alteration of organic matter during infaunal polychaete ...alteration of organic matter during...

Available online at www.sciencedirect.com

www.elsevier.com/locate/gca

Geochimica et Cosmochimica Acta 77 (2012) 396–414

Alteration of organic matter during infaunal polychaete gutpassage and links to sediment organic geochemistry. Part I:

Amino acids

Clare Woulds a,b,⇑, Jack J. Middelburg c,d, Greg L. Cowie b

a water@leeds, School of Geography, University of Leeds, Leeds LS2 9JT, UKb School of GeoSciences, University of Edinburgh, West Mains Road, Edinburgh EH9 3JW, UKc Faculty of Geosciences, Utrecht University, Budapestlaan 4, 3584 CD Utrecht, The Netherlands

d Netherlands Institute of Ecology, Korringaweg 7, 4401 NT Yerseke, The Netherlands

Received 4 November 2010; accepted in revised form 24 October 2011; available online 2 November 2011

Abstract

Of the factors which control the quantity and composition of organic matter (OM) buried in marine sediments, the linksbetween infaunal ingestion and gut passage and sediment geochemistry have received relatively little attention. This study aimedto use feeding experiments and novel isotope tracing techniques to quantify amino acid net accumulation and loss during poly-chaete gut passage, and to link this to patterns of selective preservation and decay in sediments. Microcosms containing eitherArenicola marina or Hediste (formerly Nereis) diversicolor were constructed from defaunated sediment and filtered estuarinewater, and maintained under natural temperature and light conditions. They were fed with 13C-labelled diatoms daily for 8 days,and animals were transferred into fresh, un-labelled sediment after�20 days. Samples of fauna, microcosm sediment and faecalmatter were collected after 8,�20 and�40 days, and analysed for their bulk isotopic signatures and 13C-labelled amino acid com-positions. Bulk isotopic data showed that, consistent with their feeding modes, Hediste assimilated added 13C more quickly, andattained a higher labelling level than Arenicola. Both species retained the added 13C in their biomass even after removal from thefood. A principal component analysis of 13C-labelled amino acid mole percentages showed clear differences in compositionbetween the algae, faunal tissues, and sediment plus faecal matter. Further, the two species of polychaete showed different com-positions in their tissues. The amino acids phenylalanine, valine, leucine, iso-leucine, threonine and proline showed net accumu-lation in polychaete tissues. Serine, methionine, lysine, aspartic and glutamic acids and tyrosine were rapidly lost throughmetabolism, consistent with their presence in easily digestible cell components (as opposed to cell walls which offer physical pro-tection). All sample types (polychaete tissues, sediments and faecal matter) were enriched in labelled glycine. Possible mechanismsfor this enrichment include accumulation through inclusion in tissues with long residence times, preferential preservation (i.e.selection against) during metabolism, production from other labelled amino acids during varied metabolic processes, and accu-mulation in refractory by-products of secondary bacterial production. Overall, similarities were observed between amino-aciddecay patterns in faunated microcosms, afaunal controls, and those previously reported in marine sediments. Thus, while poly-chaete gut passage did produce compound-selective accumulation and losses of certain amino acids in polychaete tissues and fae-cal matter, the impact of polychaete gut passage on sediment organic geochemistry was difficult to deconvolve from microbialdecay. Despite processing large volumes of organic matter, polychaetes may not have distinctive influence on sediment compo-sitions, possibly because metabolic processes concerning amino acids may be broadly similar across a wide range of organisms.� 2011 Elsevier Ltd. All rights reserved.

0016-7037/$ - see front matter � 2011 Elsevier Ltd. All rights reserved.

doi:10.1016/j.gca.2011.10.042

⇑ Corresponding author at: water@leeds, School of Geography, University of Leeds, University Road, Leeds LS2 9JT, UK. Tel.: +44 (0)113343 9440.

E-mail address: [email protected] (C. Woulds).

http://dx.doi.org/10.1016/j.gca.2011.10.042

mailto:[email protected]

http://dx.doi.org/016/j.gca.2011.10.042

Amino acid alteration during polychaete gut passage 397

1. INTRODUCTION

The deposition, degradation and burial of organic car-bon in marine sediments are key features of the global car-bon cycle, and represent a mechanism for long-term Csequestration. In addition, the deposition of organic matter(OM) is a significant source of energy and elemental rawmaterials for most benthic communities, and thus sedimen-tary OM dynamics exert a strong control over marine ben-thic biodiversity.

A significant amount of previous study has been devotedto understanding the factors and processes which controlthe abundance and composition of deposited and pre-served/buried sedimentary OM (Hedges and Keil, 1995;Burdige, 2005, 2007). Factors controlling OM depositioninclude water column primary productivity, water depthand sedimentation rate, and these combine with factorssuch as OM source and reactivity, dissolved oxygen concen-tration and sediment grain size to control bulk sedimentOM contents and preservation (e.g. Canfield, 1994; Hedgesand Keil, 1995; Hartnett et al., 1998). The organic geo-chemical composition of sedimentary OM is also controlledby these factors. Most notably however, OM compositionhas been shown to relate to OM degradation state. A rangeof studies have shown that the contributions of amino acidsand carbohydrates to total organic matter, and the relativeconcentrations of biochemicals within particular classes(e.g. pigments, amino acids or carbohydrates) vary system-atically with progressive degradation, and these composi-tional trends are replicated in multiple locations (e.g. Leeand Cronin, 1982; Cowie et al., 1992, 1995; Cowie andHedges, 1994; Dauwe and Middelburg, 1998; Hedgeset al., 1999; Woulds and Cowie, 2009).

Of the processes which control sedimentary OM compo-sition, one has received relatively little attention. Most geo-chemical studies consider OM degradation as a dominantlymicrobial process, and thus the role of invertebrate inges-tion and gut passage in transforming OM compositionshas been overlooked. However, benthic invertebrates havebeen shown to be extremely efficient at rapidly searchingout and ingesting freshly deposited, highly bioavailableOM (e.g. Miller et al., 2000), and in some settings almostall sedimentary OM may pass through the guts of fauna be-fore being subjected to further degradation processes (e.g.Lauerman et al., 1997). Moreover, the digestive tracts ofbenthic invertebrates contain bacteria, enzymes, acids andsurfactants (Longbottom, 1970; Mayer et al., 1997; Ahrenset al., 2001a; Voparil and Mayer, 2004), and are thusadapted to break down complex polymers, releasing simplebiochemicals for assimilation. The rate at which these gutfluids act is very rapid, with Ahrens et al. (2001a) reportingthe majority of contaminant desorption to occur in the firstminute of contact with sediment. Further, gut passage hasbeen shown to move OM between redox zones, causingphysical alteration of sedimentary OM, repackaging it intofaecal pellets, and exposing new surfaces (Aller, 1994).Thus, gut passage causes significant biochemical alterationof OM. However, while many studies have examined faunaldigestion and assimilation of sediment constituents from anecological standpoint, conclusions have only rarely been

drawn regarding the impact of faunal activity on sedimentorganic geochemistry.

Early studies of the polychaete gut processes weremostly concerned with morphology, histology, and the de-tails of peristaltic movements, although a few studies alsolooked at iron absorption and pigment assimilation (Fox,1950; Dales, 1955, and references therein). A considerableamount of research into infaunal invertebrate gut passagehas focused on determining the bioavailability and assimila-tion efficiency of heavy metal and organic pollutants (e.g.Wallace and Lopez, 1997; Chen and Mayer, 1999; Ahrenset al., 2001a; Casado-Martinez et al., 2010; Janssen et al.,2010). These and other studies have identified a wide rangeof factors that control benthic invertebrate ingestion ratesand assimilation efficiencies. These factors include foodsource (e.g. macroalgae species), and composition (Har-grave, 1970; Amouroux et al., 1989, 1997; Charles, 1993;Charles et al., 1995, 1996), surfactant micelle concentrationin gut fluids (Voparil and Mayer, 2000; Ahrens et al.,2001a), and the sediment to fluid ratio in the gut (e.g. Chenand Mayer, 1999; Ahrens et al., 2001a). Further, severalstudies suggest that animal taxon and physiology areimportant controls on faunal feeding and digestive pro-cesses (Mayer et al., 1996; Wang et al., 2002). Different taxahave been shown to ingest different particle sizes, and toassimilate compounds from different food sources withvarying efficiency (Gremare et al., 1989). Ahrens et al.(2001a) report that Nereis succinea showed significantlyhigher gut surfactant concentrations than Pectinaria goul-

dii, and Chen and Mayer (1999) found that Arenicola gutfluids released more Cu from sediment than two holothuri-ans (Chen and Mayer, 1998).

In addition, previous research supports the suggestionthat invertebrate gut passage will influence OM composi-tion. For example, Thomas and Blair (2002) showed thatpolychaete faecal pellets were enriched in amino acidsincluding glycine and leucine compared to the food mate-rial, and Cowie and Hedges (1996) studying zooplanktonobserved particularly efficient removal during gut passage(i.e. absence from faecal pellets) of compounds including ly-sine, glutamic acid and glucose. Bradshaw et al. (1989,1991) have shown that marine invertebrates preferentiallyassimilate mono- and poly-unsaturated fatty acids over sat-urated forms, and that certain sterols are not absorbed atall.

A number of methodological constraints have limitedthe questions addressed by previous studies. Firstly, manyprevious studies were conducted without the use of isotopiclabels (e.g. Bradshaw et al., 1991). This limits findings tocompositional differences between food, faecal matter,and organism tissue. An improvement was offered in theform of 14C-labelled food sources, either as algae (Har-grave, 1970), or as formaldehyde sorbed onto organic mat-ter (Charles et al., 1995). When combined with modelling ora conservative Si tracer, this was capable of distinguishingbulk carbon uptake (Amouroux et al., 1989, 1997), or over-all losses of individual amino acids (Cowie and Hedges,1996). However, it did not allow assessment of assimilationor retention of compounds in the faunal biomass, and itcannot fully resolve production or transformation of bio-

398 C. Woulds et al. / Geochimica et Cosmochimica Acta 77 (2012) 396–414

chemicals within animal tissues or by gut bacteria (e.g.Bradshaw et al., 1990). Compound-specific isotopic analysishas recently become widely available, which means thatthese limitations can now be overcome.

Secondly, experiments used to quantify absorption effi-ciency during gut passage have usually involved feedingfor very short periods (hours) in artificial environments(e.g. in tubes), where the sediment ingested can be tightlycontrolled and faecal matter can be easily collected (e.g.Ahrens et al., 2001a). Thus, such experiments do not allowstudy of animals in their natural state, or of long-term,simultaneous compositional changes in the ambient sedi-ment. This focus on organisms in previous studies, hastherefore led to a comparative lack of links being made be-tween gut passage and more widely observed compositionalchanges during OM decay.

