lable at ScienceDirect

Soil Biology & Biochemistry 74 (2014) 1e10

Contents lists avai

Soil Biology & Biochemistry

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

Can earthworm-secreted calcium carbonate immobilise Zn incontaminated soils?

L. Brinza a,*, Paul F. Schofield b, J. Fred W. Mosselmans a, Erica Donner c, Enzo Lombi c,David Paterson d, Mark E. Hodson e

aDiamond Light Source Ltd., Harwell Science and Innovation Campus, Chilton, Didcot OX11 0DE, UKbMineral and Planetary Sciences, Department of Earth Sciences, Natural History Museum, Cromwell Road, London SW7 5BD, UKcCentre for Environmental Risk Assessment and Remediation, University of South Australia, Mawson Lakes Campus, Mawson Lakes, SA 5095, AustraliadX-ray Microanalysis Beamline, Australian Synchrotron Clayton, Victoria 3168, Australiae Environment Department, University of York, Heslington, York YO10 5DD, UK

a r t i c l e i n f o

Article history:Received 23 October 2013Received in revised form28 December 2013Accepted 12 January 2014Available online 24 January 2014

Keywords:Lumbricus terrestrisCalciteMetalAdsorptionCoprecipitationXASSynchrotron

* Corresponding author.E-mail address: loredana.brinza@diamond.ac.uk (L

0038-0717/$ e see front matter � 2014 Published byhttp://dx.doi.org/10.1016/j.soilbio.2014.01.012

a b s t r a c t

We investigated the interaction of soil Zn with calcium carbonate granules secreted by the earthwormLumbricus terrestris. Earthworms were kept in agricultural soil amended with ZnSO4 to give Zn con-centrations up to 750 mg kg�1 and in two former mine site soils containing 500 and 120 mg Zn kg�1.After 21e42 days the earthworms accumulated 260e470 mg Zn kg�1. Granule production was 0.027e2.11 mg CaCO3 earthworm�1 day�1 and was positively correlated with soil and soil solution pH (r ¼ 0.66and 0.85 respectively, p � 0.01). Granule Zn concentration was 34e163 mg kg�1. Granules collected fromZn-free control soil and left in Zn-bearing soil for 28 days contained 49e60 mg Zn kg�1 suggesting thatthe majority of Zn associates with granules after their secretion. However, synchrotron X-ray fluores-cence indicates some incorporation of Zn into granules during their formation. X-ray diffraction and X-ray absorption spectroscopy indicate that the granules are predominantly calcite and the spectroscopysuggests that the Zn is either adsorbed to, or incorporated into, the calcite lattice. X-ray diffraction of theouter c. 35 mm of the granules supports incorporation of Zn into the calcite lattice. The low granuleproduction rates in the mine site soils and the granule Zn concentrations suggest that earthwormsecreted calcium carbonate is unlikely to impact significantly on Zn mobility in soils.

� 2014 Published by Elsevier Ltd.

1. Introduction

Earthworms are ecosystem engineers, contributing to a range ofsoil processes that result in ecosystem services (Blouin et al., 2013).This has led to the use of earthworms in environmental remedia-tion (e.g. Butt, 1999; Wong and Ma, 2008; Boyer and Wratten,2010). Additionally, earthworms are routinely found in contami-nated soils, apparently adapted to high-metal conditions (e.g.Arnold et al., 2008; Andre et al., 2010; Kille et al., 2013). The neteffect of earthworms at contaminated sites appears to be, at leaston the time-scale of laboratory experiments, an increase in themobility of contaminants (e.g. Sizmur and Hodson, 2009; Sizmuret al., 2011). However, over significantly longer time scales theymay have other impacts.

Calcium carbonate has been proposed as an in situ soil amend-ment to reduce Zn mobility in contaminated soils (e.g. Lee et al.,

. Brinza).

Elsevier Ltd.

2009; Padmavathiamma and Li, 2010) with reductions in Znmobility being attributed to both increases in soil pH, formation ofZn-bearing carbonates and adsorption of Zn onto the carbonatesurfaces. Many species of earthworm secrete granules of calciumcarbonate (Darwin, 1881; Canti and Piearce, 2003) raising theintriguing prospect of earthworms assisting the remediation ofmetal contaminated soil via their secretions. The granules arepredominantly calcite but also contain aragonite, vaterite andamorphous calcium carbonate (Gago-Duport et al., 2008; Lee et al.,2008). In the UK, granule production is dominated by Lumbricusrubellus and Lumbricus terrestris (Canti and Piearce, 2003). Esti-mates of granule production rate are highly dependent on esti-mates of earthworm density. Lambkin et al. (2011) reportproduction rates of 0.18e31 g CaCO3 ha�1 yr�1 on the basis oflaboratory measured production rates and assuming an earthwormdensity of 10e20 L. terrestris m�2 (Briones et al., 2008) whilstWiecek and Messenger (1972) estimated that earthworm excretedcalcium carbonate could contribute up to 0.11 g CaCO3 ha�1 yr�1 toforest soils on the basis of soil measurements. The function of thegranules remains unclear with possible functions including

L. Brinza et al. / Soil Biology & Biochemistry 74 (2014) 1e102

regulation of tissue pH and secretion of Ca (Robertson, 1936;Piearce, 1972).

A link has been suggested between the Ca-rich chloragogenoustissue in which metals such as Pb and Zn accumulate (Laverack,1963; Prentø, 1979; Morgan, 1981) and the secretion of the calcitegranules; though in previous studies using electron microprobeanalysis (e.g. Prentø,1979;Morgan,1981; Schrader,1992) no Znwasdetected in the calcium carbonate produced by L. terrestris. Thestudies of Morgan (1981) and Prentø (1979) give little or no detail ofthe soils from which the earthworms were collected but theyappear to be uncontaminated. Schrader (1992) exposed earth-worms to soil amended with sewage sludge; the concentration ofZn in the amended soil was 129 mg kg�1 which is well within therange of concentrations observed in uncontaminated soils (Mertensand Smolders (2013). Given the low soil Zn concentrations and thedetection limit of electron probe analysis (0.04%, 400 ppm) it isperhaps not surprising that no Zn was detected in the granules.However, in our previous work, analysing granules from uncon-taminated soils by acid digestion and analysis with inductivelycoupled plasma e optical emission spectrometry (Lee et al., 2008)we did detect Zn in the granules sowe know that Zn can be present.Additionally we have shown that granules can incorporate both Sr(Brinza et al., 2013) and Pb (Fraser et al., 2011) during the formationprocess and that in soils with elevated levels of Sr and Pb theconcentrations of these elements in granules can be elevated aswell. Therefore it seems possible that Zn could also accumulate inthe granules, particularly given that for inorganic systems distri-bution coefficients for the partitioning of Zn into calcite suggestthat calcite will contain more Zn than the solution from which itprecipitates (Veizer, 1983) and that smithsonite (Zn carbonate) andcalcite both have rhombohedral calcite-type crystal structures(Reeder, 1983). Mitigating against this is the fact that the ionicradius of Zn2þ at 0.74�A is further from that of Ca2þ (1.00�A) than isSr2þ (1.18�A) or Pb2þ (1.19�A) (Shannon and Prewitt, 1969; Shannon,1976). Thus involvement of Zn2þ ions in physiological pathwaysdesigned for Ca2þ and substitution of Zn into the calcite lattice maybe problematic.

Nonetheless the aim of the investigation reported here was todetermine whether earthworm secreted calcium carbonate gran-ules would incorporate Zn during their production or post-production and whether sufficient Zn was associated with gran-ules to significantly impact on Zn mobility in Zn-contaminatedsoils.

2. Methods

2.1. Earthworms and soils

Mature (clitellate) L. terrestris earthworms were purchased fromThe Recycle Works Ltd, Ribchester, PR3 3XJ, UK and kept for at leastone week in a moist mixture of peat and Kettering Loam (BoughtonLoam and Turf Management, Kettering, Northamptonshire, NN168UN, UK) (1:2 by volume) before being used in these experimentsin order for non-viable earthworms to be identified and avoided foruse in our experiments.

Three soils were used in this study: Hamble (HS), an arable soilcollected from near Theale, Berkshire (UKOrdnance surveymap co-ordinates SU-618-702); and two former mine site soils, Cwmyst-wyth, CWM, (SN-803-748), a former PbeZn mine and Devon GreatConsols, DGC, (SX-426-733), a former CueAs mine.

