. Salamanders

The woodland salamander, P. cinereus, is the most abundantvertebrate in many northern hardwood forests and an importanttop predator in the soil food web (Burton and Likens, 1975). Sala-manders collected in the chambers were highly enriched in 13Cthough less so than invertebrate predators. Salamanders seldomfeed within the soil (Fraser, 1976a). Foraging is generally con-strained to the surface of the forest floor while under moist coverobjects or proximate to rain during cooler seasons (Fraser, 1976a;Maerz et al., 2005). As a result, salamander foraging opportunitiesand diet are stochastic and seasonally variable. Maerz et al. (2005)demonstrated that P. cinereus predation on earthworms isconcentrated on rainy nights during the cooler seasons, and pre-dation on weevils is concentrated during the spring. P. cinereus arealso restricted by gape limitation to preying predominantly onsmall and medium sized earthworms, and we did not distinguishthe enrichment of large versus small earthworms. Ants are a stablecomponent of salamander diets throughout the year (Maerz et al.,2005). Nonetheless, Maerz et al. (2005) showed that nonnativeearthworms, nonnative weevils (B. pellucidis specifically), and for-micid ants are collectively the three taxa contributing most toP. cinereus diets in our study region accounting on average for morethan 60% of all prey biomass consumed by salamanders across theactive season. Fraser (1976a,b) also reported that ants and earth-worms are the dominant contributors to P. cinereus (and Plethodonhoffmani) prey biomass. Collembola, mites, and spiders are oftenthe most common prey in P. cinereus diets, though they do notcontribute considerably to the biomass of salamander prey exceptepisodically (Burton, 1976; Fraser, 1976a,b; Maglia, 1996; Maerzet al., 2005). Small soil taxa are more important in the diets ofsmaller juvenile salamanders (Maglia, 1996). Adult carabid beetlesoccur occasionally in P. cinereus diets, though on average are notimportant to salamander prey biomass (Maerz et al., 2005); how-ever, coleoptera larvae are one of the more important prey inP. cinereus diets (Burton, 1976; Fraser, 1976a, b; Maglia, 1996; Maerzet al., 2005).

The observations that all of the taxa important in P. cinereusdiets derive a significant portion of C from roots, and that 13Cenrichment in salamanders from the chambers (d13C ¼ �15.9&)was much higher than for those collected from the plots withlabeled leaf litter (Section 4.1) all support the predominance ofroots over leaf litter in powering the soil food web in forest soilswith invasive earthworms or soil weevils. Moreover, Maerz et al.(unpublished data) also found no evidence that the presence ofearthworms at our study site was associated with reduced sala-mander abundance (contrary to Maerz et al., 2009) despite dra-matic reductions in surface arthropod abundances (Burke et al.,2011) and evidence that the presence of earthworms reduced thetotal amount of leaf litter C transferred to salamanders over a fouryear period. These results suggest that despite reducing surface leaflitter and the importance of leaf litter C in soil food webs, earth-worms and other soil fauna may be supporting salamander andother consumer populations through the transfer of root-derivedresources.

In conclusion, the food web in the litter and soil of these sugarmaple mesocosms clearly derived a large amount of energy fromthe root channel, including live roots and their exudates andmycorrhizal associates. Isotopic enrichment suggested that invasive

K.J. Gilbert et al. / Soil Biology & Biochemistry 76 (2014) 45e52 51

earthworms consumed significant amounts of living roots andmycorrhizae, but the other invertebrates also derived much of theirenergy from the root channel. The high enrichment of earthwormtissue suggests that invasive earthworms target root-derived C intheir feeding activities and challenges the view that saprophagousearthworms depend mainly on surface litter as a C source.

Acknowledgments

For their assistance in the field and laboratory we thank T.Feldpausch, M. Fisk, A. Heinz, L. Martel, L. Stoschek, and K. Cecala.This research was supported by the National Science Foundation(NSF DEB-0542065), Ecosystem Studies Program.

Appendix A. Supplementary data

Supplementary data related to this article can be found at http://dx.doi.org/10.1016/j.soilbio.2014.05.005.

References

Alban, D.H., Berry, E.C., 1994. Effects of earthworm invasion on morphology, carbon,and nitrogen of a forest soil. Applied Soil Ecology 1, 243e249.

Baylis, J.P., Cherrett, J.M., Ford, J.B., 1986. A survey of the invertebrates feeding onliving clover roots (Trifolium repens L.) using 32P as a radiotracer. Pedobiologia29, 201e208.

Bohlen, P.J., Scheu, S., Hale, C.M., McLean, M.A., Migge, S., Groffman, P.M.,Parkinson, D., 2004. Non-native invasive earthworms as agents of change innorth temperate forests. Frontiers in Ecology and the Environment 2, 427e435.