The amino acids are an important biochemical class inthe marine environment, which can contribute as much as37% of the C and 81% of the N in sinking organic matter,and 10% of the C and 37% of the N in OM deposited incoastal sediments (Cowie and Hedges, 1992a). Further,they have been shown to constitute as much as �30% ofthe most labile fraction of settling organic C, and thus rep-resent an important food source for benthic communities(Cowie et al., 1992). In addition, amino acid suites havebeen shown to change in a systematic way during OM de-cay (Cowie and Hedges, 1994; Dauwe and Middelburg,1998), allowing principal component analysis to be usedas an indicator of OM degradation state and bioavailability(Dauwe and Middelburg, 1998; Dauwe et al., 1999; Van-dewiele et al., 2009). Given this pre-existing link betweenamino acid composition and OM degradation, the aminoacids provide a logical focus for the consideration of thelinks between faunal gut passage and OM decay.

Considering the limitations of previous work, the objec-tives of this study were: (1) to characterise the pattern of netamino acid accumulation in faunal tissues; (2) to character-ise the pattern of amino acid loss through metabolism; (3)to link net accumulation and metabolism by fauna to pat-terns of OM decay. Our overarching hypothesis was thatcompound selective faunal gut passage and accumulationprocesses (encompassing digestion, assimilation, anabo-lism, transformation and metabolism) play a role in deter-mining the long-term compositional patterns observed inparticulate OM degradation. This hypothesis was testedthrough long-term (40-d) feeding experiments using a 13C-labelled food source, conducted in microcosms that mim-icked the natural environment.

2. METHODS

2.1. Study species and sites

Experimental work was conducted at the NetherlandsInstitute for Ecology, Centre for Estuarine and MarineEcology (CEME), in Yerseke, The Netherlands.

Two polychaete taxa were chosen for study, the lug-worm Arenicola marina, and the ragworm Hediste (for-merly Nereis) diversicolor. They will hereafter be referredto as Arenicola and Hediste, respectively. These species

were chosen because they are both common and ubiqui-tous in European estuaries, and play a significant role inmediating sediment biogeochemistry (Jones and Jago,1993; Kristensen, 2001 and references therein; Nielsenet al., 2003). At a typical density of 40–80 ind. m�2, Aren-

icola reworks a sediment column of 17–40 cm in a year.For Hediste at a density of 1000 ind. m�2 the length ofsediment column reworked in a year is a still considerable0.7 cm (Kristensen, 2001). Further, the two species repre-sent different lifestyles and feeding modes, and so producean interesting contrast. Arenicola is a head-down conveyorfeeder, ingesting detritus nonselectively at depth andexcreting at the surface, and is relatively sedentary onceit has constructed a burrow (Jones and Jago, 1993;Retraubun et al., 1996; Riisgard and Banta, 1998). Areni-

cola also has a marked impact on sediment biogeochemis-try through its intense bioirrigation activities (e.g.Hylleberg, 1975; Rasmussen et al., 1998; Volkenbornet al., 2010). Hediste, by contrast, has been classified asa surface deposit feeder, a carnivore and a scavenger(Evans, 1971; Ronn et al., 1988), and has also been shownto filter feed in appropriate conditions (Riisgard, 1991).Hediste exhibits considerable motility around a ‘gallery’of burrows (Francois et al., 2001), and will also emergeonto the sediment surface to feed (personal obs.). Thus,the diet of Hediste includes fresh detritus, suspended par-ticulate organic matter and small invertebrates, while thatof Arenicola is dominated by buried OM and possibly‘gardened’ bacteria (Hylleberg, 1975). It is therefore rea-sonable to suspect that they will exhibit different patternsof OM alteration during gut passage.

Sediment and polychaete specimens were collected fromtwo intertidal sites in the Scheldt estuary in the south of theNetherlands. Specimens of Hediste, plus surface sediment,were collected from a location (51.55384N, 3.873247E)exhibiting relatively fine grained, muddy sediments, withan organic carbon content (% Corg) of 0.56%. A separateintertidal site (51.487262N, 4.058526E), showing coarser,less organic rich (% Corg = 0.19%), sandy sediments wasused for the collection of Arenicola individuals and accom-panying surface sediment.

2.2. Microcosm construction

Microcosm construction and experiments were con-ducted in a controlled temperature laboratory which wasset to the seasonal average temperature (15 �C), and withthe lights set to a natural light/dark cycle (typical daylength 15–16 h in May at the site latitude).

Surface (top 10 cm) sediment was defaunated before usein microcosms by being twice frozen and thawed to killmacro- and meiofauna. After homogenisation it was thenused to partly fill microcosm containers, which were toppedup with filtered estuarine water (water column at least10 cm). In the case of Hediste, microcosm containers con-sisted of 10 cm i.d. acrylic tubes with rubber stoppers atthe base, and sediment columns of �15 cm. For Arenicola,which require more space per individual, rectangular plastictanks with a surface area of 980 cm2 were used, and the sed-iment column was always longer than 20 cm. Microcosms


were allowed to stand overnight before animals wereintroduced.

Polychaete specimens (all of a similar size within eachspecies, to avoid variation in feeding and growth rates, Ah-rens et al., 2001b) were kept in filtered seawater for 3–5 days before introduction into microcosms, and were accli-mated in microcosms for 2–3 days before feeding experi-ments began. Hediste were added to microcosms at adensity of 1018 ind. m�2. This is below the ‘typical’ densi-ties of 3000–4000 ind. m�2 reported by Kristensen (2001),however it is within the natural range, being higher thanthe 357 ind. m�2 found on an Eden estuary mudflat (Defewet al., 2002). Individuals had an average wet weight of 1 g.The density of Arenicola in mesocosms was 71 ind. m�2, inline with typical reported densities of 40–80 ind. m�2 (Kris-tensen, 2001). Average individual wet weight was 6.7 g,equivalent to �13,400 mg ash-free dry weight per m2, andwithin the range of 1.2–22.5 g ash-free dry weight per m2

at the nearby Molenplaat (Herman et al., 2000).Microcosms were kept aerated by bubbling air into the

surface water. This aeration also served to keep the waterin the microcosms mixed. In order to avoid the build-upof metabolites and increasing salinity due to evaporation,the overlying water in each microcosm was replaced withnew filtered estuarine water once every �4 days. The micro-cosms were open to the atmosphere for aeration and 13C–CO2 generated could thus escape from the system, prevent-ing determination of closed C budgets.

Polychaete survival was good up to the 21/22-d time-point (see below), with one individual recovered dead fromsome microcosms. At the 37/38-d timepoint survival wasnot as high, with 3–4 individuals missing from some micro-cosms. Missing animals are presumed to have decayed inthe microcosms, as escape would not have been possible.

2.3. Feeding experiments

Polychaetes were fed by addition of 13C-labelled(64.6 atom%) diatom detritus (Skeletonema costatum),grown at the Netherlands Insitute of Ecology, Yerseke. Al-gal slurry (freeze-dried algae resuspended in filtered estua-rine water) was introduced into the water column of eachmicrocosm once per day on 8 consecutive days, after whichno additional feeding took place. The total feeding doseafter all 8 feeding days was 3.1 g C m�2 for Arenicola,and 3.5 g C m�2 for Hediste microcosms. The high dosesused were to ensure detectable label at the end of the exper-iments (see below), and the slight discrepancy between spe-cies was due to the practical constraint of 13C-labelled algaeavailability. As soon after the end of feeding as possible (onday 9 for Arenicola, day 10 for Hediste) a sub-set of themicrocosms was sacrificed and samples were preserved.The next timepoint occurred on day 21 for Arenicola andday 22 for Hediste. At this timepoint another sub-set ofmicrocosms was sacrificed. In addition, animals from theremaining microcosms were removed with great care, andtransferred to newly-constructed equivalents that did notcontain any 13C-labelled algae. This was in order to observethe retention or loss of amino acids from their tissues. Ani-mal and sediment samples from these new microcosms were

collected on day 37 for Arenicola, and on day 38 forHediste.

For Arenicola (due to the amount of algae required),there was one microcosm per timepoint, each containingmultiple animals. For Hediste, two replicate microcosmswere sacrificed at each timepoint. It is recognised that mul-tiple animals within one microcosm are pseudoreplicates,rather than true replicates. Additional replicate microcosmscould not be run due to the expense, and thus limited avail-ability, of labelled algae, as well as the labour-intensive nat-ure of amino acid analyses. We chose not to construct alarger number of small microcosms, as we were attemptingto replicate natural conditions, and wished to minimise walleffects. Examination of polychaete bulk isotope signatures(see below) from Hediste replicate microcosms shows nosignificant difference at either 10 or 22 day timepoints (t-test, p = 0.869 and 0.402, respectively), and this lack of var-iation between microcosms suggests that variation betweenindividual animals is the primary source of spread in thedata.

In addition to the faunated microcosms, two ‘control’microcosms, to which no animals were added, were con-structed (in 54 mm diameter acrylic tubes) using defaunatedsediment from each site. These were fed as for the faunatedmicrocosms. One core from each pair of controls was sec-tioned on day 9 or 10, and the other on day 37 or 38 (forArenicola and Hediste controls, respectively). These micro-cosms were smaller than the microcosms with fauna due tothe limited availability of labelled algae. This is unlikely tohave affected the data, as the absence of fauna means thatthe dominant processes will have been microbial, and thusnot affected by the small-scale patchiness of macrofaunalcommunities. Similarly, limited availability of algae is thereason why control cores were not sacrificed at the 21/22-d timepoint.

For Arenicola the process of sacrificing a microcosm en-tailed extruding and sectioning (at 0.5 cm intervals to 2 cmdepth, followed by 1 cm intervals to 10 cm depth) three sub-cores (two taken with a 50-ml syringe with the end re-moved, and one with a 54 mm i.d. core tube) from themicrocosm. Overlying water was then siphoned off, andall visible faecal matter was collected and preserved. Theremaining sediment was then dug out with great care soas not to damage the polychaetes. In the case of Hediste

microcosms, overlying water was siphoned off and sedi-ments were extruded and sectioned. Care was taken notto damage the polychaetes, which were extracted by handas they were found. All sediment samples were placed inbags and frozen. Polychaetes were transferred to large petridishes containing filtered estuarine water and left for 2–5 hto egest their gut contents. Polychaetes and gut contentswere then transferred to pre-combusted glass vials, and fro-zen. It should be noted that gut evacuation was not com-plete (�80% complete, visual estimation), as this couldhave taken days-weeks in the starvation conditions of thepetri dish. Such a delay could have produced changes inamino acid metabolism. Data from gut contents are notpresented, as the samples were too small to analyse. There-fore, the lack of this data is unlikely to have impacted sig-nificantly on quantitative budgets (see Section 3.2.3).


