The application of δ13C, TOC and C/N geochemistry to … · 2015-10-24 · The application of d13C, TOC and C/N geochemistry to reconstruct Holocene relative sea levels and paleoenvironments

The application of d13C, TOC and C/N geochemistryto reconstruct Holocene relative sea levels andpaleoenvironments in the Thames Estuary, UK

NICOLE S. KHAN,1,2,3* CHRISTOPHER H. VANE,4 BENJAMIN P. HORTON,1,2,5 CAROLINE HILLIER,6 JAMES B. RIDING4

and CHRISTOPHER P. KENDRICK7

1Sea Level Research, Department of Marine and Coastal Sciences, School of Environmental and Biological Sciences, RutgersUniversity, New Brunswick, NJ 08901, USA

2Institute of Earth, Ocean, & Atmospheric Sciences, Rutgers University, New Brunswick, NJ 08901, USA3Department of Earth and Environmental Science, University of Pennsylvania, Philadelphia, PA 19104, USA4British Geological Survey, Environmental Science Centre, Keyworth, Nottingham NG12 5GG, UK5Earth Observatory of Singapore and Asian School of the Environment, Nanyang Technological University, 639798, Singapore6Argus Ecology Ltd, Greencroft Industrial Park, Annfield Plain, Stanley, County, Durham DH9 7XN, UK7NERC Isotope Geoscience Laboratory, British Geological Survey, Environmental Research Centre, Keyworth, Nottingham NG125GG, UK

Received 29 September 2014; Revised 9 April 2015; Accepted 24 April 2015

ABSTRACT: We examined the use of d13C, TOC and C/N geochemistry of sedimentary organic matter toreconstruct former sea levels and paleoenvironments in the absence of suitable microfossil data. The moderndistribution of d13C, TOC and C/N of 33 vegetation and 74 surface sediment samples collected from four coastalwetlands in the Thames Estuary and Norfolk, UK are described. The d13C, TOC and C/N geochemistry of sedimentsvaried in relation to the input of in situ vascular vegetation versus allochthonous particulate organic matter andalgae, which was controlled primarily by tidal inundation. We reviewed published and unpublished studies toproduce an English database of vegetation (n¼ 257) and sediment (n¼ 132) d13C, TOC and C/N geochemistry.Four elevation-dependent environments in the database had statistically distinct d13C, TOC and C/N values: (1)tidal flat/low marsh (d13C: �24.9� 1.2‰; TOC: 3.6� 1.7%; C/N: 9.9� 0.8); (2) middle marsh/high (d13C:�26.2� 1.0‰; TOC: 9.8� 6.7%; C/N: 12.1� 1.8); (3) reed swamp (d13C: �27.9� 0.7‰: TOC: 36.5� 11.5%; C/N: 13.9� 1.2); and (4) fen carr (d13C: �29.0�0.6‰; TOC: 41.6� 5.7%; C/N: 17.4� 3.1). The d13C, TOC and C/N geochemistry database was applied to a Holocene sediment core collected from the Thames Estuary to producethree new sea-level index points and one limiting date, illustrating the utility of d13C, TOC and C/N values toreconstruct Holocene relative sea levels. Copyright © 2015 John Wiley & Sons, Ltd.

KEYWORDS: carbon to nitrogen ratio; Holocene sea level and paleoenvironmental reconstruction; microfossil preservation; stable

carbon isotopes; Thames Estuary.

Introduction

Reconstructions of Holocene relative sea level (RSL) provideimportant constraints for calibrating geophysical models ofEarth’s rheology and glacio-isostatic adjustment (e.g., Lam-beck et al., 1998; Engelhart et al., 2011; Milne and Peros,2013). The ice sheet that formed over the British Isles at theLast Glacial Maximum was small in global terms, but largeenough for glacial-isostacy to produce vastly contrastingpatterns in RSL across the UK during the Holocene (e.g.,Shennan, 1989; Shennan and Horton, 2002; Shennan et al.,2006; Bradley et al., 2011). For example, RSL records fromScotland, situated close to the former ice sheet center, candisplay a mid-Holocene highstand associated with isostaticrebound (e.g., Smith et al., 2002, 2012), while RSL records insoutheast England, including the Thames Estuary, show amonotonic rise throughout the Holocene associated mainlywith subsidence of the proglacial forebulge (e.g., Devoy,1979, 1992; Long, 1992; Sidell et al., 2000).Holocene variations in RSL in the Thames Estuary and

elsewhere in the UK are derived from sea-level index pointsobtained mostly from lithostratigraphic changes betweenterrestrial and marine sediments (transgressive and regressivecontacts). These changes in lithology are corroborated bymicrofossils (e.g., foraminifera, diatoms, pollen/spores), which

are used to delineate the initiation or removal of brackish andmarine conditions and to verify that the contacts are conform-able (Tooley, 1985; Shennan, 1986). However, discrepanciesand data gaps in RSL records still persist that limit interpreta-tion of the driving mechanisms of Holocene RSL dynamics inthe Thames (e.g., Devoy, 1979; Shennan, 1989; Long, 1992,1995; Haggart, 1995; Wilkinson et al., 2000; Sidell et al.,2000; Sidell, 2003) and the UK (e.g., Shennan et al., 2006;Massey et al., 2008; Gehrels, 2010). In part, this may be dueto issues with preservation of identifiable microfossilsin Holocene archives that are associated with microfossillife processes (e.g., Moore et al., 1991; Sen Gupta, 2002;Murray, 2006; Smol and Stoermer, 2010) and post-deposi-tional changes (e.g., Metcalfe et al., 2000; Roberts et al.,2006). Furthermore, the application of microfossil-basedtransfer functions (e.g., Horton et al., 1999), whichhave expanded the type of sediments from which RSLestimates can be derived, have been hindered by problemsassociated with low sample counts and/or a lack of modernanalogues (e.g., Horton and Edwards, 2006). For example, ofthe 225 sea-level index points collected as part of the LandOcean Interaction Study of Holocene coastal evolution ofthe east coast of England (Shennan and Andrews, 2000),only 52 were suitable for transfer function analyses (Hortonet al., 2000).An alternative approach to RSL reconstruction has recently

been explored that utilizes the bulk d13C, TOC and C/Ngeochemistry [stable carbon isotopes (d13C), TOC and

�Correspondence: N. S. Khan, 1Sea Level Research, as above.E-mail: [email protected]

Copyright © 2015 John Wiley & Sons, Ltd.

JOURNAL OF QUATERNARY SCIENCE (2015) ISSN 0267-8179. DOI: 10.1002/jqs.2784

organic carbon to total nitrogen (C/N)] of sedimentary organicmatter (e.g., Wilson et al., 2005a,b; Mackie et al., 2005,2007; Lamb et al., 2006, 2007; Kemp et al., 2010, 2012;Engelhart et al., 2013). d13C and C/N are able to differentiatesources of organic matter that characteristically accumulatein coastal wetland sediments (Haines, 1977; Chmura andAharon, 1995; Go~ni et al., 2000). In particular, d13C and C/Ncan distinguish between freshwater and marine organicmatter (e.g., Fry et al., 1977; Fogel and Cifuentes, 1993) andC3 and C4 vegetation (e.g., Emery et al., 1967; Malamud-Roam and Ingram, 2001). Plants that use the C3 photosynthet-ic pathway have d13C values distinct from those that use theC4 pathway (Smith and Epstein, 1971) due to biochemicalproperties of their primary CO2-fixing enzyme (Deines,1980). In addition, slower diffusion of CO2 in water and thedominant carbonate ion used by plants (HCO3

� vs. dissolvedCO2) enables distinction of terrestrial and freshwater sourcesfrom marine organic matter sources on the basis ofd13C (Fogel et al., 1992). C/N values are used to differentiateaquatic from terrestrial sources because aquatic organicmatter has a greater bulk N content than that of terrestrialorganic matter (Tyson, 1995). TOC measurements quantifythe amount of organic matter contained within sediments(TOC is approximately half the amount of organic mattercontained in a sediment or soil sample; Ostrowska andPorebska, 2012). The amount of organic matter contained insediments has traditionally been estimated using loss-on-ignition (Ball, 1964), although this method may over- orunderestimate the total organic matter and carbon content(Schumacher, 2002; Boyle, 2004); TOC values provide adirect measurement of these quantities (Veres, 2002).In the UK, pioneering research by Andrews et al. (2000)

Wilson et al. (2005a, b), Lamb et al. (2007) and Andrews(2008) described the distribution of d13C and C/N values oftidal flat, salt marsh and fen carr environments. Furtherstudies of the distribution of d13C, TOC and C/N values haveoccurred on the salt-marshes along the US Atlantic (NorthCarolina and New Jersey) and Pacific (Oregon) coasts (Kempet al., 2010, 2012; Engelhart et al., 2013; Milker et al., 2015),in each of these studies, elevation-dependent environmentswere identified with d13C, TOC and C/N, suggesting that thistechnique has potential for reconstructing sea level. Expan-sion upon these studies is needed to (i) examine the boundarybetween salt marsh and inter- to supra-tidal reed swamp orfen carr environments, which is important in the interpreta-tion of transgressive/regressive contacts in Holocene sequen-ces in the UK; (ii) explore regional, inter- or intra-sitevariability in d13C, TOC and C/N distributions; and (iii) assessbias introduced by using modern d13C, TOC and C/N valuesto interpret Holocene sequences due to post-depositionaldegradation of organic matter, a factor deemed problematicto reconstruction studies (Andrews et al., 2000; Wilson et al.,2005a,b; Lamb et al., 2007; Kemp et al., 2010, 2012;Engelhart et al., 2013).Here, we examine the distribution of d13C, TOC and C/N

of vegetation and sedimentary organic matter within coastalwetlands of the Thames Estuary and Norfolk Broads, UK. Wecompare our data with published and unpublished studiesfrom northern and southern England to produce a database ofd13C, TOC and C/N values. We apply the database to aHolocene sediment core collected from Swanscombe Marshon the Thames Estuary with poor microfossil preservation toproduce new sea-level data. We find the effect of post-depositional processes on d13C, TOC and C/N is not prohibi-tive in paleoenvironmental interpretation, and thus thismethod can be used to reconstruct Holocene RSL in the UKand other temperate regions.

Study area

The modern Thames Estuary drains an area of �16000 km2

of England into the North Sea (Fig. 1). The tidal portion of theestuary is approximately 110 km long, with the seaward limitextending �80 km downstream and the tidal limit occurringapproximately �30 km upstream from London Bridge (Mitch-ell et al., 2012). The estuary is macrotidal with spring tidalrange varying from 5.2m near the mouth of the estuary inSheerness to 6.6m at London Bridge (Admirality Tide Tables,2013). Salinity ranges from <1 at the tidal limit to �32 at theestuary mouth at half-tide (time or state halfway betweenflood and ebb) (Juggins, 1992), although large variationsoccur during the tidal cycle as well as seasonally (Mitchellet al., 2012).Key to the accurate and reliable reconstruction of Holo-

cene sea levels and paleoenvironments is finding modernenvironments that reflect those that are found in the past.Undisturbed areas of coastal wetland habitat are limited inthe region because of human modifications, changes in wave

Figure 1. Location map showing study areas in the UK and alongthe Yare River and Thames Estuary and location of transects(black dotted line) at study sites in Dartford Creek (A), Wat TylerVountry Park Nature Reserve (B), Two Tree Island (C) and TedEllis Reserve (D).