2.2. Soil characterization

The soils were oven-dried at 40 �C and sieved to <250 mm.Subsamples were dried to 105 �C overnight and all results are

expressed on an oven-dry basis. Water holding capacity(ISO11446:1993, ISO, 1993), pH (ISO 10390:2005, BSI, 2005),organic matter content (by loss on ignition, BS EN15935, BSI, 2009)and Ca and Zn concentration (by inductively coupled plasma e

optical emission spectrometry (ICP-OES) after microwave diges-tion; BS ISO 12914, BSI 2010) were determined. An in house soilcertified reference material (SS50) traceable to BCR-143R (Com-mission of the European Communities, Community Bureau ofReference) was also digested in triplicate, and gave a recovery of103% for Ca and 95% for Zn. Repeated analysis of individual samplesindicated a precision of 0.7% for Ca and 0.4% for Zn. Detection limitswere 35 and 3.7 mg kg�1 for Ca and Zn respectively.

2.3. Earthworm incubation experiments

L. terrestris earthworms wereweighed and kept individually in aMemmert ICP 600 incubator set at 16 �C in sealed, perforatedplastic bags, each containing 300 g of air-dried, <250 mm soilmoistened to 50e70% of the soil water holding capacity. For theHamble soil, incubations used either unamended Hamble soil orsoil amended with ZnSO4 solution to give target Zn concentrationsof 250, 500 and 750mg Zn kg�1 oven dry soil (HS Zn 250, HS Zn 500and HS Zn 750 respectively) in addition to the appropriate soilmoisture content. Subsamples of the amended soils were digestedusing a microwave digestor and analysed by ICP-OES as above tocheck their Zn concentration. 5 replicates were established pertreatment.

During the experiment earthworm survival was monitoreddaily. The initial aim had been to leave earthworms in the experi-mental soils for 28 days. However, several unforeseen circum-stances prevented this from happening and earthworms were leftin the soil between 21 and 42 days. Earthworms were removedfrom the soil, weighed, depurated for 48 h (Arnold and Hodson,2007), digested by aqua regia and analysed for Zn and Ca by ICP-OES. Mussel tissue (ERM-CE278, sample no. 1570) was used as acertified reference material for the earthworm digests with a re-covery of 94% for Zn. Precision was 0.9 and 0.5% for Ca and Znrespectively. Detection limits were 70 and 7.4mg kg�1 for Ca and Znrespectively. Following removal of earthworms, soil solution wasextracted overnight with 100 mm epoxy bodied MOM Rhizonsamplers. pH was measured and then the solutions were analysedfor Ca and Zn by ICP-OES. Precision was 0.28 and 1.6% for Ca and Znrespectively; detection limits were 1.0 and 1.4 mg L�1 respectively.

Granules were extracted from the soil by sieving to 500 mmfollowing the approach of Lambkin et al. (2011). Granules wereweighed (for HS Zn 500 granules were pooled prior to weighing toensure sufficient granules for analysis). A few granules from eachtreatment were reserved for either X-ray diffraction analysis orsectioning and analysis by electron microprobe analysis (EMPA;detailed in SI, Supporting information), X-ray fluorescence (XRF)and X-ray absorption spectroscopy (XAS) (see below). Theremainder were digested in 5% HNO3 and analysed for Ca and Zn byICP-OES. Dolomite BCS No386 from the Bureau of Analysed Stan-dards was digested as a certified reference material for Ca and gaverecoveries of 98%. A synthetic limestone prepared by the ChinaNational Analysis Centre for Iron and Steel (NCS DC73345,GBW07719) was digested as a Certified reference material for Znand gave recoveries of 104%. Precision was 1.0 and 1.4% for Ca andZn respectively; detection limits were 0.8 and 1.1 mg kg�1.

Two sets of thin sections were prepared. For use in the UK,demountable thin sections of the granules were produced byembedding the granules in EpoFIX (Struers) resin and grinding to athickness of 50e70 mm, that is, 25e35 mm either side of the granulecentre. The granule slices were then mounted on Chance Glass Ltd.glass slides and mechanically polished using a 1 mm particle size

L. Brinza et al. / Soil Biology & Biochemistry 74 (2014) 1e10 3

corundum slurry. For the Australian synchrotronwork, the sampleswere prepared in a similar manner but the embedding media wasPetropoxy 154 and the samples were mounted on quartz slides.

2.4. Adsorption experiments

Granules extracted from the unamended Hamble soil wereadded to triplicate samples of Hamble soil amended with ZnSO4solution to a target concentration of 500 mg Zn kg�1 and also DGCsoil. They were recovered after 28 days and analysed for their Caand Zn concentration as above. The bulk Ca and Zn in the soil andsoil pore water were determined as above for the 500 mg kg�1

amended HS soil but the DGC soil was assumed to be identical tothat in which earthworms were incubated.

Granules from the unamended soil and calcite synthesisedfollowing the method of Rodriguez-Blanco et al. (2011) were alsoused in solution-based adsorption experiments. The granules had aBET surface area of 0.83 m2 g�1 whilst the calcite had a surface areaof 0.99 m2 g�1. For each adsorption experiment 50 mL of pH 7.5NaHCO3 and HCl solutionwas pre-equilibrated with calcite at 20 �Cfor 24 h; pH was then readjusted to 7.5 and Zn(NO3)2 was added togive Zn concentrations of 100 nM, 1 mM and 10 mM together with0.05 g of either granules or calcite. After 24 h of contact time thesupernatant was separated by centrifugation, filtered and analysedfor Zn by ICP-OES. The solid was dried and subsamples weredigested in 5% nitric acid for Zn analysis by ICP-OES. Precision andaccuracy of the analyses were as reported above. These experi-ments were not replicated as they were designed to synthesise XASstandards.

2.5. X-ray diffraction

The bulk mineralogy of the granules was assessed by combiningX-ray powder diffraction (XRD) on powdered granules and in-house, non-destructive X-ray microdiffraction (mXRD) on the pol-ished granule sections that were then used for electron probeanalysis and XAS. mXRD was also used to analyse the surfaces ofwhole granules. The probing depth is no greater than approxi-mately 35 mm for these calcium carbonate granules (Fraser et al.,2011). For both XRD and mXRD NIST silicon powder SRM640 andsilver behenate were used as external standards; calibration anddata collection were performed using Diffgrab�.

XRD data were collected in reflection geometry using a NoniusPDS 120 powder diffraction system (Schofield et al., 2002) andcobalt Ka1 radiation. Individual granules were powdered in anagate pestle and mortar, mixed with acetone and thinly depositedon a circular quartz substrate. Datawere collected for aminimumof3000 s with samples spinning continuously in the plane of thesample surface (Brinza et al., 2013).

mXRD data were collected in reflection geometry using a NoniusPDS 120 powder diffraction system as described above. In this casea 100 mm diameter beam was selected by a pinhole from a 300 mmdiameter primary beam of Cu Ka radiation generated by a GeniXsystem with a Xenocs FOX2D CU 10_30P mirror operating at 50 kVand 1 mA (Lambiv-Dzemua et al., 2012). The beam footprint on thepolished granule section was 750e500 � 100 mm, while on thegranule surface the footprint was w100 � 25 mm. During datacollection of at least 800 s, the samples were spun continuously inthe plane of the sample surface.

2.6. X-ray fluorescence mapping

Quantitative laterally-resolved element mapping was carriedout on granules fromHS250, HS750, CWM and DGC using the X-rayFluorescence Microscopy (XFM) beamline at the Australian

Synchrotron (Paterson et al., 2011). The X-ray incident energy of18,500 eV was selected using a Si (111) monochromator, and thebeamwas focused by KirkpatrickeBaez focussingmirrors to a beamsize of approximately 2� 2 mm. The elemental maps were collected‘on the fly’ in the horizontal direction, with a transit time of1.9 msec/pixel, using the 384-element Maia detector (Kirkhamet al., 2010). Processing of the full spectra XRF data was done inGeoPIXE (Ryan, 2000).

Additional m-XRF elemental mapping on granules from all thesoils was performed on the Microfocus spectroscopy beamline I18at Diamond Light Source (Mosselmans et al., 2009) where m-XRFmaps were collected using a 9-element Ge detector with an inci-dent X-ray energy of 10,500 eV, The beam footprint on the samplewas 5 � 5 mm. Maps were collected with a step size of 10 mm, inraster mode. XRF data were processed in PyMCA 4.4.1 (Solé et al.,2007). These synchrotron XRF maps were used to determine suit-able points of interest for microfocus XAS data collection.

2.7. Standards for XAS

Zn-bearing calcite was synthesised after the method ofGruzensky (1967). Solutions of ZnCl2 in CaCl2 were prepared to giveZn concentrations of 1 mM100 mMand 10mMwhichwere placed ina closed container with (NH4)2CO3 powder. After 3 weeks the solidswere separated from the supernatant by filtration and prepared forICP-OES analyses. Zn-bearing vaterite was prepared following thesynthesis method described by Bots (2012). Briefly, ZnCl2 stocksolution was added to 100 mM CaCl2 to give a final Zn concentra-tion in solution of 100 mMZn. This solutionwas used to titrate amixof 1.25 M Na2SO4 and 50 mM Na2CO3 solutions (molar ratio of25:1). Vaterite precipitation took place instantly; the solid wasimmediately washed with isopropanol using a filtration kit to avoidthe formation of calcite traces and remove the excess sulphate.