2.4. Analytical techniques

Polychaete samples were defrosted and homogenisedusing a Potter tube tissue grinder prior to being freezedried. Freeze-dried Hediste sediments were gently disaggre-gated using a pestle and mortar to aid homogenisation.This was not necessary for Arenicola sediments, which wereeasily disaggregated by manually squeezing sample bags.

Sub-samples of faunal tissues, sediments and faecal mat-ter was prepared for bulk C isotopic analysis by weighinginto ultraclean silver capsules, and decarbonation with6 N (sediment and faecal matter) or 0.1 N (faunal tissues)HCl. Samples were analysed on a Europa Scientific (Crew,UK) Tracermass isotope ratio mass spectrometer (IRMS)with a Roboprep Dumas combustion sample converter atthe Scottish Crop Research Institute. Carbon contents weredetermined from IRMS peak areas and calibrated againststandards of acetanilide. This method gave an average rel-ative standard deviation of 4.6% for organic carbon quan-tity, and 0.7 per mil for d13C (n = 27, Woulds et al., 2007).

Quantification of 13C-labelled amino acids in sediment,fauna and faecal matter was conducted as described byWoulds et al. (2010). Briefly, samples were hydrolysed inoxygen-free 6 N HCl for 70 min at 150 �C. Charge-matchedinternal standards (Cowie and Hedges, 1992b) were added,and samples were dried. This was followed by ion exchangecleanup using Dowex 50W8 resin, from which amino acidswere eluted in NH4OH and again dried. Derivatisation totrifluoroacetic isopropyl esters was initiated by dissolvingthe samples in acidified propanol and heating to 110 �Cfor 70 min. Samples were then dried under a stream ofN2, and twice washed and dried with dichloromethane.Samples were then acetylated by dissolving in trifluoroace-tic anhydride and dichloromethane, and heating to 110 �Cfor 10 min. Samples were finally twice washed and driedwith dichloromethane.

Analysis was achieved using a GC–MS equipped with a30-m, 0.32 mm diameter, 0.25 lm film thickness Equity 5column (Supelco). The initial temperature of 35 �C was heldfor 1.5 min before ramping to 100 �C at 5.5 �C min�1. Thiswas followed by a ramp at 4 �C min�1 to 190 �C, and thento 280 �C at 70 �C min�1 with a 5 min hold to bake the col-umn. The injector temperature was 290 �C, and the He car-rier gas flow rate was 1.5 ml min�1.

Positive ion chemical ionisation with CH4 at 20% andselective ion monitoring were used. Quantification wasbased on a characteristic, pseudomolecular ion for eachof the natural (unlabelled) and added (13C-labelled) ver-sions of each amino acid. Preservation of these large ionsrequired the use of a relatively low ion source temperature(154 �C). For details see Woulds et al. (2010).

Naturally present (unlabelled) amino acids in eachsample were quantified using the internal standards anda single external standard containing known amounts ofall compounds. Since unlabelled standards were not suit-able for quantifying 13C-labelled amino acids (as theydid not yield the appropriate quantifier mass fragments),isotope calibration curves were constructed for this pur-pose. Known volumetric mixes of derivatives of 13C-la-belled algae and natural, unlabelled sediment were made

and analysed by GC–MS. The amino acid concentrationsin each derivative were independently quantified before-hand using GC-FID. Therefore, the amounts of algae-de-rived and sediment-derived amino acids in each mix wasknown, and calibration curves were constructed which re-lated this to the ratio of the MS responses for the labelledand unlabelled quantifier ions of each amino acid. Thesecalibration curves were used to quantify 13C-labelled ami-no acids in samples, based on the quantity of the unla-belled version, and the ratio in MS response between thelabelled and unlabelled quantifier ions. For further discus-sion see Woulds et al. (2010).

2.5. Data treatment

Results are presented as the average ± standard devia-tion of replicate/pseudoreplicate microcosms and animals.Error bars consequently include analytical uncertainty(small) and experimental variability (larger component).Normality testing, Mann–Whitney U and Kruskal–Wallistests, and principal component analyses were applied tothe data (see below), using Minitab 15.

3. RESULTS

3.1. Bulk 13C uptake

Bulk isotopic signatures show that Hediste rapidlyassimilated the added diatom detritus, with some individu-als becoming highly 13C-enriched (Fig. 1a). Arenicola indi-viduals also assimilated added algal material (Fig. 1b),but after 9 days were still not as 13C-enriched as the surface0.5 cm of sediment. In addition, Hediste (max.d13C = 2454&) attained higher maximal labelling levelsthan Arenicola (max. d13C = 1273&). These trends areunsurprising, since, as a head-down feeder, Arenicola willnot have had access to the labelled algae until it penetratedto the depth of the burrow bottoms. In contrast, Hediste

individuals were observed to emerge from their burrows�5 min after feeding, presumably to actively collect thefresh food. In general, faunal tissues showed their maximalisotopic enrichments at the end of feeding (9 or 10 days).However, there were no significant decreases in tissue isoto-pic signatures with time (Kruskall–Wallis H = 0.72, 0.18;P = 0.697, 0.912; DF = 2 for Hediste and Arenicola, respec-tively; Fig. 1). Thus, the assimilated C appears to be effi-ciently retained in the biomass, even after the 13C-labelledfood source was removed (i.e. at day 21 or 22).

3.2. Amino acids

3.2.1. Amino acid suites

The suite (expressed as mole percentage of all aminoacids detected) of 13C-labelled amino acids present variedconsiderably between faunal tissue, faecal matter, sediment,and the source algae (Fig. 2), indicating that net accumula-tion and loss processes were compound-selective. The visu-ally evident trends in Fig. 2 include an enrichment of faunaltissues in glycine at the expense of other amino acids suchas alanine in comparison to the 13C-labelled algae

1.0

1.5

2.0

2.5

3.0

3.5

4.0

Atom

%13

C

Atom

%13

C

Natural Background PolychaetesFeeding Experiment PloychaetesSurface Sediment

1.0

1.5

2.0

2.5

3.0

3.5

4.0

0 5 10 15 20 25 30 35 40

Time (d)

13C Uptake by Arenicola marina

13C Uptake by Hediste diversicolor (a)

(b)

Fig. 1. Isotopic compositions of polychaete tissues (experimental and natural background) and microcosm surface (0.5 cm) sediments withtime through the experiments.


(Fig. 2a). Amongst sediment samples there were enrich-ments in alanine, glycine and leucine and depletions inaspartic and glutamic acids compared to the source algae(Fig. 2b).

In order to conduct a rigorous analysis of compositionaldifference between samples, a principal component analysis(PCA) was carried out on 13C-labelled amino acid mole per-centage data (Fig. 3). Principal component 1 (PC1) ex-plained 32% of the variation, and principal component 2(PC2) a further 17% (totalling 49% for PC1 and PC2 to-gether). The PCA revealed clear differences in compositionbetween the 13C-labelled algae and all other sample types(fauna, sediments and faecal matter), which all showed low-er principal component 1 scores than the algae. In addition,faunal tissues were separated from algae on the PC2 axis,showing lower scores. Factor coefficients (Fig. 3b) indicatethat these differences in scores result from the algae beingrelatively enriched in aspartic acid, glutamic acid, serineand methionine compared to samples taken from the micro-cosms, suggesting that these compounds are preferentiallymetabolised. Conversely, polychaete tissues, faecal matterand sediment samples were relatively enriched in glycine,leucine and proline. Polychaete tissues were separated fromsediment and faecal matter samples by their lower PC2

scores (t-test, P < 0.001; Fig. 3a), indicating that they tendto have greater mole percentages of proline, phenylalanine,valine and lysine (Fig. 3b).

In addition, the two species of polychaete showed differ-ent labelled amino acid compositions, denoted by their sep-aration from each other along the PC1 axis (t-testP < 0.001; Fig. 3a). This is largely linked to greater molepercentages of glycine in Arenicola tissues compared tothose of Hediste (Fig. 3b). Thus, amino-acid-selective accu-mulation and transformation/loss processes are, to someextent, taxon-specific.

3.2.2. Selective net accumulation

The use of mole percentages (relative concentrations)and PCA to elucidate selective accumulation or loss of var-ious amino acids has its limitations, as a marked shift in theabundance of one major amino acid (e.g. glycine) can eitherforce or mask smaller changes in the mole percentages ofless abundant compounds. In order to avoid this pitfall,amino acid suites for faunal tissues were also comparedon an absolute basis. Variations in concentrations due todifferent levels of bulk 13C uptake by fauna were normalisedby expressing 13C-labelled amino acid concentrations as mi-cro moles of compound per mg of assimilated 13C. These

Fig. 2. Mole percentage compositions of 13C-labelled amino acids in faunal tissues, faecal matter (a), and microcosm sediments (surface anddowncore, b). Faunal and faecal matter data are averages. Surface sediment data are for the top 0.5 cm, and downcore data are for the 1.5–2 cm horizon. n = 1 for all samples in panel B, as there is no justifiable basis for averaging across different depth horizons. ALA = alanine,GLY = glycine, THR = threonine, BALA = b-alanine, SER = serine, VAL = valine, LEU = leucine, ISO-LEU = iso-leucine, MET = methi-onine, PRO = proline, PHE = phenylalanine, GABA = c-aminobutyric acid, ASP = aspartic acid, GLU = glutamic acid, TYR = tyrosine,LYS = lysine, ORN = ornithine.


were then used to derive an ‘enrichment factor’, by sub-tracting the 13C-normalised concentration found in thesource algae from the 13C-normalised concentration foundin each polychaete sample (Fig. 4). In fact, enrichment fac-tors derived from 13C-normalised absolute concentrationsshowed similar patterns (although not precisely the same)to those calculated in the same way using the data in molepercentage form (Fig. 4b).

Enrichment factors show that faunal tissues consistentlycontained certain amino acids in lower concentrations/pro-portions than those in which they were present in the sourcealgae. These amino acids were alanine, serine, methionine,aspartic acid, glutamic acid, tyrosine and lysine (mole per-centages in polychaete tissues statistically significantly low-

er than in algae for all except lysine, Mann–Whitney U,P 6 0.0067). It is likely that these compounds were eitherselected against during solubilisation and absorption inthe gut, or rapidly metabolised in polychaete guts and tis-sues. The possibility remains that they were rapidly metab-olised before ingestion. However, this seems unlikely, as thelow concentrations were present in polychaete tissues evenat the first sampling timepoint, which was only one dayafter the final dose of algae had been added to the micro-cosms. Further, visual observations suggest that Hediste

individuals tended to leave their burrows and feed on thesurface within minutes of algae being added; thus they werevery likely selectively ingesting fresh algae, at least for thefirst 8 days of the experiment.