Copyright © 2015 John Wiley & Sons, Ltd. J. Quaternary Sci., Vol. 9999(9999) 1–17 (2015)

JOURNAL OF QUATERNARY SCIENCE

and wind climate (van der Wal and Pye, 2004) and RSL rise(Woodworth et al., 2009). These anthropogenic and climate-driven changes pose difficulties in locating suitable areasalong the estuary to use as an analogue for Holocene saltmarsh, reed swamp and fen carr paleoenvironments (Devoy,1979). We tried to minimize these factors by confining ourstudy sites to nature reserves or maintained marshes wherewell-developed successions of wetland floral zones werereadily identifiable. Three sites were chosen on the ThamesEstuary (Dartford Creek, Wat Tyler Country Park and TwoTree Island) and one site containing fen carr in the NorfolkBroads (Ted Ellis Nature Reserve). The Norfolk Broads waschosen because no suitable contemporary fen carr environ-ments exist on the modern Thames (Binney et al., 2005;Waller et al., 2005; Waller and Grant, 2012). The speciescompositions of marsh floral zones in our sites are representa-tive of the low marsh (van der Wal and Pye, 2004) and highmarsh (Boorman, 2003) communities that are characteristic ofsoutheast England. With the exception of introduced C4

Spartina (Ranwell, 1972), all other vegetation follows the C3

photosynthetic pathway. Fen carr environments occur inwaterlogged conditions (McVean, 1956) and commonlydevelop because of increasing tidal influence or elevatedgroundwater level associated with sea-level rise (e.g., Walker,1970; Long and Innes, 1995; Waller et al., 2005).The study site situated at Dartford Creek, a tributary of the

Thames Estuary, is part of the Crawley marshes (Fig. 1A).Porewater salinity at the site ranged from 26 to 32. Onetransect (A–A0; Fig. 1A) of 18 stations was established thatextended through unvegetated tidal flat and low, middle, tohigh marsh floral zones. The low marsh was inhabited byPuccinellia maritima, Spergularia spp., and in localizeddepressions Eleocharis sp. The middle–high marsh wasdominated by Elymus repens and Festuca rubrum, withSpergularia spp. present. The transition to upland vegetationwas absent from this site due to levee construction.The Wat Tyler Country Park nature reserve is located north

of Canvey Island in Pitsea. One �40m transect (B–B0; Fig. 1B)of 15 stations was sampled that encompassed a full range ofsuccessional environments from tidal flat, low, middle to highmarsh, and brackish, upland transition. Porewater salinity atthe site ranged from 27 to 34. Tidal flat sediments wereunvegetated. Low marsh vegetation included Aster tripoliumand Salicornia europaea. The middle marsh was inhabited byHalimione portulacoides and Puccinellia maritima. The highmarsh was occupied by Festuca rubra and Halimioneportulacoides, and a brackish transition zone occupied byPhragmites australis occurred at the upper boundary of themarsh.The final study area from the Thames Estuary is located

in the marshes of Two Tree Island, east of Canvey Island,forming part of the Leigh National Nature Reserve. Pore-water salinity at this site was between 34 and 37. Twotransects (C–C0; D–D0; Fig. 1C) of 11 and six samplingstations, respectively, were established to account forvariability between low marsh environments colonized bythe introduced C4 grass Spartina anglica. The occurrence ofthis species will enable distinction among low, high andupland transition environments due to its discrete valuesfrom C3 vegetation, but the species is not represented inHolocene sedimentary archives (Long et al., 1999). Bothtransects extend through tidal flat, low marsh and middlemarsh zones. The low to pioneer marsh hosts Salicorniaeuropaea and Spartina anglica. The middle marsh isdominated by Festuca rubra, Halimione portulacoides, Astertripolium, Borrichia fructescens and Suaeda maritima.Washed-in algae and seaweed (Fucus vesiculosus) were

present throughout the marsh. Transect 2 (D–D0) waspositioned to avoid sampling sediments occupied by Sparti-na. Leveeing further inland prevented the formation of highmarsh, brackish transitional and upland communities.The Ted Ellis Nature Reserve, situated southeast of Norwich

along the southern bank of the River Yare in the NorfolkBroads, contains a variety of wetland habitats. The local reliefat this site represents a small elevation gradient. Porewatersalinity at the site is <0.5. Two transects were sampled fromthis site (Fig. 1D). The first transect (E–E0) was 215m longwith 15 stations and extended from a reed swamp dominatedby Phragmites australis and Phalaris arundinacea with Carexspp. present at the banks of the Yare to a Salix-dominatedfen carr environment. The second transect (F–F0) was 70mlong with nine stations and incorporated reed swamp andAlnus-dominated fen carr environments.

Methods

Stations were positioned to ensure sufficient sampling fromeach vegetation zone and to cover the full range in elevationof vegetation zones encountered at each site studied. Stationsand core locations were surveyed to a common referencedatum (m Ordnance Datum; OD) using a total station and aLeica differential geographical positioning system with real-time kinematic capabilities. Tidal datums at all ThamesEstuary sites were interpolated from the nearest tide gaugestations (Admiralty Tide Tables, 2013). Tidal datums at TedEllis Reserve were inferred from stream gauge readingsmaintained by the Environment Agency.

Modern vegetation and sediment

The vegetation cover at each sampling station was recorded(percent cover of vegetation species was estimated from a�0.5-m2 area surrounding each sampling point) and thedominant vegetation (above- and below-ground components)was sampled (n¼ 33) to provide context for the d13C and C/Nof surface sediments (Chmura and Aharon, 1995; Malamud-Roam and Ingram, 2004). Seventy-four 10�1-cm2 surfacesamples were collected for analysis of d13C, TOC and C/N ateach sampling station. Salinity was measured using a calibrat-ed refractometer at the time of sample collection. When thesample was not wet enough for the measurements, porewaterwas separated by centrifuge from the samples in the laborato-ry and its salinity was measured (Horton et al., 1999; Sawaiet al., 2004).

Collection of Core SW1

During a drilling campaign conducted by the British Geologi-cal Survey, a series of sediment cores from locations in themid- and lower estuary were collected. One core fromSwanscombe Marsh (Core SW1) was selected for analysisbecause it represented the full series of transgressive/regres-sive sequences recognized by Devoy (1979) in his type-site atTilbury. A drilling rig employing a ‘wireline’ percussion tripodarrangement (also known as ‘shell and auger’) was used toretrieve Core SW1. The recovered core was capped andsealed in the field and immediately refrigerated at 4 ˚C. Thecore was described in the laboratory using the Troels-Smith(1955) method for organic-rich sediments.

d13C, TOC and C/N geochemistry of modern andcore samples

Plant samples were treated with 5% HCl for 2 h, rinsed withdeionized water, dried in an oven at 40 ˚C and freezer-milledto a fine powder. Sediment samples were treated with 5%


HOLOCENE RELATIVE SEA LEVELS AND PALEOENVIRONMENTS IN THE THAMES ESTUARY

HCl overnight to remove inorganic carbon, and weresubsequently rinsed with at least 1500mL of deionized water,dried in an oven at 40 ˚C and milled to a fine powder using apestle and mortar. 13C/12C analyses were performed bycombustion in a Costech Elemental Analyzer coupled onlineto an Optima dual-inlet mass spectrometer at the NERCIsotope Geosciences Laboratory, Nottingham, UK. The valueswere calibrated to the Vienna Pee Dee Belemnite (VPDB)scale using within-run cellulose standard Sigma Chemical C-6413 calibrated against NBS19 and NBS22 that were includ-ed within the runs. Sample total organic C and total N weremeasured on the same instrument. C/N ratios were calibratedwith an acetanilide standard and are given as a weightpercentage. Replicate analysis on well-mixed samples indi-cates analytical precision of <0.1‰. Core SW1 was analysedfor d13C, TOC and C/N at 8-cm intervals continuouslythroughout the core. The d13C, TOC and C/N of modernvegetation and bulk sediments (including the databases), andCore SW1 data can be found in supplementary Tables S1 andS2, and Appendix S1, respectively.

Microfossil analysis of Core SW1

Microfossil (diatom, foraminifera, pollen/spore) analysis wasundertaken on Core SW1 to support, where possible,paleoenvironmental changes inferred from d13C, TOC andC/N. Samples for diatom and foraminiferal analysis weresampled above and below radiocarbon-dated contacts ofCore SW1; pollen/spore counts were performed only on thesediment at dated horizons. All samples for diatom analysisunder light microscopy were prepared following standardmethods (Zong and Horton, 1998, 1999). Diatoms wereidentified and enumerated under 1000� magnificationusing the keys of Hartley et al. (1986), van de Werff andHuls (1958–1966) and Patrick and Reimer (1966, 1975).Classification of salinity and life form follows Denys (1991/1992) and Vos and de Wolf (1993). Sample preparation,identification and classification of foraminifera followedHorton and Edwards (2006). Wet counts were completedunder a binocular microscope. A minimum of 200 diatomand foraminifera were counted per sample where possible.Clay-rich sediments were prepared for pollen and sporesfollowing the technique of Riding and Kyffin-Hughes (2004)and peat samples were prepared by disaggregation withpotassium hydroxide. Pollen and spores were grouped intofive broad physiognomic categories: trees, shrubs, herbs,aquatics and pteridophytes. Calculation of individual taxa isexpressed as percentage of the total sum of land pollen/spores. The microfossil data from Core SW1 can be foundin Appendices S2–4.

Radiocarbon age determination

Radiocarbon (14C) dates were selected to produce newsea-level data from Core SW1 (Table 1). Samples weresent to Pozna�n Radiocarbon Laboratory for 14C analysis.Three identifiable plant macrofossils inferred to be depos-ited in situ were selected for accelerator mass spectrome-try radiocarbon dating. Before analysis, the samples werecleaned under a binocular microscope to remove contam-inating material, such as older adhered organic sedimentfrom the matrix surrounding the macrofossil or youngeringrown rootlets (Kemp et al., 2013a). In the absence ofdatable macrofossil remains, one date was obtained frombulk peat substrate at �7.14m OD. Reported radiocarbonages were calibrated to sidereal years with a 2s confi-dence interval using the IntCal13 calibration curve

(Reimer, 2013) and a laboratory multiplier of 1. Ages arepresented as calibrated years before present (cal a BP),where the zero point is AD 1950 (Stuiver and Polach,1977).