Zn adsorbed on aragonite was prepared in a batch reactor using1 g L�1 aragonite in 50 mL of solution pre-equilibrated for 24 hfollowing the same methodology used in the adsorption experi-ments described above. Zn stock solutions were spiked to give100 nM, 1 mM and 10 mM Zn as initial metal concentrations. After24 h of contact time the supernatant was separated by centrifu-gation, filtered and prepared for Zn analysis by ICP-OES. The solidwas dried and subsamples were digested in 5% nitric acid for Znanalysis by ICP-OES. Attempts to produce a standard for Zn adsor-bed on calcite using the same methodology unfortunately pro-duced a Zn-bearing calcite. In our XAS modelling we therefore useda spectrum for Zn adsorbed on calcite from the literature (Elzingaand Reeder, 2002).

2.8. Micro X-ray absorption spectroscopy (mXAS)

Full spectral XANES mapping was performed at the XFMbeamline at the Australian Synchrotron on small regions of thegranules from the HS Zn 750 and CWM soils. The regions selectedfor XANES mapping showed significant Zn concentration on thebasis of simple element mapping. Eighty seven maps of the sameregions were collected with incident energies ranging between9615 and 9810 eV (i.e. across the Zn K-edge). The beamwas focusedby KirkpatrickeBaez focussing mirrors to a beam size of approxi-mately 2 � 2 mm. The maps were collected ‘on the fly’ in the hor-izontal direction. The data were processed in GeoPIXE (Ryan, 2000)which enables the extraction of XANES spectra from any region ofinterest in a mapped area (Etschmann et al., 2010). CumulativeXANES and XANES from distinct regions within each mapped areawere extracted.

Micro XAS analyses for Ca and Zn were carried out at the I18beamline at Diamond Light Source on granules from all the soils. Zn

L. Brinza et al. / Soil Biology & Biochemistry 74 (2014) 1e104

K-edge EXAFS spectra and Ca K-edge XANES spectra were collectedin fluorescence mode at various points of interest on elementalmaps. At each point of interest the beam footprint on the samplewas 5 mm � 5 mm. The energy at the Zn edge (9659 eV) was cali-brated using a Zn metal foil. Hydrozincite, Zn5(CO3)2(OH)6(BM31917), smithsonite, ZnCO3 (SigmaeAldrich Ltd), minrecordite,CaZn(CO3)2 (BM1968,111), rosasite, (CuZn)2(CO3)(OH)2(BM1956,157) and aurichalcite, (ZnCu)5(CO3)2(OH)6(BM1964,702) weremeasured in transmission mode and calcite co-precipitated with Zn, vaterite co-precipitated with Zn and Znadsorbed onto aragonite were measured in fluorescence mode. CaXANES was recorded in fluorescence mode from natural aragonite(spelaeothem aragonite from Makapansgat Valley, South Africaprovided by Dr. A. Finch, University of St Andrews), synthetic calciteand synthetic vaterite. The Ca K-edge XANES data are distorted byself-absorption. An example of correction for the self absorptioneffect on Ca-XANES is given in the Supplementary Info of Brinzaet al. (2013). However due to the fact that all the spectra wererecorded in the same way comparisons between Ca-XANES spectraare valid. Data normalization and linear combination analysis (forXANES from �20 eV to þ70 eV) were performed, and EXAFS datafitting was carried out in the Demeter 0.9.14 package (Ravel andNewville, 2005). Principal component analysis of the XANES datawas performed in the program Sixpack (Webb, 2005).

3. Results and discussion

3.1. Soil and earthworm chemistry

Zinc concentrations in the amended Hamble soils varied abouttheir target values but were significantly different to each other(p � 0.001, ANOVA, Tukey test) (Table 1). Increasing Zn amend-ments progressively reduced soil pH by up to 1 unit (r2 ¼ 0.88,p� 0.01). Zn concentrations in the DGC and CWM soils were withinthe range covered by the amended Hamble soils (DGC was signif-icantly different fromHS500 and HS750, CWM fromHS, HS250 andHS750, p� 0.05, ANOVA, Tukey test) but pHwas significantly lower(p � 0.01, ANOVA). Soil solution Zn concentrations were variablewithin individual treatments and often below detection limits. Soilsolution Zn concentrations for CWM were below detection indi-cating a lower solubility of Zn compared with HS Zn 500 which hada similar bulk Zn concentration. Metals in laboratory amended soilsoften partition more to the soil solution phase than those in fieldcontaminated soils due to differences in metal speciation, withmetals in field contaminated soils more likely to be present ininsoluble phases or strongly sorbed to soil components (Spurgeonand Hopkin, 1995; Arnold et al., 2003; Davies et al., 2003).Although there was no significant difference in bulk Zn concen-tration between HS Zn 250 and DGC, soil solution Zn concentra-tions for DGC were higher than those for HS Zn 250. This suggests

Table 1Soil properties. Values are mean � s.d. (ND ¼ Not determined).

Parameter/Soil HS HS Zn 250 HS Zn 500

Soil pH (n ¼ 3) 8.1 � 0.23 7.6 � 0.04 7.3 � 0.03WHC, % (n ¼ 3) 40 � 0.5 ND NDLOI/% (n ¼ 1) 3.0 ND NDSoil Ca, mg kg�1 6700 � 110 7300 � 860 8800 � 928Soil Zn, mg kg�1 45 � 0.81 300 � 61 470 � 82Soil solution pH 8.2 � 0.2 7.3 � 0.1Soil solution Ca, mg L�1 280 � 16 610 � 27 1500 � 74Soil solution Zn, mg L�1 <DL <DL 4.5 � 1.04n for soil and soil solutionb 3 3 5

a Soil and solution properties assumed to be the same as the DGC soil used in the earb For earthworm-bearing soils number of replicates relates to numbers of surviving e

that the Zn in DGC was more soluble, possibly due to the lower pHof this soil.

Significant mortality occurred during the experiment (Table 2).The differing durations of our experiments and the lack of syn-chronicity in our cultures makes interpretation of the mortalitycomplicated. However, some comments can be made. We are un-aware of any published work on toxicity of Zn to L. terrestris inamended soils (Nahmani et al., 2007). Studies using Eisenia fetida(e.g. Spurgeon et al., 1994; Spurgeon and Hopkin,1995) suggest thatmortality was to be expected in the HS Zn 500 and HS Zn 750 soils;whereas in the shorter duration HS Zn 250 soil experiment the Znconcentrations were only just above NOEC levels. However, as themajority of earthworm species are more sensitive to contaminantsthan E. fetida (e.g. Spurgeon andWeeks, 1998; Langdon et al., 2005)mortality in the HS Zn 250 soil should not be considered unex-pected, although the lack of mortality in the HS Zn 500 is thensomewhat anomalous, particularly given the longer duration of thisexperiment. In experiments using similar Cwmystwyth soils lastingup to 70 days (e.g. Corp and Morgan, 1991; Andre et al., 2010)comparable levels of mortality were observed as in these experi-ments although a different earthworm species was used(L. rubellus). However, in our previous work with the same DGC soil(Sizmur et al., 2011) in experiments of up to 112 days in duration nomortality of L. terrestris was observed. Weight loss occurred in thesurviving earthworms in the majority of treatments except in theHS Zn 250 and HS Zn 750 soils in which earthworms were initiallylighter than the earthworms in the other treatments (p � 0.01,ANOVA, Tukey). Additionally the earthworms used in the DGC soilexposure had a lower mass than those used by Sizmur et al. (2011).It is possible therefore that the initial weight of the earthwormsimpacted on the mortality results. Increased sensitivity of earth-worms with decreasing size has been previously reported, thoughprimarily in the comparison of juvenile and adult data (Booth andO’Halloran, 2001; Widarto et al., 2004; Van der Ploeg et al., 2011;Anderson et al., 2013). This has been attributed both to smallerorganisms having higher surface area to body ratios, and to po-tential feeding limitations related to mouth size (Kooijman, 2000;Jager et al., 2006).