-4

-3

-2

-1

0

1

2

3

4

5

-5 0 5 10PC1

PC2

Hediste

Arenicola

Hediste sed

ArenicolasedArenicolafaecal matterControls

AlgaeHedisteTissues

Sediments and Arenicola Faecal Matter

ArenicolaTissues

ALA

GLY

THR

BALA SER

VAL

LEUISO-LEU

MET

PRO

PHE

GABA ASP

GLUTYR

LYS

ORN

-0.4

-0.3

-0.2

-0.1

0

0.1

0.2

0.3

0.4

0.5

-0.4 -0.3 -0.2 -0.1 0 0.1 0.2 0.3 0.4

PC2

PC1

(b)

(a)

Fig. 3. Results of a PCA including all samples showing (a) sample scores, and (b) amino acid factor coefficients. Circles indicate groupings ofcompositionally similar samples.


In contrast to the above, several amino acids were oftenpresent in faunal tissues in greater concentrations/propor-tions than those in which they occurred in the source algae.The most notable example was glycine (mole percentages inpolychaete tissues statistically significantly higher than inalgae, Mann–Whitney U, P = 0.0044), but threonine, va-line, leucine, proline, and phenylalanine also showed thistendency (Fig. 4). This net accumulation in tissues couldhave been the result of selective assimilation, retentionand/or production in faunal tissues.

Enrichment factors could not be calculated for sedimentand faecal matter samples due to a mismatch in the samplesanalysed for bulk isotopes and amino acids. Therefore dif-ferences in mole percentage compositions between sedi-ments/faecal matter and algae have to be consideredinstead, and these are open to interfering effects betweendifferent amino acids. Despite this limitation, mole percent-age differences showed similar patterns to faunal tissueenrichment factors. These included marked losses of aspar-

tic and glutamic acids, as well as lysine, methionine, valine,serine, and threonine, and enrichments in glycine and leu-cine (Fig. 4c and d). Once again, these could be ascribedto particularly rapid metabolism of certain compoundseither by polychaetes or sediment/gut bacteria. Enrich-ments (or accumulation) of certain amino acids in the sed-iment during decay have previously also been ascribed tosorptive protection from degradation afforded by the dia-tom silica test, and this mechanism remains as a possibleexplanation for the glycine enrichment observed here (e.g.Nguyen and Harvey, 1997).

3.2.3. Quantitative budgets

Labelled amino acid concentrations from sediments andpolychaetes were combined with sediment porosity data toproduce quantitative budgets for each amino acid at the 9–10 and 21–22 d timepoints (Fig. 5). Meaningful amino acidbudgets for the 37–38 d timepoint could not be constructedbecause animals were transferred to unlabelled sediment for

-0.4

-0.2

0

0.2

0.4

0.6

0.8 Enrichment Factor(a)

(b)

Enric

hmen

t Fac

tor

Hediste 10-dHediste 22-dHediste 38-dArenicola 9-dArenicola 21-dArenicola 37-d

-10

-15

-5

0

5

10

15

Diff

in M

ole

% (S

ampl

e-al

gae)

Mole % Difference from Algae

Hediste 10-dHediste 22-dHediste 38-dArenicola 9-dArenicola 21-dArenicola 37-d

Fig. 4. Enrichment of samples in each amino acid compared to the source algae. Note, in a and b the y axis is truncated. Values for glycinewere up to 3.5 (panel a) and 50 (panel b). Plot A compares the 13C normalised concentrations of 13C-labelled amino acids, and shows similartrends to plot b, in which differences in mole percentages between polychaete tissues and source algae are plotted. Due to the similarity, onlythe latter approach was used in panels c and d, which show the differences between sediment sample and algal compositions for Hediste andArenicola microcosms, respectively. Error bars are ±1 standard deviation.


the latter half of the experiment. When all faunated micro-cosms were considered, there were significant differencesamongst amino acids in the percentage recovered (the per-centage of the total amount of each compound that wasadded in the form of algae to be recovered from faunal tis-sues and sediment) (Kruskal–Wallis DF = 16, P < 0.001).Amino acids such as glycine, and in the case of Hediste

microcosms, leucine, valine and b-alanine, showed rela-

tively high recovery, and thus were either actively produced,retained, or were resistant to degradation or protected bysorptive association with diatom tests (e.g. Nguyen andHarvey, 1997; Thomas and Blair, 2002). In contrast, methi-onine, and to a lesser extent aspartic acid, glutamic acid,tyrosine, lysine and serine showed relatively low recovery,and thus were preferentially metabolised by macrofaunaor microbes.

-25

-15

-5

5

15

25

35

45

55

Diff

in M

ole

% (S

ampl

e-al

gae)

Mole % Difference from Algae(c)

(d)

Hediste 10-d

Hediste 22-d

Hediste Control 10-d


-15

-5

5

15

25

35

45

55

Diff

in M

ole

% (S

ampl

e-al

gae)

Mole % Difference from Algae

Arenicola 9-d

Arenicola 21-d

Arenicola Control 9-d


Aren. Faecal Matter 9-d



Fig 4. (continued)


In general there was no difference in the percentage ofamino acids recovered from microcosms between faunatedmicrocosms and afaunal controls (Mann–Whitney U,P = 0.051–0.999). The only systematic trend was that con-

trols which ran until the end of the experiment (37–38 d)showed generally lower recovery than all other microcosms(significantly for methionine, phenylalanine, glutamic acidand lysine, Mann–Whitney U, P = 0.0413), and this was

0

5

10

15

20

25

30

35

Perc

enta

ge R

ecov

ered

Amino Acid Budgets Hediste

Hediste 10-d

Hediste 22-d

Hediste control 10-d


0

10

20

30

40

50

60

70

% R

ecov

ered

Amino Acid Budgets Arenicola

Arenicola 9-d

Arenicola 21-d

Arenicola control 9-d


(a)

(b)

Fig. 5. The percentage of the total amount of each amino acid added recovered from each microcosm for (a) Hediste microcosms, and (b)Arenicola microcosms.


most likely a function of the longer time available formetabolism/degradation. Thus, differences in recovery be-tween faunated microcosms and afaunal controls may haveemerged if the experiment had run for longer before thepolychaetes were moved into unlabelled sediment. In addi-tion, there was variation between Hediste and Arenicola

microcosms in the average percentage recoveries (Fig. 5),and these differences were significant in the cases of glycine,tyrosine, and c-aminobutyric acid (Mann–Whitney U,P = 0.0304).

4. DISCUSSION

This discussion will focus on the processes which pro-duced the observed bulk 13C and amino acid distributions.It is worth noting that the observations are the net result ofa complex set of simultaneous processes, therefore it is not

always possible to draw firm conclusions on the occurrence/significance of individual mechanisms. It should also benoted that the experiments were conducted under food re-plete conditions, which were judged most relevant forspringtime, when microphytobenthos and phytodetritusare readily available. Had starvation been allowed it mayhave affected polychaete feeding behaviour, and caused adecrease in growth and respiration rates, possibly leadingto a different fate for the labelled algae (Nielsen et al.,1995).

4.1. Bulk C uptake and remineralisation

The d13C values attained by Hediste were greater thanthose attained by Arenicola (Fig. 1), and this is consistentwith the attributes of the taxa. Hediste is a motile, surfacefeeder (Evans, 1971; Ronn et al., 1988), and will have had

-505

101520253035

ALA

GLY

THR

BAL

ASE

RVA

LLE

UIS

O-L

EUIS

O-L

EUM

ETPR

OPH

EG

ABA

ASP

GLU TY

RLY

SO

RN

Mol

e %

Hediste Natural

Labelled

*

*

**

**

** *

**

*

*

-100

1020304050607080

ALA

GLY

THR

BAL

ASE

RVA

LLE

U

MET

PRO

PHE

GAB

AAS

PG

LU TYR

LYS

OR

N

Mol

e %

Arenicola Natural

Labelled

*

*

* * * * * * ** * *

(a)

(b)

Fig. 6. Comparisons of the average natural and 13C-labelled aminoacid compositions of polychaete tissues for (a) Hediste, and (b)Arenicola. * indicates a statistically significant difference betweennatural and labelled mole percentages (Mann–Whitney U,P 6 0.001–0.005). Error bars are ±1 standard deviation.


more rapid access to the algae than the sedentary, head-down feeder Arenicola. In addition, Hediste is a smallerorganism, and can be expected to have a faster growth rate,consistent with its greater labelling levels. Further, by the 9/10 day timepoint, Hediste individuals showed up to �1.5at.% greater labelling level than the surface sediments,implying that they were feeding selectively on the added al-gae. In contrast, Arenicola individuals exhibited a lowerlabelling level than the surface sediment at 9/10 d, and wereonly 13C-enriched compared to surface sediments (by �0.9at.%) later, when sediment enrichment had returned tonear-background (Fig. 1). Thus there is no evidence forArenicola feeding selectively. The difference in C uptakerates and feeding behaviours between the taxa is furtherhighlighted by the fact that, after feeding, 6.7% of the totalC present in the uppermost 1 cm of Arenicola sediments wasderived from the added algae, whereas this value was only4.6% in Hediste sediments. In spite of this, Hediste attainedgreater d13C values. Finally, the data show for both speciesthat 13C is retained in faunal tissues for longer than in thesediment (see later for further discussion).

Before a discussion of the mechanisms driving aminoacid suite variations, the importance of bulk algae, andamino acid remineralisation should be considered. BulkOM remineralisation has previously been shown to con-sume as much as 30–40% of added OM over 20 days insimilar experiments to those presented here (Andersenand Kristensen, 1992). Further, Langenbuch and Portner(2002) used O2 consumption/ammonia production ratiosin annelid tissues to suggest that most or all of its cellularenergy demand was being supplied by amino acid catabo-lism. In addition, carnivorous fish may cover >40% of thetotal energy need with amino acids (Conceicao et al.,2002).

Quantitative budgets for bulk 13C are not available here,but the fact that surface sediment isotopic signatures ap-proached natural background levels after 15–20 days(Fig. 1) suggests that substantial mineralisation of bulk al-gae occurred. Amino acid quantitative budgets show that,across all microcosms, 81–94% of the bulk amino acid pool,and 39–100% of individual amino acids were lost. Further,it is likely that amino acids were preferentially lost com-pared to bulk organic carbon as, despite sustained bulk13C enrichments, the concentrations of labelled amino acidsin faunal tissues decreased between timepoints (Fig. 5).Thus, compound-selective mineralisation of OM must beconsidered as a major driver of the amino acid compositionpatterns discussed below.

4.2. Amino acids

4.2.1. Amino acid behaviour patterns

For the following discussion amino acids have beengrouped, based on similar behaviour.