Statistical analysis

Analysis of variance (ANOVA) was performed on the sixtransects from the Thames and Norfolk Broads study areas(Table 2) and the sediment database of England (Table 3)to identify significant differences in sediment d13C, TOC andC/N values among elevation-dependent environments. One-way ANOVA fits means across a grouping variable and testsif they are significantly different (Field, 2007). Nested(hierarchical) ANOVA tests whether there is significant varia-tion in means among groups and among subgroups withingroups (Sokal and Rohlf, 1969).One-way and nested ANOVA was performed on the six

transects to identify inter-site variability in the d13C, TOC andC/N values (Table 2). ‘Environment’ was defined as the maingrouping factor using the floral zones present in the studyareas: tidal flat/low salt marsh, middle/high marsh, reedswamp and fen carr. ‘Site’ (e.g., Two Tree Island: TT; DartfordCreek: DC; Wat Tyler: WT; Ted Ellis: TE) was defined as thenested (sub-grouping) factor. Environment (floral zone) wasused as the grouping factor because the primary contributionto sediment geochemical values is the in situ vegetation coverand the incorporation of allochthonous organic matter sour-ces, which is also influenced by vegetation cover (Lambet al., 2006; Li and Yang, 2009). Furthermore, the use ofpaleoenvironments to infer the past position of sea level hasbeen established by International Geological CorrelationProjects (IGCP) 61, 200, 495 and 588 (e.g., Preuss, 1979;Tooley, 1982; van de Plassche, 1982; Shennan, 1987;Gehrels and Long, 2007; Horton et al., 2009; Shennan et al.,2015).One-way ANOVA was performed on the sediment data-

base using the same environments defined above as thegrouping factor (excluding 23 samples occupied by intro-duced C4 Spartina and 11 samples from minerogenic aldercarr environments in Kent, which do not serve as a suitableanalogue to the organic-rich peat beds of the HoloceneThames) to detect significant differences in mean d13C, TOCand C/N values of depositional environments (Table 3).The F ratio (the ratio of the sum of squares of (i) the

differences between each value and its group mean to (ii)the differences between the group means and the mean ofall values in all groups) and P values (probability > F) werecomputed to determine whether differences (i.e. a signifi-cant effect) occur among environments/sites of the moderntransects and sediment database (Sokal and Rohlf, 1969).The null hypothesis (i.e. there are no differences betweenenvironments/sites) can be rejected for high F values (>1)and low P values.The post-hoc Tukey’s Honestly Significant Difference test

(Tukey, 1953; Kramer, 1956) was applied to identify differ-ences among multiple means when a significant effect wasfound (Abdi and Williams, 2010). Significant differencesamong environments/sites are identified in Tables 2 and 3by different letters (i.e. environments/sites connected by thesame letter are statistically indistinct). For example, inTable 2, on the basis of C/N, the tidal flat/low marsh (A)and mid/high marsh (A) of Two Tree Island and tidal flat/low marsh of Dartford Creek (A) are statistically indistinctfrom one another and the tidal flat/low marsh of Wat Tyler(AB), but significantly different from the mid/high marsh ofDartford Creek (BC).



Statistical analysis was completed in JMP 10.0, and datawere log-transformed where necessary to meet assumptionsof ANOVA (equal variance, normality).

Results

Characteristics of modern vegetation

Thirty-three vegetation samples from tidal flat, low, middleand high marsh, and reed swamp environments were ana-lysed for d13C and C/N composition. The mean d13C ofall samples was �25.4‰, ranging from �32.1 to �13.1‰.The mean C/N of all samples was 38.2, spanning values of6.3–122.4.Marine and tidal flat end-member vegetation samples

(n¼ 2), including brown algae and Fucus vesiculosis, rangedfrom �32.1 to �20.3‰ in d13C and from 6.3 to 14.6 in C/N.Vegetation end-members collected from low marsh

environments (n¼13), including Aster tripolium, Borrichiafructescens, Cochleria spp., Puccinellia maritima, Salicorniaeuropaea, Spartina anglica and Sueada maritima, had meand13C and C/N values of �24.6‰ and 30.5, respectively.d13C values ranged from –30.0 (Salicornia europaea) to�13.1‰ (Spartina anglica), and C/N spanned values from12.0 (Sueada maritima) to 61.9 (Borrichia fructescens).Middle to high marsh end-member vegetation (n¼ 14),

including Agrostis stolonifera, Elymus repens, Festuca rubra,Halimione portulacoides, Scirpus maritimus and Spergulariamedia, had a mean d13C of �26.0‰, which variedbetween �28.6 (Halimione portulacoides) and �24.1‰(Festuca rubra). Mean C/N was 48.8, with values rangingfrom 15.5 (Halimione portulacoides) to 122.5 (Agrostisstolonifera).Vegetation end-members collected from reed swamp envi-

ronments (n¼ 3), including Phragmites australis and Carexsp., had mean d13C and C/N values of �25.3‰ and 41.3,

respectively. d13C fell between �26.5 (Carex sp.) and�24.6‰ (Phragmites australis), and C/N values ranged from30.6 (Carex sp.) to 61.7 (Phragmites australis).

Characteristics of modern sediments

Seventy-four surface sediment samples were analysed ford13C, TOC and C/N composition. The mean d13C of allsamples was �26.1‰, ranging from �29.6 to �19.5‰. Themean TOC of all samples was 18.7%, spanning values from0.6 to 47.6%. The mean C/N of all samples was 12.2,extending from 8.0 to 26.4.

Dartford Creek transect

The Dartford Creek transect (Fig. 2) covered tidal flat, low,middle and high marsh zones from an elevation of 1.63 to3.40m OD. The tidal flat/low marsh zone (n¼7) had d13C,TOC and C/N values of �25.9 to �25.1‰, 3.0–3.9% and8.8–9.7, respectively. d13C decreased with distance landwardin the middle–high marsh zone (n¼11) from –25.9‰ atthe boundary with the low marsh to �27.7‰ at the edge ofthe high marsh. TOC values increased from the middle tohigh marsh from 7.0 to 18.0%. C/N values increased withdistance along transect in the middle–high marsh with aminimum of 11.5 at �12m to a maximum of 14.1 at �37malong the transect.

Wat Tyler transect

The transect at Wat Tyler (Fig. 3) incorporated low, middleand high marsh and Phragmites brackish transition environ-ments ranging in elevation from 1.73 to 3.22m OD. The tidalflat/low marsh zone (n¼ 8) had d13C values between �26.7and �25.2‰. TOC values in this zone increased from 3.3 to6.9% at 26m along the transect in the low marsh. C/N valuesexhibited a similar pattern to TOC; C/N increased in the

Table 1. Summary of intercalated sea-level index points and limiting data from Core SW1 (51.462842˚N, 0.306567˚E) in the lower ThamesEstuary. Reference water level (RWL), mean high water spring tide (MHWST), mean tide level (MTL) and highest astronomical tide (HAT).

Lab. code

Poz-19868 Poz-19777 Poz-19778 Poz-19779

Sample altitude (m OD) �4.76 �7.14 �9.56 �9.7114C age� 1s (a BP) 3670�35 5720� 35 6810�40 6880� 40Calibrated age range (cal a

BP)4138–3896 6573–6412 7700–7580 7817–7620

Contact type Transgressive Regressive Transgressive Terrestrial limitingMaterial dated Phragmites (directly below

saltmarsh deposit)Monocot peat (directly

above saltmarsh deposit)Phragmites (directly belowsaltmarsh deposit)

Fen wood peat (freshwaterlimiting)

Microfossils Foraminifera: switch from100% agglutinatedassemblage to calcareousdominant

Foraminifera: nopreservation



Pollen: ferns dominant,trees and herbssubdominant

Pollen: ferns dominant,trees minor

Pollen: trees dominant,herbaceous vegetation,including Chenopodium

subdominant

Pollen: ferns dominant,trees minor

Diatoms: increase in salinitypreference

Diatoms: poor preservation,polyhalobous tomesohalobous

Diatoms: no preservation Diatoms: no preservation

d13C, TOC, C/N d13C: increase from �27.8to �25.8‰

d13C: decrease from �26.3to �28.0‰

d13C: increase from �29.7to �26.8‰

d13C: decrease from �28.1to �28.7‰

TOC: decrease from 37.4 to3.3%

TOC: increase from 5.5 to42.3%

TOC: decrease from 48.5 to5.9%

TOC: increase from 24.1 to48.8%

C/N: decrease from 16.5 to13.2

C/N: increase from 13.9 to25.9

C/N: decrease from 24.8 to16.0

C/N: increase from 20.4 to26.7

Indicative meaning MHWST�20 cm (MHWSTþ HAT)/2–20 cm MHWST�20 cm >MTL



pioneer to low marsh from 10.1 to 11.4. The d13C values ofthe middle/high marsh zone (n¼4) are lower than the tidalflat/low marsh zone, ranging from �26.9 to �26.3‰. TOCvalues increased in the middle/high marsh from 15.5% at28m to 24.0% at 36m along the transect. C/N values alsogenerally increased in this zone from a minimum of 11.9 to amaximum of 14.6 at 32m. d13C values in the Phragmitesenvironment (n¼ 3) varied between �28.0 and �26.1‰.TOC values increased from 28.8% at the boundary betweenmiddle and high marsh to 33.4% at the landward edge of thetransect.

Two Tree transects 1 and 2

Two Tree transect 1 incorporates tidal flat and low andmiddle marsh environments (Fig. 4), with undulating topogra-phy due to dissecting tidal creeks and drainage ditches. Thetransect extends from 2.00 to 2.80m OD in elevation. Thetidal flat/low marsh zone (n¼ 6) has mean d13C, TOC and C/N values of �20.8‰, 2.6% and 8.5, respectively. Comparedwith the tidal flat/low marsh zone, the middle marsh zone

(n¼5) had a lower mean d13C (�23.1‰), higher mean TOC(3.7%) and higher mean C/N (9.4).Two Tree transect 2 also included tidal flat, low and middle

marsh environments, but lacked much of the undulatingtopography of transect 1 (Fig. 4). The transect ranged from2.13 to 2.79m OD. d13C values increased with distancealong the transect from a minimum of �22.0‰ in the tidalflat/low marsh (n¼ 2) to a maximum of �25.1‰ at �7malong the transect in the middle marsh (n¼ 4). TOC valuesexhibited a similar pattern, with a minimum TOC of 1.6% inthe tidal flat/low marsh, increasing to a maximum value of5.6% at the landward edge of the transect in the middlemarsh. C/N values also increased with distance along thetransect from a minimum of 9.1 in the tidal flat/low marsh toa maximum of 10.7 at the landward edge in the middlemarsh.

Ted Ellis transects 1 and 2

Ted Ellis transect 1 (Fig. 5) extends from the vegetated banksof the River Yare, across a reed swamp to a fen carr.

Table 2. Summary of statistical analyses of sediment d13C, TOC and C/N values from the six modern transects presented in this study. One-wayANOVA with environment (tidal flat/low marsh, mid/high marsh, reed swamp, fen carr) as the grouping factor and nested ANOVA with site (TwoTree: TT; Dartfort Creek: DC; Wat Tyler: WT; Ted Ellis: TE) as nested factor was used to analyse d13C, TOC, and C/N data to examine inter-sitevariability. Group means� SD are listed. Significant differences among means for the main effect of environment and nested effect of site (Tukey’sHSD) are indicated by different letters (i.e. environments/sites connected by the same letter are statistically indistinct) (see text for furtherexplanation).