In the surviving earthworms there were no significant differ-ences in the Zn concentrations in the earthworm tissues despite thedifferent durations of exposure (p � 0.001, ANOVA, Table 2). This ismost likely because of close and rapid homoeostatic control of Znlevels via rapid uptake and excretion due to the essential nature ofthe element (Neuhauser et al., 1995; Peijnenburg et al., 1999;Spurgeon and Hopkin, 1999; Vijver et al., 2005; Nahmani et al.,2009). This tight control of Zn concentrations and rapid achieve-ment of steady state Zn levels in tissues suggests that the differingexposure periods will not have a significant impact on partitioningof Zn into the granules either. Close control of Zn uptake is sup-ported by distribution coefficients calculated as the ratio of Zn/Ca in

HS Zn 750 CWM DGC Zn 500 þ GR DGC þ GRa

7.0 5.1 � 0.1 5.5 5.5 � 0.2ND 100 � 11 110 � 5.9 ND NDND ND ND7700 220 � 8.1 1400 8100 � 870 1700 � 240810 500 � 3.9 120 550 � 56 130 � 8.57.1 5.8 � 0.1 4.6 8.2 � 0.1 4.9 � 0.6940 2.1 � 0.34 150 870 � 42 200 � 5511 < DL 1.3 2.7 � 0.47 5.4 � 1.81 3 5 3 3

thworm experiments.arthworms.

Table 2Soil properties. Values are mean � s.d. (ND ¼ Not determined).

Parameter/Soil HS HS Zn 250 HS Zn 500 HS Zn 750 CWM DGC Zn500 þ GR DGC þ GRg

Incubation period/days 42 21 28 28 42 38 28 28No. surviving

earthwormsa3/3 3/5 5/5 1/5 3/5 1/5 n/a n/a

Initial mass/g 4.3 � 0.5 2.6 � 0.2 4.5 � 0.4 3.2 4.5 � 1.0 3.9 n/a n/aFinal mass/g 4.1 � 0.5 3.2 � 0.4 4.1 � 0.8 3.8 2.9 � 0.3 3.3 n/a n/aEarthworm Ca, mg kg�1 7500 � 710 10,800 � 770 14,000 � 1500 10,100 5900 � 1070 7700 n/a n/aEarthworm Zn, mg kg�1 360 � 110 270 � 35 470 � 140 420 310 � 88 260 n/a n/aGranule Ca, mg kg�1 380000 � 12000 3,20,000 � 92,000 3,70,000 � 14,000

(n ¼ 3)b3,80,000 � 6600(n ¼ 3)d

Not enoughfor digestion

Not enoughfor digestion

3,70,000 � 8600(n ¼ 2)f

3,60,000 � 10,400(n ¼ 2)h

Granule Zn, mg kg�1 34 � 3.8 71 � 7.4 84 � 11 (n ¼ 3)b 162.84 (n ¼ 3)d Not enoughfor digestion

Not enoughfor digestion

49 � 27 (n ¼ 2)f 60 � 15 (n ¼ 2)

Granule Zn,mg kg�1 by sXRF

44 � 13 48 � 12 31 � 10 6 � 3 15 � 5

Production ratesmg CaCO3 gworm

�1 day�1 0.16 � 0.02 0.84 � 0.23 0.35c 0.19 � 0.17(n ¼ 3)e

0.0071 � 0.009 0.021 n/a n/a

mg CaCO3 worm�1 day�1 0.67 � 0.18 2.11 � 0.804 1.5 0.65 � 0.48(n ¼ 3)e

0.027 � 0.006 0.085 n/a n/a

a Number of replicates ¼ number of surviving earthworms unless otherwise stated. For Zn500 þ GR and DGC þ GR n ¼ 3 unless otherwise stated.b Samples from 5 replicates bulked then three subsamples analysed.c Samples bulked so a single value despite 5 replicate experiments.d Three subsamples from the single replicate in which the earthworm survived.e Includes granules recovered from two replicates in which earthworms died after 21 and 24 days.f Granules from one replicate were lost.g Soil and solution properties assumed to be the same as the DGC soil used in the earthworm experiments.h One batch of granules held back for XAS analysis.

L. Brinza et al. / Soil Biology & Biochemistry 74 (2014) 1e10 5

the earthworm and Zn/Ca in the soil or soil solution (Table 3). In theunamended Hamble soil, Zn was preferentially accumulated by theearthworms relative to Ca from the bulk soil (D > 1) but in theamended and contaminated soils in which Zn concentrations werehigher, Zn was preferentially excluded (D < 1). A similar trend wasseen in the soil solution data for the amended soils, with highervalues for the HS500 soil compared to the HS750 soil which has ahigher soil solution Zn concentration.

3.2. Granule production rate, bulk chemistry and bulk mineralogy

There were significant differences in granule production ratebetween treatments (ANOVA, p � 0.001, Tukey test, Table 2) withgranule production increasing with soil pH (r ¼ 0.66, p � 0.01, rankSpearman correlation) and soil solution pH (r ¼ 0.85, p � 0.01,Pearson correlation). Such correlations have been noted previously(Lambkin et al., 2011), suggestive of a role for granule production inpH regulation.

Granule production in the HS Zn 250 soil was significantlygreater than in the HS, HS Zn 750, DGC and CWM soils. It is possiblethat this was due either to pH and Zn concentrations being closestto optimal for L. terrestris and/or the earthworms in this soil beingthe healthiest. Whereas the increase in mass of earthworms in HSZn 250 over the duration of the experiment seems to support this

Table 3Mean distribution coefficientsa for Zn and Ca partitioning � std .dev. Number of replicat

Components/soil HS (3) HS Zn 250 (3) HS Zn 500 (5) HS

Earthworm/soil 7.2 � 2.6 0.61 � 0.043 0.66 � 0.22 0.4Earthworm/soil solution A A 12 � 1.9 3.7Granule/soil 0.013 � 0.001 0.006 � 0.002 0.0043 � 0.0003 0.0Granule/soil solution A A 0.074 � 0.017 0.0Granule/earthworm 0.002 � 0.001 0.01 � 0.004 0.0062 � 0.0001 0.0

A Zn in soil solution was below detection limits.B Too few granules for chemical analysis.C No earthworms in this experiment.

a Distribution coefficients calculated as, e.g. [Zn]earthworm/[Ca]earthworm/[Zn]soil/[Ca]soil.

hypothesis, the higher mortality in HS Zn 250 compared to HS andHS 750 does not. Alternatively, it has been shown that Zn caninhibit calcite growth by adsorption to active growth sites (Zacharaet al., 1988, 1991). If Znwas present in calciferous gland fluid and/orbeing incorporated into granules, at a concentration related to soilor soil solution Zn concentrations, granule production would beexpected to decrease in the soils with higher Zn amendments.

Zn concentrations in granules recovered from the amended soilswere significantly different to each other (apart from HS Zn 500 vsHS Zn 250) (p� 0.01 ANOVA, Tukey, Table 2) and increased togetherwith soil Zn concentration, potentially supporting the hypothesisthat reduction in granule production rate was due to inhibition ofcalcite growth (Zachara et al., 1988, 1991). Granules are secretedapproximately every day and earthworm Zn tissue concentrationsreach steady state in a few days (see above); thus if Zn is incor-porated into granules during their formation the longer duration ofthe higher Zn concentration experiments is unlikely to haveimpacted Zn incorporation into the granules. There were goodcorrelations (r ¼ 0.95, p � 0.001, Spearman Rank) between granuleZn concentration and both soil and soil pore water Zn concentra-tions, and a poor correlation (r ¼ 0.4, p ¼ 0.3) between granule andearthworm Zn concentrations. Distribution coefficients (Table 3)indicate that, compared to Ca, Zn is preferentially excluded fromgranules relative to the bulk soil, soil solution and earthworm

es in brackets.

Zn 750 (1) CWM (1) DGC (1) Zn500 þ GR (2) DGC þ GR (2)

0 0.025 � 0.011 0.46 C CA 3.6 C C

042 B B 0.0019 � 0.0008 0.0022 � 0.000638 A, B B 0.0444 � 0.030 0.0189 � 0.00521 B B C C

Table 5The calcium carbonate mineralogy as identified by XRD (XRD-NHM and mXRD-NHM) and the trace element chemistry from EPMA of granules. The calcite 104 peak2q position is provided based upon the wavelength of Cu Ka1 radiation.

Sample name Carbonate phasesidentified from XRD

Calcite 104peak 2q (bulk)

Calcite 104peak 2q (surface)

HS control Calcite, vaterite 29.386(2)a

HS Zn 250 Calcite, vaterite 29.342(2) 29.384(2)HS Zn 500 Calcite, vaterite 29.344(2) 29.441(2)HS Zn 750 Calcite 29.344(2) 29.467(2)CYW Calcite, vaterite 29.369(2) 29.423(2)DGC Calcite, vaterite 29.349(2) 29.425(2)

a Value taken from Brinza et al. (2013).