4.2.1.1. Selective accumulation. Phenylalanine, valine, leu-cine, iso-leucine, threonine and proline (and glycine, see la-ter) showed varying degrees of net accumulation inpolychaete tissues compared to the source algae (Figs. 2–4), and it is therefore suggested that they were selectively

assimilated and/or preferentially retained. Heterotrophicorganisms must obtain certain essential amino acids fromtheir diets, and studies of flour beetles, carpet beetles andhoney bees have shown similar amino acid essentiality torats, dogs, chicks, mice and humans (Meister, 1965). It istherefore likely that the same pattern of essentiality followsfor marine polychaetes. Thus, the observed accumulation ofphenylalanine, valine, leucine, iso-leucine, and threonine isconsistent with the fact that they are all essential (Meister,1965; Phillips, 1984). Selective uptake and/or preferentialretention of essential amino acids is consistent with previ-ous findings that invertebrate growth is not always limitedduring low availability of essential amino acids, possiblydue to compensatory mechanisms which limit their degra-dation (Phillips, 1984, and references therein).

The relative concentrations in which organisms requiredifferent amino acids is often taken to be indicated by theproportions in which they are present in faunal tissues.For example, in fish, protein synthesis and growth are max-imised when the ratios of amino acids supplied closelymatches those in their tissues (e.g. Saavedra et al., 2010).Thus selective assimilation and/or preferential retention isalso indicated by the fact that Hediste contained valine, leu-cine and iso-leucine in greater proportions than those natu-rally present in its tissues (Fig. 6).

-100

-50

0

50

100

150AL

AG

LYTH

RSE

RVA

LLE

UIS

O-LE

UM

ETPR

OPH

EAS

PG

LU TYR

LYS

OR

N

Sedi

men

t Def

icie

ncy

Indx

e Hediste

Arenicola

THR

VAVVL

LEU

ISO-OO

LEU

MET PH

E

200

Fig. 7. Sediment deficiency index, comparing of the natural aminoacid suites of polychaete tissues and their equivalent sediments.Negative sediment deficiency index values indicate a deficiency inthe sediment compared to polychaete requirements. Grey shadingindicates essential amino acids (Phillips, 1984).


Selective assimilation or preferential retention are alsosuggested when polychaete amino acid requirements arecompared with amino acid availability in the sediment.Dauwe and Middelburg (1998), did this by comparing thenatural amino acid compositions of sediment and poly-chaete tissues, and calculated a sediment deficiency indexfor each compound, using Eq. (1).

Sed: Deficiency index ¼ ðSMol% � P Mol%Þ=P Mol% � 100 ð1Þ

SMol% and PMol% are the mole percentages of each (nat-urally present, not 13C-labelled) amino acid in the sedimentand the polychaete tissues, respectively. Negative values de-note amino acids which are ‘deficient’ in the sediment. Thisindex is limited, as it does not account for the fact that onlya small proportion of total sedimentary organic matter istypically labile, and available to organisms (Mayer et al.,1995; Purinton et al., 2008), nor for variations in bioavail-ability between compounds. Nevertheless, it provides aninteresting indication of potential sediment deficiency.

The sediment deficiency index showed that most of theamino acids that accumulated in polychaete tissues weredeficient in the sediments (Fig. 7, although valine, iso-leu-cine and proline were only deficient for Hediste). It followsthat selective assimilation of these essential amino acids byabundant polychaetes could be partly responsible for theirdeficiency in the sediment.

Selective uptake or retention of individual amino acidsmay also be linked to their relative importance in metabolicprocesses. For example, phenylalanine is an essential pre-cursor to the semi-essential tyrosine, which in turn is theprecursor for thyroxine, required for metabolism andgrowth, as well as of adrenaline, and of hormones that reg-ulate the nervous system and response to stress (Saavedraet al., 2010). A complete review of amino acid metabolicrates and pathways is beyond the scope of this article,and the reader is directed instead to Caspi et al. (2010).

In summary, the pattern of net accumulation is consis-tent with amino acid essentiality and natural availability,and may constitute a direct link between faunal gut passageand sediment geochemistry.

4.2.1.2. Rapid loss. A group including serine, methionine,lysine, aspartic and glutamic acids and tyrosine showed de-pleted mole percentage concentrations in polychaete tissues,faecal matter and sediment samples (Fig. 4) compared tothe source algae (especially for serine, aspartic acid andtyrosine, Fig. 2). They also showed particularly large lossesin absolute terms (Fig. 5). Thus, these amino acids weresubject to particularly rapid degradation or metabolism,which may have occurred in polychaete tissues, guts andfaecal matter, as well as in the sediment, through gut orambient microbial activity.

Studies of amino acid catabolism find that non-essentialamino acids are metabolised to a greater extent than essen-tial ones (e.g. Conceicao et al., 2002; Applebaum and Ron-nestad, 2004). Further, studies in rats have shown thatalanine, glutamate and glutamine are preferentially catabo-lised in preference to other non-essential and all essentialamino acids (Tanaka et al., 1995; Conceicao et al., 2002;Applebaum and Ronnestad, 2004). Apart from asparticacid and tyrosine, all the amino acids in the ‘rapid loss’group were essential (Phillips, 1984) and also deficient inthe sediment (Fig. 7). Thus, it seems unlikely that their ra-pid metabolism occurred in poychaete tissues, as essentialamino acids and those in short supply have been observedto be actively retained or protected in invertebrates (Phil-lips, 1984, and references therein).

The observed rapid loss of glutamic acid is potentiallyconsistent with knowledge of the catabolism of glutamate,which is converted to glutamic acid during the analyticalprocedure used here. Applebaum and Ronnestad (2004)noted less marked absorption of glutamate than of otheramino acids by Atlantic halibut larvae, and attributed itto masking of absorption by very rapid metabolism. Gluta-mate metabolism has been observed to exceed that of ala-nine, potentially because it occupies a pivotal position incatabolic pathways, where several other amino acids (gluta-mine, histidine, proline and arginine) are converted to glu-tamate before catabolism (Applebaum and Ronnestad,2004). Further, our observation of rapid loss of lysine isconsistent with the suggestion by Gomez-Requeni et al.(2004), that lysine is rapidly used in protein synthesis orcatabolism by young gilthead sea bream.

Other mechanisms for the rapid loss of certain aminoacids include sloppy feeding and dissolution (Cowie andHedges, 1996). In addition, freeze-dried (lysed) cells wereused in this study; thus, the cell contents were easily acces-sible to digestive and microbial processes. Thomas andBlair (2002) found that lysing cells increased the uptakeof the cell interior amino acids aspartic acid, glutamic acidand phenylalanine by polychaetes, and Cowie and Hedges(1996) explained the loss of aspartic acid, glutamic acidand lysine from zooplankton faecal pellets by their locationin the easily digested cell interior. Thus, this explanationmay also apply to the rapid losses of aspartic acid, glutamicacid and lysine, observed here.

4.2.1.3. Glycine. Glycine has numerous potential functionswithin polychaete tissues, such as being a precursor in thesynthesis of many important biochemicals, including por-phyrins, (precursors to pigments and haemoglobin),


creatine (essential in muscles and nerves), and purine nucle-otides that form part of the backbone of DNA (Lehninger,1982). Further, it functions as a solute in extracellular fluidsused for osmoregulation (Oglesby, 1978), and is concen-trated in the fibrous tissues of polychaetes (Mayer et al.,1995). However, it has also been suggested that glycinehas a relatively low nutritional value (Dauwe and Middel-burg, 1998), and a low energy yield during metabolism.

The most notable alteration observed in amino acidsuites was an enrichment of polychaete tissues, faecal mat-ter and sediments with glycine (Figs. 2 and 4). It also accu-mulated in polychaete tissues in greater proportions thanthose in which it was naturally present (Fig. 6), despitebeing neither essential (it can be synthesised from serine),nor deficient in the sediment (Fig. 7). Further, glycineshowed the highest percentage recovery in quantitative ami-no acid budgets (Fig. 5).

This enrichment of all sample types with 13C-labelledglycine is consistent with previous experiments where poly-chaete tissues (Thomas and Blair, 2002), and polychaeteand zooplankton faecal pellets (Cowie and Hedges, 1996;Thomas and Blair, 2002) have shown high glycine concen-trations in relation to food, and natural faunal tissue.

One explanation for glycine accumulation in faunal tis-sues is its selective assimilation during gut passage. How-ever, simultaneous glycine enrichments in faecal matterand sediments, together with its potentially low nutritionalvalue (Dauwe and Middelburg, 1998) suggest this is unli-kely. Other possibilities include glycine being more resistantto decay than other amino acids. Such resistance could re-sult from protection from enzyme attack due to its closeassociation with the diatom test (Nguyen and Harvey,1997, and references therein). However, Nguyen and Har-vey (1997) also observed glycine enrichment during decayof cyanobacteria, where silica tests are not present. Glycinemay also accumulate in polychaete tissues as a result ofmetabolic process rates. For example, its low nutritional va-lue may mean that other amino acids were preferentiallymetabolised, although this implies that polychaete tissuesretain non-useful residues. Alternately, the inclusion of gly-cine in fibrous tissues may mean that it has a longer resi-dence time in the biomass than other amino acids.

In order for the above mechanisms to fully explain theenrichment of samples with glycine, all other amino acidswould have to be metabolised considerably more rapidly.Such differential metabolism is not suggested by amino acidbudgets, in which glycine percentage recoveries were thegreatest by only a small margin (Fig. 5). Selective preserva-tion would also require glycine to have a considerably longerresidence time than the bulk 13C-labelled algae, in order toproduce the observed increase in lmoles of glycine per mgof 13C compared to the algae (Fig. 4a). Further, metabolicmechanisms could account for glycine enrichment of faunaltissues, but not necessarily of faecal matter and sediment.

Therefore, a further suggestion is made that glycine isproduced in polychaete tissues, digestive tracts, and sedi-ments, from the transformation of other labelled aminoacids. Glycine production requires only the removal ofthe functional R group from other amino acids, and thenumerous metabolic pathways in which glycine is involved

(Caspi et al., 2010) provide multiple opportunities for thisto occur. Glycine production may thus create a dynamicpool of 13C- labelled glycine, although it is unclear whythe labelled dynamic pool is larger than that which is natu-rally present.

The enrichment of faecal matter and sediments with gly-cine suggests that, in addition to production by polychaetedigestive, metabolic, and/or gut-microbial processes, it isalso produced during sediment microbial activity. Thisleads to a further mechanism for glycine accumulation infaunal tissues, whereby bacterial glycine production (possi-bly the result of gardening) is followed by selective inges-tion of bacteria by fauna (Hylleberg, 1975). Veuger et al.(2006) proposed such faunal uptake and retention to ex-plain the longer persistance of glycine in sediments thanthe more refractory peptidoglycan.