Zone n d13C (‰) TOC (%) C/N

Tidal flat/low marsh 23 �23.9�2.3 A 3.6�1.6 A 9.6�1.0 AMid/high marsh 24 �25.6�1.7 B 8.7�6.1 B 11.4�1.8 BReed swamp 19 �27.9�0.7 C 37.9�9.8 C 13.9�1.2 CFen carr 8 �29.0�0.5 D 46.3�1.0 D 18.2�3.8 D

ANOVAF ratio 164.5 170.4 67.8Probability> F <0.0001 <0.0001 <0.0001

Site ZoneTT Tidal flat/low marsh 8 �21.1�0.9 A 2.4�1.3 A 8.7�0.6 A

Mid/high marsh 9 �23.6�0.9 B 3.8�1.3 A 9.4�0.7 ADC Tidal flat/low marsh 7 �25.4�0.2 C 3.4�0.3 A 3.4�0.3 A

Mid/high marsh 11 �26.9�0.6 D 8.9�4.2 AB 12.4�0.9 BCWT Tidal flat/low marsh 8 �25.6�0.5 C 5.0�1.6 A 10.8�0.4 AB

Mid/high marsh 4 �26.8�0.3 D 18.9�4.1 BC 13.1�1.3 CDBrackish transition/Reed swamp 3 �26.7�1.1 D 31.8�2.6 CD 13.8�2.2 CD

TE Reed swamp 16 �28.1�0.3 E 39.1�10.3 DE 13.8�1.1 CFen carr 8 �29.0�0.5 F 46.3�1.0 E 18.2�3.8 D

ANOVAF ratio 91.8 5.4 6.3Probability> F <0.0001 0.0003 <0.0001

Table 3. Summary of statistical analyses of sediment d13C, TOC and C/N values from the sediment database. One-way ANOVA withenvironment (tidal flat/low marsh, mid/high marsh, reed swamp, fen carr) as the grouping factor on the northern/southern England database wasused to analyse d13C, TOC and C/N data. Samples from C4 vegetation zones (n¼23) and minerogenic fen carr environments (n¼11) wereexcluded from analysis (see text for explanation). Group means� SD are listed. Significant differences among means for the main effect ofenvironment (Tukey’s HSD) are indicated by different letters (i.e. environments connected by the same letter are statistically indistinct) (see text forfurther explanation).

Zone n d13C (‰) TOC (%) C/N

Tidal flat/low marsh 20 �24.9�1.2 A 3.6�1.7 A 9.6�1.0 AMid/high marsh 40 �26.2�1.0 B 9.8�6.7 B 12.1�1.8 BReed swamp 20 �27.9�0.7 C 36.5�11.5 C 13.9�1.2 CFen carr 18 �29.0�0.6 D 41.6�5.7 C 17.4�3.1 D

ANOVAF ratio 67.2 140.4 67.2Probability> F <0.0001 <0.0001 <0.0001



Sampling stations cover elevations from 0.05 to 0.78m OD.The reed swamp (n¼10) had a mean d13C of �28.1‰. Thebanks and river levee showed the greatest variability withinthe reed swamp with values between �28.7 and �27.7‰.Mean TOC of the reed swamp was 34.9% with the minimumof transect 1 of 18.3% occurring along the riverbanks. TOCincreased with distance inland, reaching a maximum of47.2% at �160m along the transect. C/N values withinthe reed swamp had a mean of 13.3. The fen carr (n¼ 5) hada mean d13C of �28.8‰, which increased from the transectminimum of �29.5 to �28.2‰ at the landward edge ofthe fen carr. The mean TOC within the fen carr was46.1%, including the maximum observed on transect 1 of47.6%. C/N increased within the fen carr from 13.4 to themaximum of transect 1 of 26.4 at its landward edge.Ted Ellis transect 2 is located further inland from the River

Yare than transect 1 and therefore was found at a slightlyhigher elevation of 0.45–1.34m OD (Fig. 5). Transect 2covered reed swamp and fen carr environments. The reedswamp (n¼ 6) had a mean d13C of �28.2‰ and a meanTOC value of 46.1. C/N values decreased with distance alongthe transect from 16 at the beginning of the transect to 13.5(the minimum of transect 2) at the boundary with the fen carr.At the boundary between the fen carr (n¼ 3) and reedswamp, d13C was at a minimum of �29.6‰ and increasedwith distance inland to �28.8‰. The TOC mean was 46.6%in the fen carr. There was an increase in C/N from values ofthe reed swamp with a mean of 18.3.

Discussion

d13C and C/N characteristics of vegetation fromcoastal environments

Understanding the range and variability in d13C and C/N ofend-member vegetation are fundamental to the interpretationof d13C and C/N geochemistry from bulk sediments (Chmuraet al., 1987; Chmura and Aharon, 1995; Malamud-Roam andIngram, 2001). We combined the measured d13C and C/N

values from three sites on the Thames Estuary and one sitecontaining fen carr in the Norfolk Broads with vegetationsampled from: (i) salt marshes of Kent (Andrews, 2008),Severn Estuary (Allen et al., 2007), Humber Estuary (Lambet al., 2007) and Mersey Estuary (Wilson et al., 2005a,b); (ii)fen carr environments in Kent, East Sussex and the NorfolkBroads (Andrews, 2008); and (iii) particulate organic matterof the Thames and Severn estuaries (Allen et al., 2007;Bristow et al., 2013) to create a regional database (n¼ 257)from northern and southern England (Table S1). This vegeta-tion database serves as a guide to understand the sources oforganic matter that contribute to sediment geochemicalvalues in each environment type.We find distinctions in d13C and C/N values of aquatic,

C3 salt marsh, C4 salt marsh, and leaf and wood tissuefrom fen carr vegetation within the vegetation database(Fig. 6A). Algae displayed a wide range in d13C values of�17.8� 7.8‰. This wide variation in d13C values reflects thesalinity (and thus pH and dissolved CO2) of the environmentin which it was formed (Benedict et al., 1980; Keeley andSandquist, 1992). Freshwater algae d13C values are reportedto range from �30 to �26‰, while marine algae valuesrange from �23 to �16‰ (Lamb et al., 2006). Particulateorganic matter measurements from Bristow et al. (2013)(d13C: �20.1� 4.0‰; C/N: 9.6� 1.8) were similar to algalmatter measured in this study, although they estimate thatalgal-derived organic matter only comprised a small propor-tion of the particulate organic matter pool in the Thames(<15% of particulate organic carbon and nitrogen at sam-pling sites where d13C>�17‰). Instead, Bristow et al.(2013) indicate marsh plants and seagrasses contributesignificantly to the particulate organic matter pool of theThames. Aquatic (submergent vascular) vegetation, includingseaweed and macroalgae, had mean C/N values of10.5� 4.4. These relatively low C/N values are similar tothe range of values for this vegetation type (C/N<10)reported in other temperate regions by Meyers (1994) andLamb et al. (2006). The low C/N of algal matter is due tomoderate amounts of structural carbohydrates and greater

1.5

2.0

2.5

3.0

3.5

4.0

Ele

va

tio

n (

m O

D)

-28

-27

-26

-25

13C

8

10

12

14

0

5

10

15

20

0 5 10 15 20 25 30 35 40 45 50

Corg /N

tota

l

TO

C

Distance (m)

MHWST

MHWNT

Dartford Creek Transect

TOCC/N

Floral zone:Tidal flatLow marshMiddle marshHigh marsh

A AI

Figure 2. Transect A–A0 at Dartford Creek. Elevation profile, floralzones, d13C values (‰), total organic carbon (TOC) and C/N ratiosare shown for bulk surface sediment samples along each transect.

2.0

2.5

3.0

3.5

4.0

Ele

va

tio

n (

m O

D)

-29

-28

-27

-26

-25

-24

13C

8

10

12

14

16

18

0

10

20

30

0 5 10 15 20 25 30 35 40 45 50

Co

rg /Nto

talT

OC

Distance (m)

MHWST

MHWNT

Wat Tyler Transect

TOCC/N

Floral zone:

Brackish transition

Low marshMid/High marsh

BIB

Figure 3. Transect B–B0 at Wat Tyler Country Park Nature Reserve.Elevation profile, floral zones, d13C values (‰), total organic carbon(TOC) and C/N ratios are shown for bulk surface sediment samplesalong each transect.



concentrations of N-rich protein than is present in land planttissue (Tyson, 1995).The greatest variation in the d13C of salt marsh vegetation

occurred on the basis of the photosynthetic pathway utilizedby each plant; the d13C of C4 salt marsh vegetation (Spartinaspp.) was �13.5� 0.8‰, while the d13C of C3 salt marshvegetation (all other species) was �26.8� 1.3‰. Thesevalues are consistent with measurements of C4 and C3 saltmarsh plants found in other temperate regions (e.g., Smithand Epstein, 1970, 1971; Jackson et al., 1986; Chmura andAharon, 1995). C4 plants typically range in d13C from �17 to�9‰ (Chmura and Aharon, 1995) and have mutuallyexclusive values from C3 plants (Smith and Epstein, 1971),which typically range from �32 to �21‰ (Deines, 1980).There was greater variability in the C/N of C3 salt marshvegetation (34.6� 20.7) in comparison with C4 plants(27.4�7.7), although this difference may be an artifact of thegreater number of analyses of C3 (n¼140) versus C4 (n¼ 21)vegetation types. No consistent variations in d13C or C/Nvalues of different plant species were found among vegetationtypes. As Lamb et al. (2007) observed, no difference ind13C or C/N values existed between above- and below-ground components of herbaceous C3 or C4 marsh vegetation

in the combined dataset, which enables herbaceous vegeta-tion to be characterized by one grouping (Fig. 6A).In contrast to salt marsh vegetation, C3 freshwater vegeta-

tion showed variation between plant components. Leaf(16.3� 3.5) and wood (57.5�15.3) components had distinctC/N values. This variation is related to the much greaterproportion of N-devoid lignin in wood compared with leaves(Hedges et al., 1986; Tyson, 1995; Vane et al., 2013). Thed13C values of leaf (�31.2� 1.3‰) and wood(�30.2�1.5‰) components, however, were relatively simi-lar due to greater content of 13C-depleted lipids in leaves andlignin in wood (Vane et al., 2013). These values agree withthose found by Hedges et al. (1986), M€uller and Mathesius(1999), M€uller and Voss (1999) and Mackie et al. (2005).Freshwater vegetation (�30.9�3.4‰) had distinctly lowerd13C values than C3 salt marsh vegetation (�26.8� 1.3‰),which may be related to stress imposed on salt marshvegetation from greater ambient salinity (van Groenigen andvan Kessel, 2002). Decreased stomatal conductance (the‘openness’ of the stomatal aperture; Schlesinger, 1997) due toincreased salinity stress causes more CO2 inside the leaf toreact with CO2-fixing enzymes and less fractionation ofisotopes to occur (Guy et al., 1980), causing plant d13C to

7

9

11

0 5 10 15

Co

rg /Nto

tal

Distance (m)

1.5

2.0

2.5

3.0

3.5

4.0

Ele

va

tio

n (

m O

D)

-25

-24

-23

-22

-21

-20

-19

13 C

0

2

4

6

0 10 20 30 40 50 60

TO

C

Distance (m)

Two Tree Transect 1 Two Tree Transect 2

MHWST

MHWNT

TOCC/N

Floral zone:Tidal flatLow marshMid marsh

MHWST

MHWNT

-26

CI DIDC

Figure 4. Transect 1 (C–C0) and transect2 (D–D0) at Two Tree Island. Elevationprofile, floral zones, d13C values (‰),total organic carbon (TOC) and C/Nratios are shown for bulk surface sedi-ment samples along each transect.