L. Brinza et al. / Soil Biology & Biochemistry 74 (2014) 1e106

tissue. Distribution coefficients had values 1e2 orders ofmagnitudelower than the values we determined for Sr partitioning intogranules (Brinza et al., 2013) and one or two orders of magnitudelower than values for Zn in calcite precipitated from solution ininorganic systems (Veizer, 1983) suggesting a potentially biologi-cally mediated exclusion of the Zn from calcite, a so called “vital-effect” (Weiner and Dove, 2003). As insufficient granules wereproduced in the DGC and CWM soils for digestion and analysis,samples were kept for XRF and XAS instead.

The concentrations of Zn in the “clean” granules (granulesformed in Zn-free, unamended HS) added to both Hamble soilamended to 500mg/kg Zn and DGC are significantly different to theconcentrations in granules recovered from HS Zn 750 (ANOVA,p � 0.01) but not to the Zn concentrations in granules from HS,HS250, HS500 or each other. This means that, on the basis of Znconcentrations we are unable to state with any degree of confi-dence that the Zn in the granules is incorporated into them duringthe formation process, prior to secretion into the soil (though seeXRF maps, Section 3.3 below). If Zn associates with granules post-secretion on timescales greater than a week, the longer durationof the HS Zn 500 and HS Zn 750 experiments relative to HS Zn 250may be significant in explaining differences in granule Zn concen-tration. However there is no significant difference in Zn concen-trations between HS Zn 250 and HS Zn 500 granules. A post-secretion origin for the Zn in the granules would contrast withour previous work with Pb (Fraser et al., 2011) and Sr (Brinza et al.,2013) in which the Pb and Sr concentrations in granules secretedinto soil were far higher than those in granules extracted from“clean” Hamble soil and added to Pb- and Sr-amended Hamble soil.However, the results are consistent with the size of the ionic radii ofPb2þ and Sr2þ being closer than that of Zn2þ to the ionic radius ofCa2þ thereby facilitating the substitution of the latter two ions forCa2þ. The results are also consistent with previous EMPA-basedstudies (and our own, reported in S1, Supplementaryinformation) that failed to detect Zn in the calciferous gland tis-sues or granules, albeit with a detection limit of c. 0.04% (400 ppm)(Prentø, 1979; Morgan, 1981; Schrader, 1992). Results from thegranule and calcite adsorption experiment (Table 4) also indicatethat the granules have the capacity to adsorb Zn though to a slightlylesser degree than calcite. Distribution coefficients for Zn andcalcite were similar to (but slightly lower than) those obtained forgranules and soil solution.

The incorporation of Zn into the granules was further investi-gated using XRD. The results from bulk XRD as well as XRD ongranule surfaces are summarized in Table 5 and indicate that cal-cium carbonate granules biomineralized by L. terrestris consistedmainly of calcite; although vaterite was also present in the majorityof granules analysed. Quartz and feldspar were also occasionallydetected, although these phases are considered as non-

Table 4Results of the batch adsorption experiment performed on granules and calcite.

Initial Znconcentration

CalciteorGranule

Zn concentration Zn/Ca

In solution/ On solid distribution

mg L�1 mg kg�1a mg kg�1b mg m�2c Coefficient

100 nM Calcite <DL e <DL <DL e

Granule <DL e <DL <DL e

1 mM Calcite <DL e <DL <DL e

Granule <DL e <DL <DL e

10 mM Calcite 0.56 170 110 0.17 0.018Granule 0.63 97 65 0.12 0.007

a Calculated on basis of loss of Zn from solution.b Calculated on basis of analysis of solid material.c Calculated on basis of analysis of solid material.

biosynthesized inclusions (Lee et al., 2008; Brinza et al., 2013).There was no significant difference in the mineralogy of bio-mineralized granules from the Zn amended Hamble soil and theCWM and DGC soils. Indeed the mineralogy of the granules fromthis study is similar to that of granules from Sr amended soils(Brinza et al., 2013) in that most of the granules were predomi-nantly calcite with additional vaterite. However, Fraser et al. (2011)found that granules biomineralized by L. terrestris in artificial soilamended with Pb contained large quantities of aragonite with thecalcite. The presence of trace elements is known to influence theformation of calcite and aragonite, for example Ries et al. (2008)report that Mg favours the production of aragonite in inorganicseawater systems. Aragonite and cerrusite (PbCO3) have the samestructure, as do calcite and smithsonite (ZnCO3) (Speer,1983) so it ispossible that the presence of Pb stabilised the aragonite structurewhereas any Zn incorporation would occur in the calcite structure.However strontianite (SrCO3) also has the aragonite structure sothe relationship between trace element chemistry and formation ofspecific calcium carbonate polymorph is not straight forward.

The calcite 104 peak position is a good indicator of relativechanges in the size of the calcite unit cell (Fraser et al., 2011), andthe 2q positions of this peak measured from the XRD data of thegranules in this study are reported in Table 5. A shift in the calcite104 peak to larger 2q values would indicate a decrease in the size ofthe calcite unit cell, and vice versa. The effective ionic radii of Ca2þ

and Zn2þ in octahedral coordination with oxygen are 1.00 �A and0.74 �A respectively (Shannon and Prewitt, 1969; Shannon, 1976)and consequently the substitution of Ca2þ by Zn2þ would be ex-pected to cause a decrease in the unit cell of calcite. Indeed, the unitcell of smithsonite is approximately 23% smaller than that of calcite(Reeder, 1983). While there is no significant change in the calcite104 peak position from the bulk XRD analysis of the granules(Table 5) there does appear to be a systematic increase in the 104peak position from the calcite at the surface of the granules as afunction of increased Zn loading (HS Zn 250, HS Zn 500 and HS Zn750 in Table 5). This suggests that Zn may be substituting into thecalcite at the surfaces of these granules with the proportion of Znsubstituting for Ca increasing as the concentration of Zn added tothe soil increases. This is consistent with post-secretion incorpo-ration of Zn onto the granule surface as suggested by the granulechemistry data. As Znwas below detection levels for EMPA analysis(0.04 atom %), to further investigate the distribution and mode ofincorporation of Zn within the granules we carried out XRF map-ping and Ca and Zn micro- XAS to determine the distribution andspeciation of the Zn.

3.3. Spatially resolved granule chemistry

The sXRF data for all the granules show concentrations of Znaround the edges of the grains, consistent with adsorption of Zn tothe granules post-granule formation (Fig. 1). In the granules from

Fig. 1. Micro-XRF maps obtained from thin sections of granules using the XFMbeamline at the Australian synchrotron. Maps are constructed to highlight Zn distri-bution in light colours. a) granules collected from HS Zn 250. Areas 1 and 2 show well-defined concentric zoning. The average Zn concentration across these granule thinsections was 44 mg kg�1. b) granules collected from HS Zn 750. Areas 1 and 3 showwell-defined concentric zoning near the edge of the sample. The average Zn concen-tration across these granule thin sections was 31 mg kg�1. c) granules collected fromCWM with Areas 1 through 3 showing hotspots. The average Zn concentration acrossthese granule thin sections was 9 mg kg�1. d) granules collected from DGC with Areas 1and 2 showing hotspots. The average zinc concentration across these granule thinsections was 15 mg kg�1. The three regions of interest indicated as Areas 1e3 in Fig. 1band c were selected for analysis by XANES speciation mapping. (For interpretation ofthe references to colour in this figure legend, the reader is referred to the web versionof this article.).

L. Brinza et al. / Soil Biology & Biochemistry 74 (2014) 1e10 7

HS Zn 250 and HS Zn 750, there is evidence of concentric zoningboth at the edge of the granule (e.g. Areas 1 and 3 in Fig. 1b) andtowards the centre of the granules (Fig. 1a and b). The presence ofconcentric zoning within the granules suggests that some Zn can beincorporated into the granules during their formation. Thisconclusion is also supported by our XANES analysis (Section 3.4 andS2, Supplementary information) where points with high Ca con-centrations and a Ca XANES spectrum consistent with calcite con-tained low to medium concentrations of Zn. The concentric zoningis similar to that seen for other elements (e.g. Sr, Mg, Mn, Brinzaet al., 2013) in granules and other inorganic calcite crystals(Barker and Cox, 2011). It can be explained by the incorporation oftrace elements into the calcite during mineral growth depleting thesurrounding fluid in particular elements at a rate faster than theelements are replenished by diffusion from the bulk fluid (Shoreand Fowler, 1996). The concentric zoning around the edges of thegranules could have been produced during production of thegranules. Equally, such zoning is consistent with adsorption of Zn tothe granule surface post-formation followed by incorporation of Zninto the calcite either via dissolution and reprecipitation or solidstate diffusion (Stipp, 1998; Hoffmann and Stipp, 2001). The lack ofsuch zoning in the granules from CWM and DGC may be due to thelower concentration of Zn in the soil solution of these soils limitingthe amount of Zn incorporation during granule production andsorption to the granule surfaces post-formation. Fig. 1 also showshotspots of Zn within the body of granules. These correspond toregions of low Ca concentration and are consistent with the Znbeing contained in other mineral particles which have beenincorporated into the granules as inclusions.