Finally, glycine enrichment in sediments and faecal mat-ter may also result from secondary microbial production,and preservation of the persistent, glycine-bearing cellmembrane compound peptidoglycan (Pedersen et al.,2001). Veuger et al. (2006) conducted in situ labelling ofan intertidal microbial community and observed an accu-mulation of labelled glycine in the sediment. However, theynote that this may not have been in the form of peptidogly-can, but rather as simple amino acids within refractory deg-radation products (Pedersen et al., 2001; Veuger et al.,2006).

It is not clear which of the above mechanisms domi-nates, however the occurrence of strong glycine enrich-ments in faunal tissues and faecal matter suggest thatfaunal tissues and digestive tracts are at least importantlocations for the production and/or retention of glycine.

4.2.1.4. Non-protein amino acids. The non-protein aminoacids b-alanine (BALA) and c-aminobutyric (GABA) acidare the decay products of aspartic and glutamic acids(respectively). They originate from microbial activity (Whe-lan, 1977; Vandewiele et al., 2009), and their mole percent-age contribution to the total amino acid pool is an indicatorof advanced OM degradation state (Cowie et al., 1992;Cowie and Hedges, 1994).

In this study, BALA and GABA were present in lowconcentrations in the source algae, due to slight degrada-tion in the non-axenic cultures before harvesting. Theydid not become enriched in faunal tissues compared tothe source algae (Fig. 4a and b), but did become enrichedin some Arenicola microcosm sediments and faecal matter(Fig. 4c and d). Further, Arenicola sediments and faecalmatter showed slight (although not statistically significant,Mann–Whitney U) increases in BALA and GABA betweenthe 9 d and 21 d timepoints (Fig. 4).

As BALA and GABA are produced during microbialdegradation, they might have been expected to show highrecoveries in quantitative amino acid budgets. However,recoveries of both were similar to many protein amino acidsin Hediste microcosms (Fig. 5), and only GABA showedrelatively high recovery in Arenicola microcosms. Thus,microbial production of non-protein amino acids was rela-tively subdued, however this is consistent with their normalassociation with more advanced stages of OM decay than

Hediste Sediments PCA

-5

-4

-3

-2

-1

0

1

2

3

-6 -4 -2 0 2 4 6

PC1

PC2

Faunated sed 9-d

Faunated sed 22-d

Control 9-d

Control 38-dAfaunal Controls

Afaunal Controls

Faunated Sediment

-3

-2

-1

0

1

2

3

4

5

6

-6 -5 -4 -3 -2 -1 0 1 2 3 4 5

PC2

PC1

Arenicola Sediments PCAFaunated sed 10-dFaunated sed 21-dControl10-d Control 37-dAren. Faecal matter 9-dAren. Faecal matter 21-d Aren. Faecal matter 37-d

(b)

(a)

Fig. 8. Results of a PCA including only sediment and faecal matter samples, showing sample scores for (a) Hediste, and (b) Arenicola

sediments. Circles indicate groupings of compositionally similar samples. Note, such groupings are not present for Arenicola sediments.


were observed here (Cowie and Hedges, 1994; Dauwe et al.,1999).

The relatively high concentrations of GABA in Arenico-

la microcosms suggest that Arenicola sediments, or the ani-mals themselves, harbour or stimulate particularly activemicrobial communities. The notable presence of GABA inArenicola faecal matter (Figs. 2 and 4) suggests that a sig-nificant proportion of the microbial activity may occurwithin the Arenicola gut, and/or on freshly egested material,which is consistent with findings that enteric microbes con-tribute particular fatty acids to the faeces of several marineinvertebrates (Bradshaw et al., 1989), and that gut passagealters the bacterial species composition of ingested material(Plante and Mayer, 1994; Plante, 2010). The fact that fau-nated microcosms generally showed higher GABA recover-ies than the equivalent afaunal controls (Fig. 5, notstatistically significant, Mann–Whitney U) also supportsprevious suggestions that fauna stimulate sediment micro-bial activity (Aller and Aller, 1998).

4.2.2. Taxon-specific effects

In several cases Hediste and Arenicola samples displayeddifferent distributions of labelled amino acids. Hediste tis-

sues contained greater mole percentages of labelled valine,lysine, leucine, iso-leucine, methionine, proline, asparticand glutamic acids, and threonine than Arenicola tissues,and lower mole percentages of glycine (Figs. 2 and 3,Mann–Whitney U, P 6 0.001–0.007). In addition, Hediste

showed enrichment in valine, leucine, iso-leucine and pro-line compared to the source algae that Arenicola did not(Fig. 4b). Further, Hediste accumulated valine, leucineand iso-leucine in greater proportions than those naturallypresent in its tissues, and Arenicola did not show this effect(Fig. 6). Their natural amino acids suites were also differ-ent, with Arenicola showing a greater glycine content, andHediste slightly more proline (Fig. 6).

Thomas and Blair (2002) also found different OM alter-ation patterns amongst polychaete taxa as a result of gutpassage, with terebellid faecal pellets showing higher pro-portions of threonine and glycine than those of maldanidsand spionids. Further, maldanid tissues showed greater la-belled glycine mole percentages compared to spionids andterebellids. These differences were attributed to a differencein gut architecture between plug flow reactors (maldanidsand spionids) and mixed reactor plug flow digestors (tere-bellids). In addition they suggested that differences could


further arise from differences in feeding behaviour and par-ticle selection, as well as metabolic processes. Similarly, He-

diste and Arenicola exhibit contrasting feeding behaviours.Hediste is a particle-selective surface deposit feeder, scaven-ger/carnivore and filter feeder, while Arenicola is a head-down deposit feeder with limited particle selectivity (Evans,1971; Ronn et al., 1988; Riisgard, 1991; Jones and Jago,1993; Riisgard and Banta, 1998). They may also have differ-ent anabolism:catabolism ratios, therefore, it is not surpris-ing that they exhibited different patterns of net amino acidaccumulation and loss.

4.3. Impact of faunal gut passage on sedimentary organic

geochemistry

The sustained isotopic enrichment of bulk fauna tissue(Fig. 1) observed here shows that benthic polychaetes area short-term reservoir for freshly deposited C. Also, the dif-ferences in labelled amino acid compositions between fau-nal tissues, food, and faecal matter suggest that faunalingestion and gut passage of OM leads to net accumulationof certain compounds in the biomass. This pathway for or-ganic C through the marine sedimentary system is under-acknowledged by geochemists, but is the prime interest ofbenthic ecologists studying secondary production (Demingand Baross, 1993; Cartes et al., 2011).

It was expected that the impact of gut passage on thepattern of OM decay would show as a difference in sedi-ment 13C-labelled amino acid suites between faunatedmicrocosms and afaunal controls. Thus, PCAs of faunatedand control sediments, plus faecal matter data were con-ducted, one for each taxon (Fig. 8). For Hediste data,PC1 explained 49% of the variation, and PC2 explained17%, and for Arenicola data the values were 54% and15%, respectively. The distinction in PC scores between fau-nated and afaunal sediments was greater for Hediste data,in which the two types of sediment grouped in separateclusters (Fig. 8). It is difficult to determine which aminoacids drove the difference in scores, as each cluster coversa large area, however factor coefficients suggest that faunat-ed sediments showed higher concentrations of alanine,methionine, aspartic and glutamic acids and tyrosine, andlower concentrations of glycine. Faunated and afaunalArenicola sediments did not show clear differences in PCscores, except that some 21-d faunated sediments showedhigher PC1 scores than the controls (Fig. 8). These scoresare attributable to higher GABA and proline concentra-tions, which are associated with microbial activity (Veugeret al., 2006; Vandewiele et al., 2009). Therefore, Arenicola

may have influenced sediment labelled amino acids suitesthrough stimulation of microbial activity. Thus, the evi-dence weakly suggests that the presence of fauna influencedthe suite of 13C-labelled amino acids remaining in the sedi-ment. However, longer experiments are required to identifyfirm compositional differences.

Aspects of the amino acid loss patterns observed hereare consistent with previous observations of OM decay.Most notably, glycine generally becomes a more dominantconstituent of the amino acid suite as degradation pro-gresses (e.g. Dauwe and Middelburg, 1998; Lee et al.,

2000; Veuger et al., in press), possibly due to its associationwith the diatom test (Cowie et al., 1992). Here, a range ofmechanisms for the enrichment of sediments, faunal tissuesand faecal matter with glycine have been discussed, whichsuggest that faunal gut passage and gut bacterial processesmay also contribute towards glycine accumulation duringdecay. In addition, as here, proline was observed to accu-mulate during a recent year-long decay experiment (Veugeret al., in press).

In addition, methionine, glutamic acid, tyrosine, phenyl-alanine, leucine and iso-leucine are often observed to de-grade relatively rapidly in sediments (Cowie et al., 1992;Dauwe and Middelburg, 1998; Lee et al., 2000). For glu-tamic acid, tyrosine, methionine and lysine, this is consis-tent with the preferential losses observed in this study(Figs. 4 and 5). In contrast, phenylalanine was not observedto be rapidly lost, while leucine, and sometimes iso-leucine,were enriched in faunal tissues and sediments (Fig. 4). Fur-ther, serine and aspartic acid have previously been observedto accumulate during OM degradation (Cowie et al., 1992;Dauwe and Middelburg, 1998; Lee et al., 2000), but in thisstudy they were rapidly lost (Figs. 4 and 5).

These inconsistencies with previous observations maypartially result from variations in study site, sediment typeand geochemistry, time scales, OM source, and biologicalcommunity. They also reflect variations in findings amongstprevious amino acid decay studies. For example, in SaanichInlet, Cowie et al. (1992) found that glutamic acid was notparticularly reactive, in contrast to Dauwe and Middel-burg’s (1998) observations in North Sea sediments. Otherstudies have found a lack of change in amino acid suitesduring marked OM degradation in both coastal and deep-sea settings (Henrichs and Farrington, 1987; Cowie et al.,1995; Horsfall and Wolff, 1997), and Suthhof et al. (2000)found that only tyrosine was particularly labile during earlydecay. Moreover, the data presented here focus on veryearly decay of fresh phytodetritus, and thus decay patternsvary from sedimentary environments receiving OM whichhas been degraded for days to months during sinking. Thus,there is still a gap in our understanding of amino acid suitealteration over the whole range of decay timescales. Closingthis gap requires further experiments lasting months toyears, such as those recently conducted by Veuger et al.(in press).

5. CONCLUSIONS

In summary, bulk remineralisation of algal detritus wasan important process, which was probably a major driver ofOM compositional changes. Despite this, polychaetes re-tained 13C in their tissues, and thus faunal biomass repre-sents an understudied route for OM through thesedimentary system. Polychaete tissues displayed net accu-mulation and losses of certain amino acids, the patternsof which tended to align with amino acid essentiality anddeficiency in the sediment. The most notable alteration ofthe labelled amino acid suite was an enrichment of all sam-ple types with glycine. Potential mechanisms for this in-clude selective accumulation through inclusion in tissueswith long residence times, selection against during


metabolism, production from other labelled amino acidsduring metabolic processes, and accumulation in refractoryby-products of secondary bacterial production.