0.0

0.5

1.5

Ele

va

tio

n (

m O

D)

-30

-29

-28

-27

13C

15

20

25

30

35

40

45

50

0 50 100 150 200

TO

C

Distance (m)

0 10 20 30 40 50 60

Co

rg /Nto

tal

Distance (m)

Ted Ellis Transect 1 Ted Ellis Transect 2Floral zone:

Reed swampFen carr

TOCC/N

1.0

10

14

18

22

26

30

EIE F FI

MHWST

Figure 5. Transect 1 (E–E0) andtransect 2 (F–F0) at Ted EllisNature Reserve. Elevation profile,floral zones, d13C values (‰),total organic carbon (TOC) andC/N ratios are shown for bulksurface sediment samples alongeach transect.



increase (Farquhar et al., 1982; Guy and Reid, 1986).Although other factors may influence plant d13C values(Twiddy, 1996), this variation in d13C may be useful indistinguishing C3 plant types accumulating under saline andfreshwater conditions.

d13C, TOC and C/N characteristics of bulksediments from coastal wetlands

Site-specific variations of modern sediments

The study sites in the Thames and Norfolk Broads displaysite-specific variations in d13C, TOC and C/N values ofsediments (Table 2). In particular, the d13C of tidal flat/lowmarsh and middle/high marsh sediments of Two Tree Islanddiffered from those of the Dartford Creek and Wat Tyler sites.We infer that these differences in d13C values are related tovariations in pH (and speciation of carbonate ions) andsuspended terrestrial particulate organic matter among sites.At Dartford Creek and Wat Tyler, we presume dissolved CO2

(d13C¼��8‰) is the main inorganic C source of phyto-plankton. Moving towards the estuary mouth at Two TreeIsland, HCO3

� (d13C¼�0‰) becomes a more prominentinorganic C source. This effect may explain the higher tidalflat/low marsh sediment d13C values at Two Tree Island,which are dominated by allochthonous particulate organics.The influence of suspended terrestrial material also decreasesalong the axis of the estuary and could cause sedimentd13C to rise (Wilson et al., 2005b; Zong et al., 2010). Theseexplanations agree with d13C measurements of suspendedorganic matter on the Thames Estuary, which showedincreasing d13C down estuary (Middelburg and Herman,2007). The presence of C4 Spartina anglica (d13C¼�13.2‰)at Two Tree Island may also contribute to its relatively high

d13C values, as well as the distribution and size of the tidalcreeks/channels, which may influence the deposition andredistribution of organic matter sources (Wolaver et al., 1986;Ember et al., 1987; Middelburg et al., 1997). This latter factoris supported by the relatively low TOC and C/N values ofthese sites (Table 2) because if Spartina were substantiallyincorporated into sediment, higher TOC and C/N valueswould be expected. The lower elevation of sampling stationsat Two Tree Island, and thus absence of a developed highmarsh causes increased tidal flux at the Two Tree Islandmiddle marsh and prevents the accumulation of organicmatter by greater export of dissolved and particulate organicmatter and macro-detritus (Boorman et al., 2000). A combi-nation of these factors probably resulted in the observedinter-site variability, although we cannot rule out that mixingof anthropogenic pollution from historical events at Two TreeIsland (Scrimshaw and Lester, 1995) could also cause a shiftin observed d13C values.

Regional database of northern and southern England

We combined our d13C, TOC and C/N sediment values withanalyses from coastal wetlands of the Mersey Estuary (Wilsonet al., 2005ab), Humber Estuary (Lamb et al., 2007), andKent, East Sussex and Norfolk (Andrews, 2008) to create aregional database (n¼ 132) from northern and southernEngland (Table S2). This database incorporates study areasfrom a wide range of physiographic conditions to ensure coresamples have a suitable modern analogue. The rationalebehind the database approach to reconstruct RSL (rather thanlocal Thames transects alone) stems from finding the widestrange of modern environments that can accurately serve asan analogue for Holocene paleoenvironments. This same

-34

-29

-24

5 15 25 35 45 55 65 75 85 95

LeafRootTwigStem or whole plant

C3 salt marsh/Reed swampFen carr

AquaticHabitat/environment:

Plant component:C4 salt marsh

POMC4 marsh

Aquatic

C3 herbaceous salt marsh/reed swamp

C3 freshwater leaves C3 freshwater wood5 10 15 20 25 30 35

C/N

Fen carrMid-high marshPhragmites reed swampTidal flat/low marsh

C3 freshwater leaves C3 freshwater wood

C3 herbaceous salt marsh/reed swamp

C4 marsh

13C

C/N

Sediment depositionalenvironment:

B

-19

-14

Particulate organicmatter (POM)

Particulate organicmatter (POM)

Aquatic

A

Figure 6. d13C (‰) and C/N of plants and surface sediments. (A) d13C and C/N of vegetation end-members grouped by the habitat or environmentfrom which it was collected and the plant component analysed. Aquatic vegetation includes brown and green algae and the genera Fucus andEnteroporpha. Salt marsh/reed swamp vegetation include the genera Agrostis, Aster, Atriplex/Halimione, Borrichia, Cochleria, Elymus, Festuca,Limonium, Puccinellia, Phragmites, Salicornia, Scirpus, Seriphidium, Spartina, Spergularia, Suaeda and Triglochlin. Fen carr vegetation includesthe genera Alnus, Betula, Dryopteris, Rubus, Salix and Urtica. Values come from this study and work by Wilson et al. (2005a,b), Lamb et al.(2007); Andrews (2008) and Bristow et al. (2012). Patterned boxes represent the mean�2SD of each vegetation type in the England database:aquatic (light blue diagonal lines), particulate organic matter (dark blue diagonal lines), C4 salt marsh (orange diagonal lines), C3 salt marsh(purple diagonal lines), C3 freshwater leaves (red diagonal lines) and C3 freshwater wood (green diagonal lines). Two samples with C/N>100were excluded from the plot. (B) d13C and C/N values of surface sediment from this study, grouped by depositional environment andsuperimposed on the ranges defined in (A). Note the change in C/N values on the x-axis. This figure is available in colour online atwileyonlinelibrary.com.



issue has arisen surrounding the use of local vs. regionalmodern training sets for microfossil transfer function-basedRSL reconstructions (Horton and Edwards, 2005; Woodroffeand Long, 2010; Barlow et al., 2013; Kemp et al., 2013b;Watcham et al., 2013). Local training sets tend to producemore precise reconstructions, although the accuracy of thesereconstructions remains questionable in the UK (Horton andEdwards, 2005; Wilson and Lamb, 2012) because they sufferfrom a lack of good analogues; therefore, we emphasize thisregional approach in the application of the compiled modernd13C, TOC and C/N geochemistry database to Holocenearchives. Two criteria governed inclusion of studies into thesediment database: (i) studies must utilize identical samplepreparation methods before analysis to minimize bias incomparison of d13C and C/N values (Brodie et al., 2011;Craven et al., 2013); and (ii) information regarding theenvironment from which the sample was collected (i.e. tidalflat, low marsh, etc., or the dominant vegetation occupyingthe sampling site) and/or sample elevation must be given.Tidal flat/low marsh and middle/high marsh environmentswere grouped together because their d13C, TOC and C/Nranges overlapped and ANOVAs showed the values to beindistinguishable from one another. However, we identifiedstatistically significant differences in sediment d13C, TOC andC/N values among the tidal flat/low marsh group, the middle/high marsh group, reed swamp and fen carr floral zones ofthe database (excluding samples occupied by introducedSpartina spp. and samples from minerogenic alder carrenvironments; Table 3).Tidal flat and low marsh sediments had mean (� SD)

d13C values of �24.9� 1.2‰, TOC values of 3.6�1.7% andC/N values of 9.6� 1.0. The relatively high d13C and lowTOC and C/N values of these sediments result from minimalincorporation of in situ vegetation cover into sediments,greater import of allochthonous particulate organic matterand algae, and high rates of minerogenic sedimentation(Wilson et al., 2005a) (Fig. 6B). In addition, in situ organicmatter from tidal flat and low marsh environments may beexported due to tidal action (Boorman et al., 2000; Bristowet al., 2013) or organic matter may be altered or brokendown due to high rates of microbial activity, stimulated bygreater nutrient import from particulate organic matter sour-ces (Ember et al., 1987; Lamb et al., 2006; Kemp et al.,2010). Tidal flat and low marsh d13C, TOC and C/N values inthe northern/southern England database were similar to thosefrom the Pacific coast of Oregon (d13C: �24.1� 1.7‰, TOC:2.5� 1.8%; C/N: 10.4� 2.7; Engelhart et al., 2013) andnorthwest Europe (e.g., Waarde Marsh of the southwestNetherlands: d13C: �22.0‰; TOC: 1.2%; C/N: 17; Middel-burg et al., 1997). Conversely, d13C was lower and TOC andC/N were higher than sediments from the US Atlantic coast(e.g., North Carolina: d13C: �17.6 to �16.2‰; TOC: 4.0–10.7%; C/N: 14.8–16.1; Kemp et al., 2010 New Jersey: d13C:�18.9 to �15.4‰; TOC: 7.0–24.0%; C/N: 12.1–17.3; Kempet al., 2012; Massachusetts: d13C: �19.5‰; TOC: 5.6%; C/N:18.3; Middelburg et al., 1997). This difference with the tidalflats and low marshes of the US Atlantic coast can beexplained by the dominant occurrence of C4 grasses (predom-inantly Spartina alterniflora and S. patens) and greaterincorporation of vascular vegetation into sediments and/orreduced minerogenic input (Haines, 1976; Chmura et al.,1987; Ember et al., 1987; Chmura and Aharon, 1995;Middelburg et al., 1997; Kemp et al., 2010, 2012).Sediments from the middle/high marsh zone had

d13C values of �26.2� 1.0‰, TOC values of 9.8�6.7% andC/N values of 12.1�1.8. d13C values show little alterationfrom their autochthonous vegetation counterparts (Fig. 6B),

although lower TOC and C/N values suggest a secondarycontribution to sedimentary organic matter from allochtho-nous marine or riverine particulate organic matter sources (C/N< 10; Meyers, 1994) (Fig. 6B). High minerogenic sedimen-tation indicates transport of allochthonous material to themarsh surface, including marine and riverine dissolved andparticulate organic matter (Boorman et al., 2000). Alternative-ly, diagenesis where immobile nitrogen is retained duringsubsequent loss of carbon through oxidation (Chmura et al.,1987; Ember et al., 1987) may explain why bulk sediment C/N values fall within the lower range of vegetation. Theincorporation of fungal mycelium has also been shown toincrease N, with a subsequent drop in C by 10% in degradedmaterial (Vane et al., 2001). Mid/high marsh (absent of C4

vegetation) d13C values of northern/southern England weresimilar to those from Oregon (d13C: �27.3� 1.4‰; Engelhartet al., 2013) and North Carolina (d13C: �26.3�2.1‰; Kempet al., 2010), although TOC and C/N values were much lowerthan high marsh environments of Oregon (TOC: 12.4�4.0%,C/N: 13.6� 1.4; Engelhart et al., 2013) and North Carolina(TOC: 16.6� 11.6%, C/N: 18.4� 2.3; Kemp et al., 2010),which is consistent with the minerogenic and organogenicnature of UK and US marshes, respectively (Allen and Pye,1992; Middelburg et al., 1997). The organic matter content ofmarshes varies between regions in part because lowertemperatures and the shorter growing season in the UK limitbiomass productivity, making mineral matter delivered bytides a more dominant source of sediment accumulation(Allen, 1990; Allen and Pye, 1992; French, 1993; French andSpencer, 1993; Middelburg et al., 1997). However, otherlocal factors that affect mineral sediment deposition, such asthe supply of mineral matter and distance to its source,vegetation type and density, barriers to surface flows, andpost-depositional reworking and erosion by waves or tides,may also account for this regional variation in d13C, TOC andC/N values.Sediments from the reed swamp zone had d13C values of