XRF quantitative measurements of Zn concentration for thegranule slices mapped at the Australian synchrotron are includedin Table 1. The values are taken from 60 to 100 mm slices of twogranules for each sample type. It can be seen that the values forthe granules taken from the DGC or CWM soils (ca. 15 and16 mg kg�1 respectively) are lower than those for granules fromHS Zn 250, 500 and 750, most likely reflecting the higher Znconcentrations in the soil solution of the Zn-amended HS samples.The XRF determined Zn concentrations for the HS granules arelower than the value determined by acid digestion and ICP-OESanalysis. This is undoubtedly due to the fact that the XRF ana-lyses are from a substantially smaller volume of material than thatdigested for the ICP-OES analysis. As the distribution of the Zn isvery heterogeneous the smaller sample volume of the sXRFmeasurement is more sensitive to natural variability in the system.Furthermore, the sections were obtained close to the equatorialportion of the granule, which effectively oversamples the core arearelative to surface layer. Given the lower concentration of Zn in thecore of the granule, it is not surprising that the concentrationsestimated by the sXRF measurements are lower than those ob-tained by bulk ICP-OES analysis.

3.4. Zn speciation in the granules

For all the samples investigated by XAS Ca XANES collected atthe same points at which Zn EXAFS were also collected areconsistent with calcite being the only calcium carbonate phasepresent in the granules (Gebauer et al., 2010). Zn concentrations ingranules from the HS Zn 250 soil were too low to produce XAS dataof sufficient quality for further analysis. The XANES spectra andEXAFS modelling gathered from granules extracted fromHS Zn 500and HS Zn 750 suggest that the Zn is present substituting for Ca inthe calcite lattice and associated with inclusions of soil minerals.The Zn in the CWM granules appears to be associated with soilminerals and that in the DGC granules is adsorbed to the calcite. Fulldetails of the interpretation of the XANES and EXAFS spectra

L. Brinza et al. / Soil Biology & Biochemistry 74 (2014) 1e108

obtained in this study are given in S2 and S3, Supplementaryinformation.

For the Zn amended HS granules, the majority of analyses wereperformed at points within the body of the granules where the XRFanalysis indicated that Ca concentrations were normal and Znconcentrations were relatively low or medium. The XANES (Fig. S1,Supporting information) and EXAFS (Fig. S3, and Table S1,Supporting information) were consistent with the Zn substitutingfor Ca in the calcite lattice suggesting the incorporation of Zn intogranule calcite during formation. XANES and EXAFS were also ob-tained from points at the edge of granules or in areas where the XRFmaps suggested that there might be a hole or void in the granule.These points had a low Ca content, a medium or high Zn contentand a high Fe signal and returned Zn XANES and EXAFS spectrasimilar to those obtained for Zn adsorbed on various aluminaphases such as layered double hydroxides (Li et al., 2012). Thus theZn at these points is probably in soil inclusions and not associatedwith calcium carbonate. A couple of Zn XANES spectra (Fig. S2,Supporting information), one from a granule edge with normal Calevels, one from the interior of a granule with low Ca levels weresuggestive of Zn being present in several phases with the majorityof Zn being adsorbed to the calcite, some substituting for Ca in thecalcite lattice and a small amount being associated with soilminerals.

XANES and EXAFS spectra (Figs. S1 and S3, Supportinginformation) obtained from CWM granules were all consistentwith Zn being associated with soil minerals or sphalerite (ZnS, oneof the ores mined at Cwmystwyth in its final active years, Hughes,1993), presumably as inclusions in the granules. The XANES spectraobtained for the DGC granules suggest that the Zn is presentadsorbed on the calcite, the EXAFS data were of poor quality forthese points but are consistent with this interpretation.

4. Conclusions

In this study we have detected Zn associated with earthworm-produced calcite granules recovered from Zn amended and Zn-contaminated soils. Partition coefficients suggest a vital effect,excluding Zn from granules during their formation though XRF andXANES data indicating the presence of Zn in the calcite structurewithin the body of the granules indicates some incorporation of Zninto the calcite during granule formation. The concentration of Znaround granule edges and as hotspots, the similarity of Zn con-centrations between granules secreted into Zn-bearing soils and“clean” granules added to these soils and XAS data consistent withZn adsorbed to calcite or associated with other soil componentssuggests that the majority of the Zn adsorbs to the granules post-secretion or is present as inclusions of other soil minerals. This isin contrast to previous work on Pb which appears to adsorb togranule surfaces and then become incorporated into the latticeprior to secretion (Fraser et al., 2011) and Sr which appears to beincorporated into the granules during formation (Brinza et al.,2013). The concentric zoning of Zn close to the edge of granulesindicates some inward diffusion of Zn, most likely via solid statediffusion or a dissolution e reprecipitation mechanism followingadsorption.

Regardless of the mode of incorporation of the Zn, association ofthe Zn with the granule carbonate will reduce Zn mobility andavailability in soils as observed in amendment-based remediationtrials involving lime (e.g. Lee et al., 2009; Padmavathiamma and Li,2010). The low concentration of Zn in the soil solution of the CWMand DGC soils limits the amount of Zn available for earthwormuptake and associationwith granules. Additionally, at the pH of theCWM and DGC soils, granules are likely to dissolve in the soil overtime and therefore adsorption of Zn and subsequent Zn:calcite

formation is unlikely. Additionally, granule production rates of0.027 and 0.085 mg CaCO3 per earthworm per day in the CWM andDGC soils indicate that any impact granules have on Zn mobilityand availability will be at extremely short length scales, immedi-ately adjacent to granules and on granule surfaces. Assuming adensity of 10e20 L. terrestris per m�2 (Briones et al., 2008), a soildensity of 1000 kg m�3, and a granule concentration of 50 mg kg�1

Zn, the annual production of granules would immobilise a negli-gible fraction of the total Zn in the soil. Even assuming that none ofthe granules dissolved it would take hundreds of thousands ofyears for the amount of Zn immobilised by the granules to reach the% level in the CWM and DGC soils.

Acknowledgements

We thank the Diamond Light Source for the provision ofbeamtime under grants NT2123-1 and NT5731-1. We are grateful toDr Adrian Finch, (University of St Andrews) and Professor EvertElzinga (Rutgers University) for sharing standards and spectra withus. We thank Dr Tina Geraki (Diamond Light Source) for help withaspects of the data analysis. Part of this research was undertakenusing the XFM Beamline at the Australian Synchrotron, Victoria,Australia. We thank Dr Daryl Howard and Dr Chris Ryan for assis-tance with the XRF mapping and XANES imaging.

Appendix A. Supplementary data

Supplementary data related to this article can be found at http://dx.doi.org/10.1016/j.soilbio.2014.01.012.

References

Anderson, C.J., Kille, P., Lawlor, A.J., Spurgeon, D.J., 2013. Life-history effects ofarsenic toxicity in clades of the earthworm Lumbricus rubellus. EnvironmentalPollution 172, 200e207.

Andre, J., Sturzenbaum, S.R., Kille, P., Morgan, A.J., Hodson, M.E., 2010. Metal bio-accumulation and cellular fractionation in an epigeic earthworm (Lumbricusrubellus): the interactive influences of population exposure histories, site-specific geochemistry and mitochondrial genotype. Soil Biology and Biochem-istry 42, 1566e1573.

Arnold, R.E., Hodson, M.E., 2007. Effect of time and mode of depuration on tissuecopper concentrations of Eisenia andrei, Lumbricus rubellus and Lumbricus ter-restris. Environmental Pollution 148, 21e30.

Arnold, R.E., Hodson, M.E., Langdon, C.J., 2008. A Cu tolerant population of theearthworm Dendrodrilus rubidus (Savigny, 1862) at Coniston copper mines,Cumbria, UK. Environmental Pollution 152, 713e722.

Arnold, R.E., Hodson, M.E., Black, S., Davies, N.A., 2003. The influence of mineralsolubility and soil solution concentration on the toxicity of copper to Eiseniafetida Savigny. Pedobiologia 47, 622e632.

Barker, S.L.L., Cox, S.F., 2011. Oscillatory zoning and trace element incorporation inhydrothermal minerals: insights from calcite growth experiments. Geofluids 11,48e56.