Overall, however, although Arenicola and Hediste arelarge, active and abundant polychaetes, it was difficult tocharacterise their impact on the decay pathways of freshOM as distinct from decay during microbial activity only.It is therefore suggested that, while the finer details of netamino acid accumulation and loss in polychaete tissues ap-pear to be taxon-specific, there is an overall broad similarityin the metabolism of amino acids by polychaetes andbacteria.

ACKNOWLEDGEMENTS

The authors thank Pieter van Rijswijk and Bert Sinke for assis-tance with sample collection and conducting experiments at theNetherlands Institute for Ecology, Centre for Estuarine and Mar-ine Ecology. Also Steve Mowbray for assistance with analyticalwork. They also thank three anonymous reviewers and associateeditor Bob Aller, whose comments led to substantial improvementsto the manuscript. This work was funded through a NERC fellow-ship grant to C. Woulds and supported by the Darwin Centre forBiogeosciences.

REFERENCES

Ahrens M. J., Hertz J., Lamoureux E. M., Lopez G. R., McElroyA. E. and Brownawell B. J. (2001b) The effect of body size ondigestive chemistry and absorption efficiencies of food andsediment-bound organic contaminants in Nereis succinea (Poly-chaeta). J. Exp. Mar. Biol. Ecol. 263, 185–209.

Ahrens M. J., Hertz J., Lamoureux E. M., Lopez G. R., McelroyA. E. and Brownawell B. J. (2001a) The role of digestivesurfactants in determining bioavailability of sediment-boundhydrophobic organic contaminants to 2 deposit-feeding poly-chaetes. Mar. Ecol. Prog. Ser. 212, 145–157.

Aller R. C. (1994) Bioturbation and remineralisation of sedimen-tary organic matter: effects of redox oscillation. Chem. Geol.

114, 331–345.

Aller R. C. and Aller J. Y. (1998) The effect of biogenic irrigationintensity and solute exchange on diagenetic reaction rates inmarine sediments. J. Mar. Res. 56, 905–936.

Amouroux J. M., Gremare A. and Amouroux J. (1989) Modelingof consumption and assimilation in abra-alba (mollusca,bivalvia). Mar. Ecol. Prog. Ser. 51, 87–97.

Amouroux J. M., Cuzon G. and Gremare A. (1997) Association ofpulse chase design, compartmental analysis and analog mod-eling to assess absorption and assimilation efficiencies inPenaeus stylirostris fed an artificial diet. Aquaculture 149, 71–

86.

Andersen F. O. and Kristensen E. (1992) The importance ofbenthic macrofauna in decomposition of microalgae in acoastal marine sediment. Limnol. Oceanogr. 37, 1392–1403.

Applebaum S. L. and Ronnestad I. (2004) Absorption, assimilationand catabolism of individual free amino acids by larval Atlantichalibut (Hippoglossus hippoglossus). Aquaculture 230, 313–322.

Bradshaw S. A., O’hara S. C. M., Corner E. D. S. and Eglinton G.(1989) Assimilation of dietry sterols and faecal contribution oflipids by the marine invertebrates Neomysis integer, Scrobicu-

laria plana and Nereis diversicolor. J. Mar. Biol. Assoc. U.K. 69,

891–911.

Bradshaw S. A., O’hara S. C. M., Corner E. D. S. and Eglinton G.(1990) Dietry lipid changes during herbivory and corophagy bythe marine invertebrate Nereis diversicolor. J. Mar. Biol. Assoc.

U.K. 70, 771–787.

Bradshaw S. A., O’hara S. C. M., Corner E. D. S. and Eglinton G.(1991) Effects on dietry lipids of the marine bivalve Scrobicu-

laria plana feeding in different modes. J. Mar. Biol. Assoc. U.K.

71, 635–653.

Burdige D. J. (2005) Burial of terrestrial organic matter in marinesediments: a re-assessment. Global Biogeochem. Cycles 19,

GB4011, 1–7.

Burdige D. J. (2007) Preservation of organic matter in marinesediments: controls, mechanisms, and an imbalance in sedimentorganic carbon budgets?. Chem. Rev. 107 467–485.

Canfield D. E. (1994) Factors influencing organic carbon preser-vation in marine sediments. Chem. Geol. 114, 315–329.

Cartes J. E., Mamouridis V. and Fanelli E. (2011) Deep-seasuprabenthos assemblages (Crustacea) off the Balearic Islands(western Mediterranean): mesoscale variability in diversity andproduction. J. Sea Res. 65, 340–354.

Casado-Martinez M. C., Smith B. D., Luoma S. N. and RainbowP. S. (2010) Bioaccumulation of arsenic from water andsediment by a deposit-feeding polychaete (Arenicola marina):a biodynamic modelling approach. Aquat. Toxicol. 98, 34–43.

Caspi R., Altman T., Dale J. M., Dreher K., Fulcher C. A., GilhamF., Kaipa P., Karthikeyan A. S., Kothari A., KrummenackerM., Latendresse M., Mueller L. A., Paley S., Popescu L., PujarA., Shearer A. G., Zhang P. and Karp P. D. (2010) TheMetaCyc database of metabolic pathways and enzymes and theBioCyc collection of pathway/genome databases. Nucleic Acids

Res. 38, D473–D479.

Charles F. (1993) Utilization of fresh detritus derived from Cystose-

ira mediterranea and Posidonia oceanica by the deposit-feedingbivalve Abra ovata. J. Exp. Mar. Biol. Ecol. 174, 43–64.

Charles F., Gremare A. and Amouroux J. M. (1995) Utilization ofc-14 formaldehyde to infer ingestion rates and absorptionefficiencies by benthic deposit-feeders. Mar. Ecol. Prog. Ser.

127, 121–129.

Charles F., Gremare A. and Amouroux J. M. (1996) Ingestion ratesand absorption efficiencies of Abra ovata (Mollusca: Bivalvia)fed on macrophytobenthic detritus. Estuar. Coast. Shelf Sci. 42,

83–102.

Chen Z. and Mayer L. M. (1998) Mechanisms of Cu solubilizationduring deposit feeding. Environ. Sci. Technol. 32, 770–775.

Chen Z. and Mayer L. M. (1999) Assessment of sedimentary Cuavailability: a comparison of biomimetic and AVS approaches.Environ. Sci. Technol. 33, 650–652.

Conceicao L. E. C., Ronnestad I. and Tonheim S. K. (2002)Metabolic budgets for lysine and glutamate in unfed herring(Clupea harengus) larvae. Aquaculture 206, 305–312.

Cowie G. L. and Hedges J. I. (1992a) Sources and reactivities ofamino acids in a coastal marine environment. Limnol. Ocea-

nogr. 37, 703–724.

Cowie G. L. and Hedges J. I. (1992b) Improved amino acidquantification in environmental samples: charge-matchedrecovery standards and reduced analysis time. Mar. Chem. 37,

223–238.

Cowie G. L. and Hedges J. I. (1994) Biochemical indicators ofdiagenetic alteration in natural organic matter mixtures. Nature

369, 304–307.

Cowie G. L. and Hedges J. I. (1996) Digestion and alteration of thebiochemical constituents of a diatom (Thalassiosira weissflogii)ingested by a herbivorous zooplankton (Calanus pacificus).Limnol. Oceanogr. 41, 581–594.

Cowie G. L., Hedges J. I. and Calvert S. E. (1992) Sources andrelative reactivities of amino acids, neutral sugars and lignin inan intermittently anoxic marine environment. Geochem. Cos-

mochem. Acta 56, 1963–1978.

Cowie G. L., Hedges J. I., Prahl F. G. and De Lange G. J. (1995)Elemental and major biochemical changes across an oxidation


front in a relict turbidite: an oxygen effect. Geochem. Cosmo-

chem. Acta 59, 33–46.

Dales R. P. (1955) Feeding and digestion in terebellid polychaetes.J. Mar. Biol. Assoc. U.K. 34, 55–79.

Dauwe B. and Middelburg J. J. (1998) Amino acids and hexosam-ines as indicators of organic matter degradation state in NorthSea sediments. Limnol. Oceanogr. 43, 782–798.

Dauwe B., Middelburg J. J., Herman P. M. J. and Heip C. H. R.(1999) Linking diagenetic alteration of amino acids and bulkorganic matter reactivity. Limnol. Oceanogr. 44, 1809–1814.

Defew E. C., Tolhurst T. J. and Paterson D. M. (2002) Site-specificfeatures influence sediment stability of intertidal flats. Hydrol.

Earth Syst. Sci. 6, 971–981.

Deming J. W. and Baross J. A. (1993) Deep-sea smokers – windowsto a subsurface biosphere. Geochim. Cosmochim. Acta 57, 3219–

3230.

Evans S. M. (1971) Behaviour in polychaetes. Q. Rev. Biol. 46, 379–

405.

Fox D. L. (1950) Comparative metabolism of organic detritus byinshore animals. Ecology 31, 100–108.

Francois F., Poggiale J. C., Durbec J. P. and Stora G. (2001) A newmodel of bioturbation for a functional approach to sedimentreworking resulting from macrobenthic communities. In Organ-

ism–Sediment Interactions (eds. J. Y. Aller, S. A. Woodin andR. C. Aller). University of South Carolina Press, pp. 73–86.

Gomez-Requeni P., Mingarro M., Calduch-Giner J. A., Medale F.,Martin S. A. M., Houlihan D. F., Kaushik S. and Perez-Sanchez J. (2004) Protein growth performance, amino acidutilisation and somatotropic axis responsiveness to fish mealreplacement by plant protein sources in gilthead sea bream(Sparus aurata). Aquaculture 232, 493–510.

Gremare A., Amouroux J. M. and Amouroux J. (1989) Modelingof consumption and assimilation in the deposit-feeding poly-chaete Eupolymnia nebulosa. Mar. Ecol. Prog. Ser. 54, 239–248.

Hargrave B. T. (1970) Utilization of benthic microflora by Hyalella

azteca (amphipoda). J. Animal Ecol. 39, 427.

Hartnett H. E., Keil R. G., Hedges J. I. and Devol A. H. (1998)Influence of oxygen exposure time on organic carbon preser-vation in continental margin sediments. Nature 391, 572–574.

Hedges J. I. and Keil R. G. (1995) Sedimentary organic matterpreservation: an assessment and speculative synthesis. Mar.

Chem. 49, 81–115.

Hedges J. I., Sheng Hu F., Devol A. H., Hartnett H. E. andTsmakis E. (1999) Sedimentary organic matter preservation: atest for selective degradation under anoxic conditions. Am. J.

Sci. 299, 529–555.