�27.9�0.7‰, TOC values of 36.5� 11.5% and C/N valuesof 13.9� 1.2. Although Phragmites australis, the dominantreed swamp vegetation, has plant tissues with d13C and C/Nvalues within the same range as C3 salt marsh vegetation,d13C, TOC and C/N values of sediments within this floral zonevary from those of the salt marsh. This variation may beexplained by the reduced tidal influence on the Phragmitesreed swamp zone. Sediments accumulating in this zonerepresent in situ vegetation, rather than a combination of insitu vascular vegetation and allochthonous marine or fluvialparticulate organic matter and algae. Similar to the middle/high marsh, C/N values of bulk sediment are within the lowerrange of vegetation, which may be related to early diagenesis.Phragmites TOC and C/N values from the UK are consistentwith those from a brackish transitional zone occupied byPhragmites in New Jersey (TOC: 23.9�9.7%, C/N: 13.8� 0.6;Kemp et al., 2012), but d13C values of �25.1� 2.0‰ fromNew Jersey (Kemp et al., 2012) are slightly higher. This trendis seen in the upland border in marshes of Massachusetts,where d13C values average �24.5‰, perhaps due to import ofadjacent C4 vegetation or increased salinity stress on Phragmi-tes vegetation itself (Farquhar et al., 1989).Fen carr bulk sediments had d13C values of �29.0� 0.6‰,

TOC of 41.6� 5.7% and C/N values of 17.4� 3.1. Sedimen-tation in this zone is dominantly organogenic, indicated byhigh TOC values and d13C values representative of the C3

freshwater vegetation that occupy this zone. d13C values ofsediments from the fen carr fall within the higher range ofd13C values of their modern vegetation counterparts (Fig. 6B),which suggests the preferential degradation of lignin by



white-rot and soft-rot fungi (Hatakka, 1994; Vane, 2003;Vane et al., 2003, 2005, 2006). Lignin tends to be 4–7‰more depleted in d13C relative to bulk plant material (Benneret al., 1987), and thus its preferential decay in sedimentswould cause d13C values to increase. C/N values of fen carrsediment were within the range of freshwater leaf material(Fig. 6B), which indicates that most sedimentary organicmatter is either primarily derived from leaves or that thewoody material incorporated into sediments is significantlyaltered to cause a large drop in its C/N content, which is alsoconsistent with the breakdown of N-devoid lignocellulosiccompounds. Fen carr d13C, TOC and C/N values from the UKwere similar to upland sediments from Oregon (d13C:>�28.5‰, TOC: 30� 4.6%, C/N: 20.4�3.7), although UKfen carr d13C values were much lower than those fromfreshwater environments from the San Francisco Bay (�27.7to �23.3‰; Cloern et al., 2002), upland sediments from NewJersey (�26.5 to �25.1‰; Kemp et al., 2012) and freshwatermarshes of Louisiana (average: �27.8‰). These discrepan-cies underline the importance of regional and/or site-specificinvestigations of the modern distribution of d13C, TOC and C/N bulk sediment values before interpretation of the sedimen-tary record (Engelhart et al., 2013).

The use of d13C, TOC and C/N to produce sea-levelindex points from the Thames Estuary

We assess the use of the modern distribution of d13C, TOCand C/N values from the UK (excluding samples withinfluence from C4 Spartina) alongside microfossil indicatorsto produce three sea-level index points and one limiting datefrom Core SW1 collected from the Swanscombe marshes onthe southern shore of the River Thames (Fig. 7). Core SW1suffers from microfossil preservation problems and thereforethese indicators can only provide supporting data.Sea-level index points delimit the unique position of RSL

over time and space. The horizontal age component of anindex point is obtained from radiocarbon dating of the sampleand its associated 2s calibrated age range. The verticalcomponent of an index point is estimated using the indicativemeaning of a sample, which describes its relationship to atidal datum (e.g., mean high water spring tide, MHWST) atthe time of deposition using the mid-point (reference waterlevel, RWL) and range over which the indicator is found inthe contemporary environment (indicative range). If a sampleis deposited in a terrestrial environment, it is classified as aterrestrial limiting date, providing only an upper limit on theposition of RSL. Shennan (1982, 1986) and Horton et al.(2000) established the indicative meaning for litho- andbiostratigraphical sequences commonly used to produce sea-level index points (Table 4). We estimated the indicativemeanings of dated transgressive/regressive contacts usingd13C, TOC and C/N values.The transgressive contact of a Phragmites or monocot peat

directly below a clastic salt marsh deposit (from Shennan,1982, 1986; Horton et al., 2000; Shennan and Horton, 2002)corresponds to a d13C, TOC and C/N transition from greaterthan �27‰ to less than �27‰, less than 18% to greater than18%, and less than 14 to greater than 14, respectively(Table 4). For regressive contacts, represented by a Phragmitesor monocot peat directly above a clastic salt marsh deposit,d13C, TOC and C/N change from less than �27‰ to greaterthan �27‰, greater than 18% to less than 18%, and greaterthan 14 to less than 14, respectively. For terrestrial limitingdates, d13C must stay below the threshold value of �27‰,TOC must remain above 18% and C/N must remain higherthan 14. These values are based on the average of the upper

and lower bounds of the mid–high marsh and reed swampenvironments in the sediment database. Using the mean andtwo standard deviations, we calculated the lower and upperlimit of each zone. The lower limit of the mid–high marshzone (e.g., d13C¼�29.3‰) and upper limit of the reedswamp (e.g., d13C¼�24.2‰) were then averaged (e.g., meand13C¼�26.8‰) and rounded to the nearest whole number toobtain the threshold value (e.g., d13C¼�27‰) that appears inTable 4.The gradational transgressive contact between a peat with

unidentifiable plant macrofossils and an overlying gray mudwith organics at �4.76m OD returned a date of 4138–3896 cal a BP (median: 4017) (Fig. 7B). Across the contact,d13C increased from �27.8 to �25.8‰, TOC decreased from37.4 to 3.3% and C/N fell from 16.5 to 13.2. This shift invalues is consistent with a transition from Phragmites reedswamp to a tidal flat/marsh environment. Foraminifera anddiatoms were absent from the peat between �5.30 and�5.10m OD. However, at the transgressive contact, forami-nifera switched from an agglutinated to calcareous-dominatedassemblage, which is consistent with a change in environ-ment from the upper limits of tidal influence to a middle tolow marsh or tidal flat environment (Horton and Edwards,2006). Similarly, the diatom assemblages suggest an increasein salinity preference with the number of polyhalobous andmesohalobous taxa increasing and number of oligohaloboustaxa decreasing across the contact (Denys, 1991/1992).Pollen and spores are present within the peat and theassociations are dominated by ferns with Corylus and Pinus,and importantly Chenopodium attained >15% abundance,reflecting the local presence of salt marsh vegetation (Birkset al., 2001). The geochemical and microfossil data supportDevoy’s (1979) paleoenvironmental interpretation, whichsuggested a change in depositional environment from fen-wood to reed swamp with open salt marsh communities atthe transgressive contact. This contact is interpreted as aPhragmites or monocot peat directly below a clastic saltmarsh deposit (Shennan, 1986; Horton et al., 2000). There-fore, the RWL is MHWST �20 cm with an indicative range of�20 cm (Table 1).A bulk monocot peat overlying a gray mud at the gradual

regressive contact at �7.14m OD produced an age of 6573–6412 cal a BP (median: 6523) (Fig. 7C). d13C decreasedacross the contact from �26.3 to �28.0‰, TOC increasedfrom 5.5 to 42.3% and C/N increased from 13.9 to 25.9. Theupper range of these TOC and C/N values is higher thanobserved ranges of the modern environments within ourdatabase (i.e. no matching analogue) (Fig. 7E,F). The valueswithin the peat between �6.51 and �6.99m OD areconsistent with d13C and C/N values of Holocene oakfenwood settings found in the Humber Estuary (Fig. 7E,F;Andrews et al., 2000), which is also supported by the pollen-based interpretation of an oak fenwood peat (Devoy, 1979).We were unable to sample modern oak fenwood environ-ments in the Thames or Norfolk Broads. Nevertheless, thed13C, TOC and C/N values across the regressive contactsuggest a transition from a middle/high salt marsh to reedswamp or fen carr environment. Foraminifera were notpreserved, and diatoms were sparse with too few individualsto produce quantitative counts (n¼�10), although thosepreserved prefer saline to brackish environment, with littlechange across the contact. Undifferentiated fern sporesdominate the pollen/spore assemblage. The geochemical datasupport the inference of Devoy (1979) who suggested theregressive contact of the peat indicated sedge fen/reed swampwith local salt marsh communities. Therefore, we assignedthis dated contact the RWL of a Phragmites or monocot peat



TOC

Meso

halobous

Polyhalo

bous

Dia

tom

s

40

17

± 1

21

65

23

± 1

11

0 1

0 2

0 3

0 4

0 5

0

76

40

± 6

0

77

18

± 9

9

-29

-2

8 -2

7

15

2

0

25

B C D

Org

an

ic-r

ich

mu

dM

ud

Sa

nd

Gra

vel

Ra

dic

arb

on

dat

e(c

al a

BP

)

Altitude (m OD)

Lithology

-11

-10

-9

-8

-7

-6

-5

-4

-3

-2

-1

0

1

2

A

-26

-29

-2

8 -2

7 -2

6

-29

-2

8 -2

7 -2

6

0 1

0 2

0 3

0 4

0 5

0

0 1

0 2

0 3

0 4

0 5

0

15

2

0

25

15

2

0

25

-30

-29

-28

-27

-26

-25

-24

-22

010

2030

4050

Tida

l flat

/Lo

w m

arsh

Mid

/Hig

h M

arsh

Brac

kish

tran

sitio

n/Re

ed s

wam

p

Fen

carr

TOC

-30

-29

-28

-27

-26

-25

-24

-23

-22 0

510

1520

2530

3540

Tida

l flat

/Lo

w m

arsh Mid

/Hig

h M

arsh

Brac

kish

tran

sitio

n/Re

ed s

wam

p

Fen

carr

E F

BBBBBBBBBB

BB BBBB

BBBB

BBBBCC

CCCC

CC

CCCC

CCCCCCCC CC

CCCC

CCDD DDDDDD DD

DDDD

DD

DD

DD

DDDDDD

DD

BBBBBBBBBB BB

BBBB

BB BB BBBB

CCCC

CC

CC

CCCC

CCCCCC

CCCC

CCCC CCDD DDDDDD DD

DDDD

DD DD

DD

DDDDDD

DD

BBBBBBBBBBBBBBBB /////

DDDDDDDDDDDDDDDDDDDDDDDDDD

BBB

CCCCCCCC

CCCCCCCCC CCC

BBBBB

CCC

B

CC

CC

DDDDD

DDDDDDDD

DDD

DDDD

DDDDDDDDDD

DDDDDDDDD

DDDDDD

DDDD

CCCCCCC DDDDDDDDDDDDDDDD

CCCCCCCCCCCCC

CCCC

CCCCC DDDDDDDDDDDDDDDDDD

CCCCCCCC//////////

BBBBBBBBCCCC

CCCCC

CCHol

ocen

e oa

kfe

nwoo

d

BBBBBBBB

DDDDDDDDD

DDDDDDDDDDDDD

BBBBBCCCCCCCCC

BBBBB

DDDDDDDDDDDDDDDDDD

BBBBBCC

Hol

ocen

e oa

kfe

nwoo

d

Pe

at

-5.3

-5.1

-4.9

-4.7

-4.5

-4.3

-7.7

-7.5

-7.3

-7.1

-6.9

-6.7

-6.5

-10

.3

-10

.1

-9.9

-9.7

-9.5

-9.3

-9.1

30

30

30

Olig

ohalobous

Halo

phobous

-23

Re

lative

Ab

un

da

nce

(%

)