Bonkowski, M., Villenave, C., Griffiths, B., 2009. Rhizosphere fauna: the functionaland structural diversity of intimate interactions of soil fauna with plant roots.Plant and Soil 321, 213e233. http://dx.doi.org/10.1007/s11104-009-0013-2.

Bradford, M.A., Strickland, M.S., DeVore, J.L., Maerz, J.C., 2012. Root carbon flow froman invasive plant to belowground foodwebs. Plant and Soil 359 (1e2), 233e244.

Brussaard, L., Pulleman, M.M., Ouedraogo, E., Mando, A., Six, J., 2007. Soil fauna andsoil function in the fabric of the food web. Pedobiologia 50 (6), 447e462.

Burke, J.L., Maerz, J.C., Milanovich, J.R., Fisk, M.C., Gandhi, K.J.K., 2011. Invasion byexotic earthworms alters biodiversity and communities of litter- and soil-dwelling oribatid mites. Diversity 3 (1), 155e175. http://dx.doi.org/10.3390/d3010155.

Burton, T.M., 1976. An analysis of the feeding ecology of the salamanders (Amphibia,Urodela) of the Hubbard Brook experimental forest, new hampshire. Journal ofHerpetology 10 (3), 187e204.

Burton, T.M., Likens, G.E., 1975. Salamander populations and biomass in the Hub-bard Brook experimental Forest, New Hampshire. Copeia, 541e546.

Cortez, J., Bouche, M.B., 1992. Do earthworms eat living roots? Soil Biology andBiochemistry 24, 913e915. http://dx.doi.org/10.1016/0038-0717(92)90014-O.

Coyle, D.R., Mattson, W.J., Raffa, K.F., 2008. Invasive root feeding insects in naturalforest ecosystems of North America. In: Johnson, S.N., Murray, P.J. (Eds.), RootFeeders: an Ecosystem Perspective. CABI.org, pp. 134e151.

Curry, J.P., Schmidt, O., 2007. The feeding ecology of earthworms e a review.Pedobiologia 50 (6), 463e477.

De Deyn, G.B., Quirk, H., Oakley, S., Ostle, N., Bardgett, R.D., 2011. Rapid transfer ofphotosynthetic carbon through the plant-soil system in differently managedspecies-rich grasslands. Biogeosciences 8, 1131e1139. http://dx.doi.org/10.5194/bg-8-1131-2011.

Dempsey, M.A., Fisk, M.C., Fahey, T.J., 2011. Earthworms increase the ratio of bacteriato fungi in northern hardwood forest soils, primarily by eliminating the organichorizon. Soil Biology and Biochemistry 43 (10), 2135e2141.

Dempsey, M.A., Fisk, M.C., Yavitt, J.B., Fahey, T.J., Balser, T.C., 2013. Exotic earth-worms alter soil microbial community composition and function. Soil Biologyand Biochemistry 67, 263e270.

Edwards, C.A., Bohlen, P.J., 1996. Biology and Ecology of Earthworms, vol. 3.Chapman & Hall, London, UK.

Eisenhauer, N., Schlaghamersky, J., Reich, P.B., Frelich, L.E., 2011. The wave towards anew steady state: effects of earthworm invasion on soil microbial functions.Biological Invasions 13, 2191e2196. http://dx.doi.org/10.1007/s10530-011-0043-4.

Fahey, T.J., Yavitt, J.B., Sherman, R.E., Maerz, J.C., Groffman, P.M., Fisk, M.C.,Bohlen, P.J., 2013a. Earthworms, litter and soil carbon in a northern hardwoodforest. Biogeochemistry 114, 269e280.

Fahey, T.J., Yavitt, J.B., Sherman, R.E., Groffman, P.M., Wang, G., 2013b. Partitioning ofbelowground C in young sugar maple forest. Plant and Soil 367, 379e389.

Fahey, T.J., Yavitt, J.B., Sherman, R.E., Maerz, J.C., Groffman, P.M., Fisk, M.C., Bohlen, P.,2013c. Earthworm effects on the conversion of litter C and N into soil organicmatter in a sugar maple forest. Ecological Applications 23 (5), 1185e1201.

Fahey, T.J., Jacobs, K.R., Sherman, R.E., 2012. Fine root turnover in sugar mapleestimated by 13C isotope dilution. Canadian Journal of Forest Research 42 (10),1792e1795. http://dx.doi.org/10.1139/x2012-128.

Fahey, T.J., Yavitt, J.B., Sherman, R.E., Groffman, P.M., Fisk, M.C., Maerz, J.C., 2011.Transport of carbon and nitrogen between litter and soil organic matter in anorthern hardwood forest. Ecosystems 14, 326e340. http://dx.doi.org/10.1007/s10021-011-9414-1.