Blouin, M., Hodson, M.E., Delgado, E.A., Baker, G., Brussard, L., Butt, K.R., Dai, J.,Dendooven, L., Peres, G., Tondoh, J.E., Cluzeau, D., Brun, J.-J., 2013. A review ofearthworm impact on soil function and ecosystem services. European Journal ofSoil Science 64, 161e182.

Booth, L.H., O’Halloran, K., 2001. A comparison of biomarker responses in theearthworm Aporrectodea caliginosa to the organophosphorus insecticidesDiazinon and Chlorpyrifos. Environmental Toxicology and Chemistry 20, 2494e2502.

Bots, P., 2012. Experimental Investigations of Calcium Carbonate Mineralogy in Pastand Present Oceans (Ph.D thesis). The University of Leeds.

Boyer, S., Wratten, S.D., 2010. The potential of earthworms to restore ecosystemservices after opencast mining e A review. Basic and Applied Ecology 11, 196e203.

Brinza, L., Quinn, P.D., Mosselmans, J.F.W., Schofield, P.F., Hodson, M.E., 2013.Incorporation of strontium in earthworm-secreted calcium carbonate granulesproduced in strontium-amended and strontium-bearing soil. Geochimica etCosmochimica Acta 113, 21e37.

Briones, M.J.I., Ostle, N.J., Piearce, T.G., 2008. Stable isotopes reveal that the calcif-erous gland of earthworms is a CO2-fixing organ. Soil Biology and Biochemistry40, 554e557.

BSI, 2010. ISO 12914. Soil Quality. Microwave Assisted Extraction of Aqua RegiaSoluble Fraction for the Determination of Elements. British Standards Online.

L. Brinza et al. / Soil Biology & Biochemistry 74 (2014) 1e10 9

BSI, 2009. BS EN 15935. Soil, Sludge, Waste and Treated Biowaste. Determination ofLoss on Ignition. British Standards Online.

BSI, 2005. ISO 10390: 2005. Soil Quality. Determination of pH. British StandardsOnline.

Butt, K.R., 1999. Inoculation of earthworms into reclaimed soils: the UK experience.Land Degradation and Development 10, 565e575.

Canti, M.G., Piearce, T.G., 2003. Morphology and dynamics of calcium carbonategranules produced by different earthworm species. Pedobiologia 47, 511e521.

Corp, N., Morgan, A.J., 1991. Accumulation of heavy metals from polluted soils by theearthworm, Lumbricus rubellus: can laboratory exposure of ‘control’ wormsreduce biomonitoring problems? Environmental Pollution 74, 39e52.

Darwin, C., 1881. The Formation of Vegetable Mould, through the Action of Worms,With Observations on Their Habits. The Echo Library, Teddington.

Davies, N.A., Hodson, M.E., Black, S., 2003. Is the OECD acute worm toxicity testenvironmentally relevant? The effect of mineral form on calculated leadtoxicity. Environmental Pollution 121, 49e54.

Elzinga, E.J., Reeder, R.J., 2002. X-ray absorption spectroscopy study of Cu2þ andZn2þ adsorption complexes at the calcite surface: implications for site-specificmetal incorporation preferences during calcite crystal growth. Geochimica etCosmochimica Acta 66, 3943e3954.

Etschmann, B.E., Ryan, C.G., Brugger, J., Kirkham, R., Hough, R.M., Moorhead, G.,Siddons, D.P., De Geronimo, G., Kuczewski, A., Dunn, P., Paterson, D., deJonge, M.D., Howard, D.L., Davey, P., Jensen, M., 2010. Reduced as components inhighly oxidized environments: evidence from full spectral XANES imaging us-ing the Maia massively parallel detector. American Mineralogist 95, 884e887.

Fraser, A., Lambkin, D.C., Lee, M.R., Schofield, P.F., Mosselmans, J.F.W., Hodson, M.E.,2011. Incorporation of lead into calcium carbonate granules secreted byearthworms living in lead contaminated soils. Geochimica et CosmochimicaActa 75, 2544e2556.

Gago-Duport, L., Briones, M.J.I., Rodríguez, J.B., Covelo, B., 2008. Amorphous calciumcarbonate biomineralization in the earthworm’s calciferous gland: pathways tothe formation of crystalline phases. Journal of Structural Biology 162, 422e435.

Gebauer, D., Gunawidjaja, P.N., Ko, J.Y.P., Bacsik, Z., Aziz, B., Liu, L., Hu, Y.,Bergström, P.L., Tai, C.-W., Sham, T.-K., Edén, D.M., Hedin, N., 2010. Proto-calciteand proto-vaterite in amorphous calcium carbonates. Angewandte Chemie In-ternational Edition 49, 8889e8891.

Gruzensky, P.M., 1967. Growth of calcite crystals. In: Peiser, H.S. (Ed.), CrystalGrowth. Pergamon, Oxford, pp. 365e367.

Hoffmann, U., Stipp, S.L.S., 2001. The behaviour of Ni2þ on calcite surfaces. Geo-chimica et Cosmochimica Acta 65, 4131e4139.

Hughes, S.J., 1993. The Cwmystwyth Mines, vol. 17. Northern Mine Research Society,British Mining, Talybont, Aberystwyth.

ISO, 1993. ISO 11465:1993 (1993) Soil Quality. Determination of Dry Matter andWater Content on a Mass Basis e Gravimetric Method. ISO.

Jager, T., Heugens, E.H.W., Kooijman, S.A.L.M., 2006. making sense of ecotoxico-logical test results: towards application of process-based models. Ecotoxicology15, 305e314.

Kille, P., Andre, J., Anderson, C., Ang, H.N., Bruford, M.W., Bundy, J.G., Donnelly, R.,Hodson, M.E., Juma, G., Lahive, E., Morgan, A.J., Stürzenbaum, S.R.,Spurgeon, D.J., 2013. DNA sequence variation and methylation in an arsenictolerant earthworm population. Soil Biology and Biochemistry 57, 524e532.

Kirkham, R., Siddons, D., Dunn, P.A., Kuczewski, A.J., Dodanwela, R., Moorhead, G.F.,Ryan, C.G., De Geronimo, G., Beuttenmuller, R., Pinelli, D., Pfeffer, M., Davey, P.,Jensen, M., de Jonge, M.D., Howard, D.L., Kusel, M., McKinlay, J., 2010. The Maiaspectroscopy detector system: engineering for integrated pulse capture, low-latency scanning and real-time processing. AIP Conference Proceedings 1234,240e243.

Kooijman, S.A.L.M., 2000. Dynamic Energy and Mass Budget in Biological Systems.Cambridge University Press, Cambridge.

Lambiv-Dzemua, G., Gleeson, S.A., Schofield, P.F., 2012. The importance of lith-iophorite over asbolane in the ore zone mineralogy of the Nkamouna Co-Mnlaterite deposit, southeast Cameroon. Mineralium Deposita 48, 155e171.

Lambkin, D.C., Gwilliam, K.H., Layton, C., Canti, M.G., Piearce, T.G., Hodson, M.E.,2011. Production and dissolution rates of earthworm-secreted calcium car-bonate. Pedobiologia 54S, S119eS129.

Langdon, C.J., Hodson, M.E., Arnold, R.E., Black, S., 2005. Survival and behaviour ofthree species of earthworm in lead treated soils using the OECD acute earth-worm toxicity test. Environmental Pollution 138, 368e375.

Laverack, M.S., 1963. The Physiology of Earthworms. Pergamon, Oxford.Lee, M.R., Hodson, M.E., Langworthy, G.N., 2008. Earthworms produce granules of

intricately zoned calcite. Geology 36, 943e946.Lee, S.-H., Lee, J.-S., Choi, Y.J., Kim, J.-G., 2009. In situ stabilisation of cadmium-, lead-

, and zinc-contaminated soil using various amendments. Chemosphere 77,1069e1075.

Li, W., Livi, K.J., Xu, W., Siebecker, M.G., Wang, Y., Phillips, B.L., Sparks, D.L., 2012.Formation of crystalline ZneAl layered double hydroxide precipitates on g-alumina: the role of Mineral dissolution. Environmental Science and Technol-ogy 46, 11670e11677.

Mertens, J., Smolders, E., 2013. Zinc. In: Alloway, B.J. (Ed.), Heavy Metals in Soils:Trace Metals and Metalloids in Soils and Their Bioavailability. Springer, Dor-drecht, The Netherlands, pp. 465e493.

Mosselmans, F., Cavill, S., Peach, A., Keylock, S., Gregory, S., Quinn, P., Rosell, J.R.,Atkinson, K., Dent, A., Diaz-Moreno, S., Leicester, P., 2009. I18-the microfocusspectroscopy beamline at the Diamond Light Source. Journal of SynchrotronRadiation 16, 818e824.