Henrichs S. M. and Farrington J. W. (1987) Early diagenesis ofamino acids and organic matter in two coastal marinesediments. Geochem. Cosmochem. Acta 51, 1–15.

Herman P. M. J., Middelburg J. J., Widdows J., Lucas C. H. andHeip C. H. R. (2000) Stable isotopes’ as trophic tracers:combining field sampling and manipulative labelling of foodresources for macrobenthos. Mar. Ecol. Prog. Ser. 204, 79–92.

Horsfall I. M. and Wolff G. A. (1997) Hydrolysable amino acids insediments form the Porcupine Abyssal Plain, Northeast Atlan-tic Ocean. Org. Geochem. 26, 311–320.

Hylleberg J. (1975) Selective feeding by Abarenicola pacifica withnotes on Abarenicola vagabunda and a concept of gardening inlugworms. Ophelia 14, 113–137.

Janssen E. M. L., Croteau M.-N., Luoma S. N. and Luthy R. G.(2010) Measurement and modeling of polychlorinated biphenylbioaccumulation from sediment for the marine polychaeteNeanthes arenaceodentata and response to sorbent amendment.Environ. Sci. Technol. 44, 2857–2863.

Jones S. E. and Jago C. F. (1993) In situ assessment of modificationof sediment properties by burrowing invertebrates. Mar. Biol.

115, 133–142.

Kristensen E. (2001) Impact of polychaetes (Nereis spp. andArenicola marina) on carbon biogeochemistry in coastal marinesediments. Geochem. Trans. 2, 92.

Langenbuch M. and Portner H. O. (2002) Changes in metabolicrate and N excretion in the marine invertebrate Sipunculus

nudus under conditions of environmental hypercapnia: identi-fying effective acid–base variables. J. Exp. Biol. 205, 1153–1160.

Lauerman L. M. L., Smoak J. M., Shaw T. J., Moore W. S. andSmith K. L. J. (1997) 234Th and 210Pb evidence for rapidingestion of settling particles by mobile epibenthic megafaunain the abyssal NE Pacific. Limnol. Oceanogr. 42, 589–595.

Lee C. and Cronin C. (1982) The vertical flux of particulate organicnitrogen in the sea – decomposition of amino-acids in the peruupwelling area and the equatorial atlantic. J. Mar. Res. 40, 227–

251.

Lee C., Wakeham S. G. and Hedges J. I. (2000) Composition andflux of particulate amino acids and chloropigments in equato-rial Pacific seawater and sediments. Deep Sea Res. Part I 47,

1535–1568.

Lehninger A. L. (1982) Principles of Biochemistry. Worth, NewYork.

Longbottom M. R. (1970) Distribution of the digestive enzymes inthe gut of Arenicola marina. J. Mar. Biol. Assoc. U.K. 50, 121–

128.

Mayer L. M., Schick L. L., Sawyer T. and Plante C. J. (1995)Bioavailable amino acids in sediments: a biomimetric, kinetics-based approach. Limnol. Oceanogr. 40, 511–520.

Mayer L. M., Schick L. L., Self R. F. L., Jumars P. A., Findlay R.H., Chen Z. and Sampson S. (1997) Digestive environments ofbenthic macroinvertebrate guts: enzymes, surfactants anddissolved organic matter. J. Mar. Res. 55, 785–812.

Mayer L. M., Chen Z., Findlay R. H., Fang J. S., Sampson S., SelfR. F. L., Jumars P. A., Quetel C. and Donard O. F. X. (1996)Bioavailability of sedimentary contaminants subject to deposit-feeder digestion. Environ. Sci. Technol. 30, 2641–2645.

Meister A. (1965) Biochemistry of the Amino Acids, 2nd edition.Academic.

Miller R. J., Smith C. R., Demaster D. J. and Fornes W. L. (2000)Feeding selectivity and rapid particle processing by deep-seamegafaunal deposit feeders: a 234Th tracer approach. J. Mar.

Res. 58, 653–673.

Nguyen R. T. and Harvey H. R. (1997) Protein and amino acidcycling during phytoplankton decomposition in oxic and anoxicwaters. Org. Geochem. 27, 115–128.

Nielsen O. I., Kristensen E. and Holmer M. (2003) Impact ofArenicola marina (Polychaeta) on sediment sulfur dynamics.Aquat. Microb. Ecol. 33, 95–105.

Nielsen A. M., Eriksen N. T., Iversen J. J. L. and Riisgard H. U.(1995) Feeding, growth and respiration in the polychaetesNereis diversicolor (facultative filter-feeder) and n-virens(omnivorous) – a comparative-study. Mar. Ecol. Prog. Ser.

125, 149–158.

Oglesby L. C. (1978) Salt water balance. In Physiology of Annelids

(ed. P. J. Mill). Academic Press, London, pp. 555–658.

Pedersen A. G. U., Thomsen T. R., Lomstein B. A. and JorgensenN. O. G. (2001) Bacterial influence on amino acid enantiomer-ization in a coastal marine sediment. Limnol. Oceanogr. 46,

1358–1369.

Phillips N. W. (1984) Role of different microbes and substrates aspotential suppliers of specific, essential nutrients to marinedetritivores. Bull. Mar. Sci. 35, 283–298.


Plante C. J. (2010) Landscape and smaller-scale effects of lugworm(Arenicola marina) deposit feeding on benthic bacterial assem-blages. J. Mar. Res. 68, 743–765.

Plante C. J. and Mayer L. M. (1994) Distribution and efficiency ofbacteriolysis in the gut of Arenicola marina and 3 additionaldeposit feeders. Mar. Ecol. Prog. Ser. 109, 183–194.

Purinton B. L., DeMaster D. J., Thomas C. J. and Smith C. R.(2008) (14)C as a tracer of labile organic matter in Antarcticbenthic food webs. Deep Sea Res. Part II 55, 2438–2450.

Rasmussen A. D., Banta G. T. and Andersen O. (1998) Effects ofbioturbation by the lugworm Arenicola marina on cadmiumuptake and distribution in sandy sediments. Mar. Ecol. Prog.

Ser. 164, 179–188.

Retraubun A. S. W., Dawson M. and Evans S. M. (1996) The roleof the burrow funnel in feeding processes in the lugwormArenicola marina (L). J. Exp. Mar. Biol. Ecol. 202, 107–118.

Riisgard H. U. (1991) Suspension feeding in the polychaete Nereis-

diversicolor. Mar. Ecol. Prog. Ser. 70, 29–37.

Riisgard H. U. and Banta G. T. (1998) Irrigation and depositfeeding by the lugworm Arenicola marina, characteristics andsecondary effects on the environment. A review of currentknowledge. Vie Milieu 48, 243–257.

Ronn C., Bonsdorff E. and Nelson W. G. (1988) Predation as amechanism of interference within infauna in shallow brackishwater soft bottoms-experiments with an infauna predator,Nereis-diversicolor Muller O.F. J. Exp. Mar. Biol. Ecol. 116,

143–157.

Saavedra M., Conceicao L. E. C., Barr Y., Helland S., Pousao-Ferreira P., Yufera M. and Dinis M. T. (2010) Tyrosine andphenylalanine supplementation on Diplodus sargus larvae: effecton growth and quality. Aquacult. Res. 41, 1523–1532.

Suthhof A., Jennerjahn T. C., Schafer P. and Ittekkot V. (2000)Nature of organic matter in surface sediments from thePakistan continental margin and the deep Arabian Sea: animoacids. Deep Sea Res. Part II 47, 329–351.

Tanaka H., Shibata K., Mori M. and Ogura M. (1995) Metabolismof essential amino-acids in growing rats at graded-levels ofsoybean protein isolate. J. Nutr. Sci. Vitaminol. 41, 433–443.

Thomas C. J. and Blair N. E. (2002) Transport and digestivealteration of uniformly 13C-labelled diatoms in mudflat sedi-ments. J. Mar. Res. 60, 517–535.

Vandewiele S., Cowie G., Soetaert K. and Middelburg J. J. (2009)Amino acid biogeochemistry and organic matter degradation

state across the Pakistan margin oxygen minimum zone. Deep

Sea Res. Part II 56, 376–392.

Veuger B., Van Oevelen D., and Middelburg J. J. (in press) Fate ofmicrobial nitrogen, carbon, hydrolysable amino acids, mono-saccharides, and fatty acids in sediment. Geochem. Cosmochem.

Acta.

Veuger B., van Oevelen D., Boschker H. T. S. and Middelburg J. J.(2006) Fate of peptidoglycan in an intertidal sediment: anin situ C-13-labeling study. Limnol. Oceanogr. 51, 1572–1580.

Volkenborn N., Polerecky L., Wethey D. S. and Woodin S. A.(2010) Oscillatory porewater bioadvection in marine sedimentsinduced by hydraulic activities of Arenicola marina. Limnol.

Oceanogr. 55, 1231–1247.

Voparil I. M. and Mayer L. M. (2000) Dissolution of sedimentarypolycyclic aromatic hydrocarbons into the lugworm’s (Areni-

cola marina) digestive fluids. Environ. Sci. Technol. 34, 1221–

1228.

Voparil I. M. and Mayer L. M. (2004) Commercially availablechemicals that mimic a deposit feeder’s (Arenicola marina)digestive solubilization of lipids. Environ. Sci. Technol. 38,

4334–4339.

Wallace W. G. and Lopez G. R. (1997) Bioavailability ofbiologically sequestered cadmium and the implications of metaldetoxification. Mar. Ecol. Prog. Ser. 147, 149–157.

Wang W. X., Yan Q. L., Fan W. H. and Xu Y. (2002)Bioavailability of sedimentary metals from a contaminatedbay. Mar. Ecol. Prog. Ser. 240, 27–38.

Whelan J. K. (1977) Amino-acids in a surface sediment core ofAtlantic abyssal plain. Geochim. Cosmochim. Acta 41, 803–810.

Woulds C. and Cowie G. L. (2009) Sedimentary pigments on thePakistan margin: controlling factors and organic matterdynamics. Deep Sea Res. Part II 56, 347–357.

Woulds C., Cowie G. and Middelburg J. J. (2010) A method for thequantitative detection of C-13 labelled amino acids and aldosesin marine sediments and fauna. Org. Geochem. 41, 13–21.

Woulds C., Cowie G. L., Levin L. A., Andersson J. H., MiddelburgJ. J., Vandewiele S., Lamont P. A., Larkin K. E., Gooday A. J.,Schumacher S., Whitcraft C., Jeffreys R. M. and Schwartz M.C. (2007) Oxygen as a control on seafloor biological commu-nities and their roles in sedimentary carbon cycling. Limnol.

Oceanogr. 52, 1698–1709.

Associate editor: Robert C. Aller

alteration of organic matter during infaunal polychaete ...alteration of organic matter during...

Documents