Fain, J.J., Volk, T.A., Fahey, T.J., 1994. Fifty years of change in an upland forest insouth-central New York: general patterns. Bulletin of the Torrey Botanical Club121 (2), 130e139.

Fisk, M.C., Fahey, T.J., Groffman, P.M., Bohlen, P.J., 2004. Earthworm invasion, fine-root distributions, and soil respiration in north temperate forests. Ecosystems 7,55e62.

Fraser, D.F., 1976a. Empirical evaluation of the hypothesis of food competition insalamanders of the genus plethodon. Ecology 57, 459e471. http://dx.doi.org/10.2307/1936431.

Fraser, D.F., 1976b. Coexistence of salamanders in the genus plethodon: a variationof the Santa Rosalia theme. Ecology 57 (2), 238e251.

Greiner, H., Kashian, D., Tiegs, S., 2012. Impacts of invasive Asian Amynthas hil-gendorfi and European Lumbricus rubellus earthworms in a North Americantemperate deciduous forest. Biological Invasions 14, 2017e2027.

Gunn, A., Cherrett, J.M., 1993. The exploitation of food resources by soil meso-invertebrates and macro-invertebrates. Pedobiologia 37 (5), 303e320.

Hale, C.M., Frelich, L.E., Reich, P.B., 2005. Effects of European earthworm invasion onsoil characteristics in northern hardwood forests of Minnesota, USA. Ecosys-tems 8, 911e927.

Horowitz, M.E., Fahey, T.J., Yavitt, J.B., Feldpausch, T., Sherman, R.E., 2009. Patterns oflate-season photosynthate movement in sugar maple (Acer saccharum Marsh)saplings. Canadian Journal of Forest Research 39, 2294e2298.

Horwath, W.R., Paul, E.A., Harris, D., Norton, J., Jagger, L., Horton, K.A., 1996. Defininga realistic control for the chloroform-fumigation incubation method usingmicroscopic counting and 14C-substrates. Canadian Journal of Soil Science 76,459e467.

Hunter, M.D., 2008. Root herbivory in forest ecosystems. In: Johnson, S.N.,Murray, P.J. (Eds.), Root Feeders: an Ecosystem Perspective, pp. 68e95. Ch. 5.CABI.org.

Hyvönen, R., et al., 2007. The likely impact of elevated [CO2], nitrogen deposition,increased temperature and management on carbon sequestration in temperateand boreal forest ecosystems: a literature review. New Phytologist 173 (3), 463e480.

James, S.W., 1995. Systematics, biogeography, and ecology of Neartic earthwormsfrom eastern, central, southern, and southwestern United States. In:Hendrix, P.F. (Ed.), Earthworm Ecology and Biogeography in North America.Lewis, Boca Raton (FL), pp. 29e52.

Jenkinson, D.S., Powlson, D.S., 1976. The effects of biocidal treatments on meta-bolism in soildI. Fumigation with chloroform. Soil Biology and Biochemistry 8(3), 167e177.

Kramer, S., Marhan, S., Ruess, L., Armbruster, W., Butenschoen, O., Haslwimmer, H.,Kuzyakov, Y., Pausch, J., Scheuneman, N., Schoen, J., Schmalwasser, A.,Totsche, K., Walker, F., Scheu, S., Kandeler, E., 2012. Carbon flow into microbialand fungal biomass as a basis for the belowground food web of agroecosystems.Pedobiologia 55 (2), 111e119.

Kuzyakov, Y., Domanski, G., 2000. Carbon input by plants into the soil: review.Journal of Plant Nutrition and Soil Science 163, 421e431.

Lawrence, A.P., Bowers, M.A., 2002. A test of the ‘hot’ mustard extraction method ofsampling earthworms. Soil Biology and Biochemistry 34 (4), 549e552.

Lawrence, B., Fisk, M.C., Fahey, T.J., Suarez, E.R., 2003. Influence of non-native earth-worms on mycorrhizal colonization of sugar maple (Acer saccharum). New Phy-tologist 157, 145e153. http://dx.doi.org/10.1046/j.1469-8137.2003.00649.x.

Lee, K.E., 1985. Earthworms: Their Ecology and Relationships with Soils and LandUse. Academic Press, Sydney.

Loss, S.R., Blair, R.B., 2011. Reduced density and nest survival of ground-nestingsongbirds relative to earthworm invasions in northern hardwood forests.Conservation Biology 25, 983e992. http://dx.doi.org/10.1111/j.1523-1739.2011.01719.x.

Loss, S.R., Niemi, G.J., Blair, R.B., 2012. Invasions of non-native earthworms related topopulation declines of ground-nesting songbirds across a regional extent innorthern hardwood forests of North America. Landscape Ecology 27, 683e696.