Morgan, A.J., 1981. A morphological and electron-microprobe study of the inorganiccomposition of the mineralised secretory products of the calciferous gland andchloragogenous tissue of the earthworm, Lumbricus terrestris L. Cell and TissueResearch 220, 829e844.

Nahmani, J.Y., Hodson, M.E., Devin, S., Vijver, M., 2009. Uptake kinetics of metals bythe earthworm Eisenia fetida exposed to field-contaminated soils. Environ-mental Pollution 157, 2622e2628.

Nahmani, J.Y., Hodson, M.E., Black, S., 2007. A review of studies performed to assessmetal uptake by earthworms. Environmental Pollution 145, 402e424.

Neuhauser, E.F., Cukic, Z.V., Malecki, M.R., Loehr, R., Durkin, P.R., 1995. Bio-concentration and biokinetics of heavy metals in the earthworm. Environ-mental Pollution 89, 293e301.

Padmavathiamma, P.K., Li, L.Y., 2010. Effect of amendments on phytoavailability andfractionation of copper and zinc in a contaminated soil. International Journal ofPhytoremediation 12, 697e715.

Paterson, D., de Jonge, M.D., Howard, D.L., Lewis, W., McKinlay, J., Starritt, A.,Kusel, M., Ryan, C.G., Kirkham, R., Moorhead, G., Siddons, D.P., 2011. The X-rayfluorescence microscopy beamline at the Australian synchrotron. AIP Confer-ence Proceedings 1365, 219e222.

Peijnenburg, W.J.G.M., Baerselman, R., de Groot, A.C., Jager, T., Posthuma, L., vanVeen, R.P.M., 1999. Relating environmental availability to bioavailability: soil-type-dependent metal accumulation in the Oligochaete Eisenia andrei. Eco-toxicology and Environmental Safety 44, 294e310.

Piearce, T.G., 1972. The calcium relations of selected Lumbricidae. Journal of AnimalEcology 41, 167e188.

Prentø, P., 1979. Metals and phosphate in the chloragosomes of Lumbricus terrestrisand their possible physiological significance. Cell and Tissue Research 196, 123e134.

Ravel, B., Newville, M.A., 2005. ATHENA, ARTEMIS, HEPHAESTUS: data analysis forX-ray absorption spectroscopy using IFEFFIT. Journal of Synchrotron Radiation12, 537e541.

Reeder, R.J., 1983. Carbonates: mineralogy and chemistry. Reviews in Mineralogy 11.Ries, J.B., Anderson, M.A., Hill, R.T., 2008. Seawater Mg/Ca controls polymorph

mineralogy of microbial CaCO3: a potential proxy for calcite-aragonite seas inPrecambrian time. Geobiology 6, 106e119.

Robertson, J.D., 1936. The function of the calciferous glands of earthworms. Journalof Experimental Biology XIII, 279e297.

Rodriguez-Blanco, J.D., Shaw, S., Benning, L.G., 2011. The kinetics and mechanisms ofamorphous calcium carbonate (ACC) crystallization to calcite, via vaterite.Nanoscale 3, 265e271.

Ryan, C.G., 2000. Quantitative trace element imaging using PIXE and the Nuclearmicroprobe. International Journal of Imaging Systems and Technology 11, 219e230. Special Issue on Advances in Quantitative Image Analysis.

Schofield, P.F., Knight, K.S., Covey-Crump, S.J., Cressey, G., Stretton, I.C., 2002. Ac-curate quantification of the modal mineralogy of rocks when image analysis isdifficult. Mineralogical Magazine 66, 189e200.

Schrader, S., 1992. Energy-dispersive X-ray microanalysis of the calciferous glandsof Lumbricus terrestris L. (Oligochaeta) contaminated with heavy metals. SoilBiology and Biochemistry 24, 1755e1759.

Shannon, R.D., 1976. Revised effective ionic radii and systematic studies in inter-atomic distances in halides and chalcogenides. Acta Crystallographica B32,755e767.

Shannon, R.D., Prewitt, C.T., 1969. Effective ionic radii in oxides and fluorides. ActaCrystallographica B25, 925e946.

Shore, M., Fowler, A.D., 1996. Oscillatory zoning in minerals; a common phenom-enon. Canadian Mineralogist 34, 1111e1126.

Sizmur, T., Tilston, E.L., Charnock, J., Palumbo-Roe, B., Hodson, M.E., 2011. Impacts ofepigeic, anecic and endogeic earthworms on metal and metalloid mobility andavailability. Journal of Environmental Monitoring 13, 266e273.

Sizmur, T., Hodson, M.E., 2009. Do earthworms impact metal mobility and avail-ability in soil? A review. Environmental Pollution 157, 1981e1989.

Solé, V.A., Papillon, E., Cotte, M., Walter, P., Susini, J., 2007. A multiplatform code forthe analysis of energy-dispersive X-ray fluorescence spectra. Spectrochim. Acta,Part B 62, 63e68.

Speer, J.A., 1983. Crystal chemistry and phase relations of orthorhombic carbonates.Reviews in Mineralogy 11, 145e190.

Spurgeon, D.J., Hopkin, S.P., 1999. Comparisons of melta accumulation and excretionkinetics in earthworms (Eisenia fetida) exposed to contaminated field andlaboratory soils. Applied Soil Ecology 11, 227e243.

Spurgeon, D.J., Hopkin, S.P., 1995. Extrapolation of the laboratory-based OECDearthworm toxicity test to metal-contaminated field sites. Ecotoxicology 4,190e205.

Spurgeon, D.J., Hopkin, S.P., Jones, D.T., 1994. Effects of cadmium, copper, lead andzinc on growth, reproduction and survival of the earthworm Eisenia fetida(Savigny): assessing the environmental impact of point-source metal contam-ination in terrestrial ecosystems. Environmental Pollution 84, 123e130.

Spurgeon, D.J., Weeks, J.M., 1998. Evaluation of factors influencing results fromlaboratory toxicity tests with earthworms. In: Sheppard, S.C., Bembridge, J.D.,Holmstrup, M., Posthuma, L. (Eds.), Advances in Earthworm Ecotoxicology,SETAC Technical Publications Series, pp. 15e25. Pensacola, Florida, USA.

Stipp, S.L.S., 1998. Surface analytical techniques applied to calcite: evidence of solid-state diffusion and implications for isotopes methods. Palaeogeography,Palaeoclimatology, Palaeoecology 140, 441e457.

Van der Ploeg, M.J.C., Baveco, J.M., van der Hout, A., Bakker, R., Rietjens, I.M.C.M.,van den Brink, N.W., 2011. Effects of C60 nanoparticle exposure on earthworms

L. Brinza et al. / Soil Biology & Biochemistry 74 (2014) 1e1010

(Lumbricus rubellus) and implications for population dynamics. EnvironmentalPollution 159, 198e203.

Veizer, J., 1983. Trace elements and isotopes in sedimentary carbonates. Reviews inMineralogy 11, 265e299.

Vijver, M.G., Vink, J.P.M., Jager, T., Wolterbeek, H.T., van Straalen, N.M., vanGestel, C.A.M., 2005. Biphasic elimination and uptake kinetics of Zn and Cd inthe earthworm Lumbricus rubellus exposed to contaminated floodplain soil. SoilBiology and Biochemistry 37, 1843e1851.

Webb, S.M., 2005. SIX pack: a graphical user interface for XAS analysis using IFEFFIT.Physica scripta 2005, 1011.

Weiner, S., Dove, P.M., 2003. An overview of biomineralization processes and theproblem of the vital effect. Reviews in Mineralogy and Geochemistry 54, 1e29.

Widarto, T.H., Homstrup, M., Forbes, V.E., 2004. The influence of nonylphenol onlife-history of the earthworm Dendrobaena octaedra Savigny: linking effects

from the individual- to the population-level. Ecotoxicology and EnvironmentalSafety 58, 147e159.

Wiecek, C.S., Messenger, A.S., 1972. Calcite contributions by earthworms to forestsoils in Northern Illinois. Soil Science Society of America Proceedings 36, 478e480.

Wong, M.H., Ma, Y., 2008. Land reclamation using earthworms in metal contami-nated soils. In: Naidu, R. (Ed.), Developments in Soil Science. Elsevier, Bur-lington, pp. 719e734.

Zachara, J.M., Cowan, C.E., Resch, C.T., 1991. Sorption of divalent metals on calcite.Geochimica et Cosmochimica Acta 55, 1549e1562.

Zachara, J.M., Kittrick, J.A., Harsh, J.B., 1988. The mechanism of Zn2þ adsorption oncalcite. Geochimica et Cosmochimica Acta 52, 2281e2291.