Foraminifera

Fora

min

ifera

l ass

embl

age

Ab

sen

t

Ca

lca

reo

us-

do

min

an

t

Ag

glu

tin

ate

d-d

om

ina

nt

20

40

60

80

20

40

60

20

40

60

80

20

40

60

20

40

60

80

20

40

60

DIA

TOM

S AB

SEN

TO

R SP

ARSE

20

20

20

40

20

20

20

20

20

DIA

TOM

S AB

SEN

TO

R SP

ARSE

Figure 7. Derivation of sea-level index points from transgressive/regressive contacts in Core SW1. (A) Stratigraphy of Core SW1. (B) Litho,chemo-, bio- and chronostratigraphy of the transgressive contact at �4.76m OD. (C) Litho, chemo-, bio- and chronostratigraphy of the regressivecontact at �7.14m OD. (D) Litho, chemo-, bio-, and chrono-stratigraphy of the transgressive and regressive contacts at �9.56 and �9.71m OD.(E) Comparison between modern range of bulk sediment d13C and TOC and SW1 core sediment. Shaded boxes represent the full range of modernd13C and TOC values of each environment in the northern/southern England database. Letters (B, C, D) indicate geochemical values acrosstransgressive/regressive contacts shown in (B) (�4.76m OD), (C) (�7.14m OD) and (D) (�9.56 and �9.71m OD), respectively. (F) Comparisonbetween modern ranges of d13C and C/N of bulk sediments and SW1 core sediment. Shaded boxes represent full range of modern d13C and C/Nvalues of each environment in the northern/southern England database. Letters (B, C, D) indicate geochemical values across transgressive/regressive contacts shown in (B) (�4.76m OD), (C) (�7.14m OD) and (D) (�9.56 and �9.71m OD), respectively. This figure is available incolour online at wileyonlinelibrary.com.



directly above clastic salt marsh deposit (Shennan, 1986;Horton et al., 2000), which is (MHWST�HAT)/2–20 cm, andan indicative range of� 20 cm (Table 1).A date of 7700–7580 cal a BP (median: 7640) was

obtained from a Phragmites fragment at the gradual transgres-sive contact between a peat and mud at �9.56m OD(Fig. 7D). Across the contact, d13C increased from �29.7 to�26.8‰, TOC decreased from 48.5 to 5.9% and C/N fellfrom 24.8 to 16.0. TOC values are slightly higher than themodern range of reed swamp and fen carr environments,suggesting Holocene environments accumulated greateramounts of organic matter than their modern equivalents.Foraminifera and diatoms are absent from the peat andoverlying mud, although diatoms between �9.40 and�9.10m are dominated by polyhalobous and mesohaloboustaxa indicative of a tidal flat environment. Alnus and Corylusdominate pollen within the peat, with Poaceae (grass)subdominant. The change in d13C, TOC and C/N values isconsistent with a transition in paleoenvironment from reedswamp to salt marsh conditions, which agrees with pollenanalysis from this study and Devoy (1979). Devoy (1979)suggested the depositional environment at the transgressivecontact was a sedge fen/reed swamp. We interpret thiscontact to represent an index point with an RWL of MHWST– 20 cm and indicative range of� 20 cm (Table 1).A wood fragment within an alder carr peat at �9.71m OD

was dated to 7817–7620 cal a BP (median: 7718) (Fig. 7D).d13C values 8 cm above and below the date ranged from�28.1 to �28.7‰, TOC ranged between 24.1 and 48.8%and C/N increased from 20.4 to 26.7. These values areconsistent with modern fen carr environments, although againcore TOC values were slightly higher than modern fen carrcounterparts. Foraminifera and diatoms were absent from thepeat, which was also palynologically sparse with only Alnuspollen and undifferentiated fern spores preserved. The d13C,TOC and C/N values permit interpretation of this dated woodto be a freshwater limiting point, which formed above meantide level (Table 1).We find good agreement between modern and Holocene

sediment d13C, TOC and C/N values (Fig. 7E,F), which isfurther supported by pollen-based interpretations of Devoy(1979). This correspondence suggests that these geochemicalindicators can be used to identify past depositional environ-ments in the Thames Estuary. However, post-depositionalchange may cause inaccurate interpretations. The concentra-tion of refractory plant lignin due to preferential breakdownof more labile cellulose and hemicellulose compounds isoften considered the major factor causing post-depositionalchange in d13C values in coastal sedimentary archives (e.g.,Benner et al., 1987; Wilson et al., 2005a,b; Lamb et al.,2006, 2007). Lignin is 4–7‰ more depleted in d13C valuesrelative to bulk plant material, so as the more labilecomponents are broken down and respired by bacteria and

fungi, the d13C of the remaining organic matter reflects theincreasing relative lignin content. The preferential degrada-tion of lignin by white-rot fungi in aerobic environments mayalso occur, which would result in a relative increase insediment d13C (Hatakka, 1994; Vane, 2003; Vane et al.,2005, 2006). Cellulose, hemicellulose and lignin are devoidof N, so loss of these compounds would result in decreasedTOC values and increased C/N values. Nevertheless, we findno systematic offsets between Holocene sediment frommodern d13C, TOC and C/N ranges that are consistent withthe changes described above, which implies that post-depositional change is not prohibitive in paleoenvironmentalinterpretation in the Thames.The utility of d13C, TOC and C/N in the production of

Holocene sea-level index points is clearly illustrated in theapplication to Core SW1. Poor preservation of microfossils,an issue found elsewhere throughout the Thames Estuary(Devoy, 1979) and the UK (Horton et al., 2000; Metcalfeet al., 2000; Roberts et al., 2006), inhibited interpretation ofstratigraphic contacts in the core. Based on lithology andplant macrofossils, inference could have been made ontransgressive/regressive contacts in the core (e.g., Shennanand Horton, 2002), but in the absence of supporting informa-tion from microfossils, uncertainties would exist in ourinterpretation. d13C, TOC and C/N values that are based onan extensive modern data set provide additional confidencein the interpretation of contacts within the core.

Conclusions

We investigated the use of d13C, TOC and C/N values frombulk sedimentary organic matter to reconstruct RSL andpaleoenvironmental change by sampling vegetation and bulksediment from modern transects at four coastal wetlands onthe Thames Estuary and Norfolk Broads, UK. We reviewedpublished and unpublished studies from the UK to compile asediment d13C, TOC and C/N geochemistry database fromcoastal wetlands of northern and southern England. Tidal flat/low marsh, mid/high marsh, reed swamp and fen carrenvironmental zones had statistically distinct d13C, TOC andC/N values due to the relative amounts of in situ vegetationand tidal-derived allochthonous particulate organic matterand algae incorporated into sediments. Intra-site variability insediment d13C values was observed related to variations insite salinity and the presence of the C4 species Spartinaanglica.d13C, TOC and C/N values of tidal flat/low marsh (d13C:

�24.9� 1.2‰; TOC: 3.6�1.7%; C/N: 9.9�0.8), middlemarsh/high marsh (d13C: �26.2� 1.0‰; TOC: 9.8�6.7%; C/N: 12.1� 1.8), reed swamp (d13C: �27.9� 0.7‰: TOC:36.5� 11.5%; C/N: 13.9� 1.2) and fen carr (d13C:�29.0� 0.6‰; TOC: 41.6� 5.7%; C/N: 17.4�3.1) environ-ments from the bulk sediment database (removing sites

Table 4. Summary of the indicative meanings used to calculate the position of relative sea level of index points and limiting data in the UKdatabase (Shennan, 1982, 1986; Horton et al., 2000; Shennan and Horton, 2002). Updated with corresponding changes in d13C, TOC and C/Nacross (�20 cm above or below) contacts. Mean high water spring tide (MHWST), mean tide level (MTL) and highest astronomical tide (HAT).

Sample typeIndicativerange

Referencewater level d13C TOC Corg/Ntotal

Phragmites or monocot peatDirectly above clasticsalt marsh deposit

�20 cm (MHWST�HAT)/2–20 cm

Decrease from > to< �27‰

Increase from < to>18%

Increase from < to>14

Directly below clasticsalt marsh deposit

�20 cm MHWST – 20 cm Increase from < to>�27‰

Decrease from > to<18%

Decrease from > to<14

Terrestrial limiting >MTL MTL ��27‰ �18% �14



occupied by Spartina spp.) were used to interpret sequencesfrom a Holocene sediment core with poor microfossilpreservation collected from Swanscombe marshes on theThames Estuary. d13C, TOC and C/N were consistent withinterpretations based on microfossils, where preserved. Holo-cene environments appeared to accumulate greater amountsof organic matter than their modern equivalents. The geo-chemical dataset was used to estimate the indicative mean-ings of radiocarbon-dated samples in the core to producethree new sea-level index points and one terrestrial limitingdate. We find that d13C, TOC and C/N of bulk sedimentaryorganic matter can together be used as an effective tool in thepaleoenvironmental interpretation of Holocene sediments inthe absence of microfossil indicators.

Supporting information

Additional supporting information can be found in the onlineversion of this article:Table S1. Central and southern England vegetation and

particulate organic matter (POM) d13C and C/N from thisstudy and others.Table S2. Northern and southern England database of

sediment d13C, TOC, C/N from this study and others.Appendix S1. d13C, TOC and C/N of samples from

Holocene sediment cores collected from Swanscombe marsh,Thames Estuary, UK. Sample altitude given in meters OD.Appendix S2. Foraminiferal counts listed by depth.Appendix S3. Relative abundance of dominant (>2%)

diatom species listed by sample depth (m OD).Appendix S4. Listing of the formal and common names of

the Quaternary pollen and spore taxa observed in this study.

Acknowledgements. C.H.V., J.B.R. and C.P.K. publish with permis-sion of the Executive Director of the British Geological Survey (NERC).This study was funded by the British Geological Survey Climate andLandscape Change theme and a British Geological Survey UniversityFunding Initiative studentship. Additional funding for this study wasprovided by NOAA grant NA11OAR4310101 and NSF grant EAR1419366. The NERC Isotope Geosciences Laboratory (NIGL) signifi-cantly supported this work. The modern plant data from RomneyMarsh were taken from the MSc thesis of Jonathan Andrews, and thesupervisory team (Antony Long, Martyn Waller and Melanie Leng) hasgiven permission for its use in this paper. We thank Martyn Waller andJason Kirby for valuable guidance during site selection and twoanonymous reviewers whose comments improved the manuscript.This paper is a contribution to IGCP project 588 ‘Preparing for coastalchange’ and PALSEA 2. The authors have no financial or conflict ofinterest to report. We dedicate this paper to Fred Scatena, our dearcolleague, friend and mentor who passed away before we completedthis manuscript.