MacFadyen, A., 1955. Methods for extracting soil arthropods. Soil Zoology. Butter-worth’s Scientific Publications, London.

Maerz, J.C., Nuzzo, V., Blossey, B., 2009. Declines in woodland salamander abun-dance associated with nonnative plant and earthworm invasions. ConservationBiology 23, 975e981.

Maerz, J.C., Karuzas, J.M., Madison, D.M., Blossey, B., 2005. Introduced invertebrates areimportant prey for a generalist predator. Diversity and Distributions 11, 83e90.

Maglia, A.M., 1996. Ontogeny and feeding of the red-backed salamander, Plethodoncinereus. Copeia 1996, 576e586.

Matamala, R., Gonzalez-Meler, M., Jastrow, J., Norby, R., Schlesinger, W., 2003. Im-pacts of fine root turnover on forest NPP and soil C sequestration potential.Science 302, 1385e1387.

Phillips, R.P., Fahey, T.J., 2006. Tree species and mycorrhizal associations influencethe magnitude of rhizosphere effects. Ecology 87 (5), 1302e1313.

Phillips, R.P., Finzi, A.F., Bernhardt, E.S., 2011. Enhanced root exudation inducesmicrobial feedbacks to N cycling in a pine forest under long-term CO2 fumi-gation. Ecology Letters 14, 187e194.

K.J. Gilbert et al. / Soil Biology & Biochemistry 76 (2014) 45e5252

Pollierer, M.M., Dyckmans, J., Scheu, S., Haubert, D., 2012. Carbon flux through fungiand bacteria into the forest soil animal food web as indicated by compound-specific 13C fatty acid analysis. Functional Ecology 26, 978e990. http://dx.doi.org/10.1111/j.1365-2435.2012.02005.x.

Pollierer, M.M., Langel, R., Scheu, S., Maraun, M., 2009. Compartmentalization of thesoil animal food web as indicated by dual analysis of stable isotope ratios(15N/14N and 13C/12C). Soil Biology and Biochemistry 41 (6), 1221e1226.

Pollierer, M.M., Langel, R., Körner, C., Maraun, M., Scheu, S., 2007. The under-estimated importance of belowground carbon input for forest soil animal foodwebs. Ecology Letters 10 (8), 729e736.

Ruess, L., Chamberlain, P.M., 2010. The fat that matters: soil food web analysis usingfatty acids and their carbon stable isotope signature. Soil Biology andBiochemistry 42, 1898e1910.

Scheu, S., 2002. The soil food web: structure and perspectives. European Journal ofSoil Biology 38, 11e20.

Schmidt, O., Curry, J.P., Dyckmans, J., Rota, E., Scrimgeour, C.M., 2004. Dual stableisotope analysis (d13C and d15N) of soil invertebrates and their food sources.Pedobiologia 48, 171e180.

Stevens, G.N., Jones, R.H., 2006. Patterns in soil fertility and root herbivory interactto influence fine-root dynamics. Ecology 87, 616e624.

Symondson, W.O.C., Glen, D.M., Erickson, M.L., Liddell, J.E., Langdon, C.J., 2000. Doearthworms help to sustain the slug predator Pterostichus melanarius

(Coleoptera: Carabidae) within crops? Investigations using monoclonal anti-bodies. Molecular Ecology 9, 1279e1292.

Tierney, G.L., Fahey, T.J., Groffman, P.M., Hardy, J.P., Fitzhugh, R.D., Driscoll, C.T.,Yavitt, J.B., 2003. Environmental control of fine root dynamics in a northernhardwood forest. Global Change Biology 9, 670e679.

Tllalka, M., Hensman, D., Darrah, P.R., Watkinson, S.C., Fricker, M.D., 2003. Non-circadian oscillations in amino acid transport have complementary profiles inassimilatory and foraging hyphae of Phanerochaete velutina. New Phytologist158 (2), 325e335.

Wardle, D.A., Bardgett, R.D., Klironomos, J.N., Setälä, H., van der Putten, W.H.,Wall, D.H., 2004. Ecological linkages between aboveground and belowgroundbiota. Science 304 (5677), 1629e1633. http://dx.doi.org/10.1126/science.1094875.

Wells, C.E., Glenn, D.M., Eissenstat, D.M., 2002. Soil insects alter fine root demog-raphy in peach (Prunus persica). Plant Cell and Environment 25, 431e439.

Whalen, J.K., Janzen, H.H., 2002. Labeling earthworms uniformly with 13C and 15N:implications for monitoring nutrient fluxes. Soil Biology and Biochemistry 34,1913e1918.

Wheeler, W.M., 1911. An ant-nest coccinellid (Brachycantha quadripunctata Mels.).Journal of the New York Entomological Society 19, 169e174.