Abbreviations. ANOVA, analysis of variance; MHWST, mean highwater spring tide; OD, Ordnance Datum; RSL, relative sea level; RWL,reference water level; TOC, total organic carbon.

References

Abdi H, Williams LJ. 2010. Tukey’s honestly significant difference(HSD) test. In Encyclopedia of Research Design, Salkind N (ed.).Sage Publications, Thousand Oaks, CA.

Admirality Tide Tables. 2013. Volume 1, UK and Ireland. UKHydrographic Office: London.

Allen JRL. 1990. Salt-marsh growth and stratification: a numericalmodel with special reference to the Severn Estuary, southwestBritain. Marine Geology 95: 77–96.

Allen JRL, Lamb AL, Dark P. 2007. Seasonality of 13C and C/N ratiosin modern and mid-Holocene sediments in the Severn EstuaryLevels, SW Britain. The Holocene 17: 139–144.

Allen JRL, Pye K. 1992. Saltmarshes: Morphodynamics, Conservation,and Engineering Significance. Cambridge University, Press.

Andrews JB. 2008. Use of carbon isotope and C/N geochemistry inreconstructing vegetation communities: a mid- to late- Holocenepalaeoenvironmental investigation from Romney Marsh, Kent, UK.Masters thesis, Durham University.

Andrews JE, Samways G, Dennis PF, et al. 2000. Origin, abundanceand storage of organic carbon and sulphur in the HoloceneHumber Estuary: emphasizing human impact on storage changes InHolocene Land-Ocean Interaction and Environmental ChangeAround the North Sea, Shennan I, Andrews JE (eds). GeologicalSociety of London: London; 145–170.

Ball DF. 1964. Loss-on-ignition as an estimate of organic matter andorganic carbon in non-calcareous soils. Journal of Soil Science 15:84–92.

Barlow NLM, Shennan I, Long AJ, et al. 2013. Salt marshes as LateHolocene tide gauges. Global and Planetary Change 106: 90–110.

Benedict CR, Wong WW, Wong JH. 1980. Fractionation of the stableisotopes of inorganic carbon by seagrasses. Plant Physiology 65:512–517.

Benner R, Fogel ML, Sprague EK, et al. 1987. Depletion of 13C inlignin and its implications for stable carbon isotope studies. Nature329: 708–710.

Binney HA, Waller MP, Bunting MJ, et al. 2005. The interpretation offen carr pollen diagrams: the representation of the drylandvegetation. Review of Palaeobotany and Palynology 134: 197–218.

Birks HH, Peglar SM, Boomer I, et al. 2001. Palaeolimnologicalresponses of nine North African lakes in the CASSARINA Project torecent environmental changes and human impact detected by plantmacrofossil, pollen, and faunal analyses. Aquatic Ecology 35: 405–430.

Boorman L. 2003 Saltmarsh Review. An Overview of Coastal Salt-marshes, Their Dynamic and Sensitivity Characteristics for Conser-vation and Management (JNCC Report No. No. 334). Joint NatureConservation Committee, Peterborough, UK.

Boorman L, Hazelden J, Boorman M. 2000. EUROSAM Final Report,Task 6: Analysis of the nutrient control of the organic carbon cyclein the saltmarsh (ANCOSM) and the modelling of the nitrogencycle (NITROMOD). L A B Coastal: Cambridgeshire, UK.

Bradley SL, Milne GA, Shennan I, et al. 2011. An improved glacialisostatic adjustment model for the British Isles. Journal of Quaterna-ry Science 26: 541–552.

Bristow LA, Jickells TD, Weston K, et al. 2013. Tracing estuarineorganic matter sources into the southern North Sea using C and Nisotopic signatures. Biogeochemistry 113: 9–22.

Brodie CR, Casford JSL, Lloyd JM, et al. 2011. Evidence for bias in C/N, d13C and d15N values of bulk organic matter, and onenvironmental interpretation, from a lake sedimentary sequence bypre-analysis acid treatment methods. Quaternary Science Reviews30: 3076–3087.

Boyle J. 2004. A comparison of two methods for estimating theorganic matter content of sediments. Journal of Paleolimnology 31:125–127.

Chmura GL, Aharon P. 1995. Stable carbon isotope signatures ofsedimentary carbon in coastal wetlands as indicators of salinityregime. Journal of Coastal Research 11: 124–135.

Chmura GL, Aharon P, Socki RA, et al. 1987. An inventory of13C abundances in coastal wetlands of Louisiana, USA: vegetationand sediments. Oecologia 74: 264–271.

Cloern JE, Canuel EA, Harris D. 2002. Stable carbon and nitrogenisotope composition of aquatic and terrestrial plants of the SanFrancisco Bay estuarine system. Limnology and Oceanography 47:713–729.

Craven KF, Edwards RJ, Goodhue R, et al. 2013. Evaluating theinfluence of selected acid pre-treatment methods on C/N andd13C of temperate inter-tidal sediments for relative sea levelreconstruction. Irish Journal of Earth Sciences 31: 25–42.

Deines P. 1980. The isotopic composition of reduced organic carbon.In Handbook of Environmental Isotope Geochemistry, the Terrestri-al Environment, Fritz P, Fontes J (eds). Elsevier Scientific PublishingCompany: New York; 329–406.

DeLaune R. 1986. The use of d13C signature of C- 3 and C-4 plants indetermining past depositional environments in rapidly accreting



marshes of the Mississippi River deltaic plain, Louisiana, U.S.A.Chemical Geology Isotope Geoscience Section 59: 315–320.

Denys L. 1991/1992. A check-list of the diatoms in the Holocenedeposits of the western Belgian coastal plain with a survey of theirapparent ecological requirements. Professional Paper. Berchem,Belgium.

Devoy RJN. 1979. Flandrian Sea level changes and vegetationalhistory of the Lower Thames Estuary. Philosophical Transactions ofthe Royal Society of London B Biological Sciences 285: 355–407.

Devoy RJN. 1992. Questions of coastal protection and the humanresponse to sea-level rise in Ireland and Britain. Irish Geography25: 1–22.

Ember LM, Williams DF, Morris JT. 1987. Processes that influencecarbon isotope variations in salt marsh sediments. Marine EcologyProgress Series 36: 33–42.

Emery KO, Wigley RL, Bartlett AS, et al. 1967. Freshwater peat onthe continental shelf. Science 158: 1301–1307.

Engelhart S, Horton B, Kemp A. 2011. Holocene sea-level changesalong the United States’ Atlantic coast. Oceanography 24: 70–79.

Engelhart SE, Horton BP, Vane CH, et al. 2013. Modern foraminifera,d13C, and bulk geochemistry of central Oregon tidal marshes andtheir application in paleoseismology. Palaeogeography Palaeocli-matology Palaeoecology 377: 13–27.

Farquhar G, O’Leary M, Berry J. 1982. On the relationship betweencarbon isotope discrimination and the intercellular carbon dioxideconcentration in leaves. Functional Plant Biology 9: 121–137.

Farquhar GD, Hubick KT, Condon AG, et al. 1989. Carbon isotopefractionation and plant water-use efficiency In Stable Isotopes inEcological Research, Rundel PW, Ehleringer JR, Nagy KA (eds).Springer Verlag: Berlin.

Field A. 2007. Analysis of variance (ANOVA). In Encyclopedia ofMeasurement and Statistics, Salkind NJ, Rasmussen K (eds). SagePublications: Thousand Oaks, CA.

Fogel ML, Cifuentes LA. 1993. Isotope fractionation during primaryproduction. In Organic Geochemistry, Topics in Geobiology, EngelMH, Macko SA (eds). Springer: New York; 73–98.

Fogel ML, Cifuentes LA, Velinsky DJ, et al. 1992. Relationship ofcarbon availability in estuarine phytoplankton to isotopic composi-tion. Marine Ecology Progress Series 82: 291–300.

French JR. 1993. Numerical simulation of vertical marsh growth andadjustment to accelerated sea-level rise, North Norfolk, U.K. EarthSurface Processes and Landforms 18: 63–81.

French JR, Spencer T. 1993. Dynamics of sedimentation in a tide-dominated backbarrier salt marsh, Norfolk, UK. Marine Geology110: 315–331.

Fry B, Scalan RS, Parker PL. 1977. Stable carbon isotope evidence fortwo sources of organic matter in coastal sediments: seagrasses andplankton. Geochimica et Cosmochimica Acta 41: 1875–1877.

Gehrels WR. 2010. Late Holocene land- and sea-level changes in theBritish Isles: implications for future sea-level predictions. Quaterna-ry Science Reviews 29: 1648–1660.

Gehrels WR, Long AJ. 2007. Quaternary land-ocean interactions: sea-level change, sediments and tsunami. Marine Geology 242: 1–4.

Go~ni MA, Thomas KA, Goni MA. 2000. Sources and transformationsof organic matter in surface soils and sediments from a tidal estuary(North Inlet, South Carolina, USA). Estuaries 23: 548–564.

Guy RD, Reid DM. 1986. Photosynthesis and the influence of CO2-enrichment on d13C values in a C3 halophyte. Plant Cell andEnvironment 9: 65–72.

Guy RD, Reid DM, Krouse HR. 1980. Shifts in carbon isotope ratiosof two C3 halophytes under natural and artificial conditions.Oecologia 44: 241–247.

Haggart BA. 1995. A re-examination of some data relating toHolocene sea-level changes in the Thames estuary. In TheQuaternary of the Lower Reaches of the Thames. QuaternaryResearch Association Field Guide, Bridgeland DR, Allen P, HaggartBA (eds). QRA: London; 329–335.

Haines EB. 1976. Relation between the stable carbon isotopecomposition of fiddler crabs, plants, and soils in a salt marsh.Limnology and Oceanography 21: 880–883.

Haines EB. 1977. The origins of detritus in Georgia salt marshestuaries. Oikos 29: 254–260.

Hartley B, Ross R, Williams DM. 1986. A check-list of the freshwater,brackish and marine diatoms of the British Isles and adjoiningcoastal waters. Journal of the Marine Biological Association of theUnited Kingdom 66: 531–610.

Hatakka A. 1994. Lignin-modifying enzymes from selected white-rotfungi: production and role from in lignin degradation. FEMSMicrobiology Reviews 13: 125–135.

Hedges JI, Clark WA, Quay PD, et al. 1986. Compositions and fluxesof particulate organic material in the Amazon River. Limnologyand Oceanography 31: 717–738.

Horton BP, Edwards RJ. 2005. The application of local and regionaltransfer functions to the reconstruction of Holocene sea levels,north Norfolk, England. Holocene 15: 216–228.

Horton BP, Edwards RJ. 2006. Quantifying Holocene Sea-LevelChange Using Intertidal Foraminifera: Lessons from the British Isles.Cushman Foundation for Foraminiferal Research.

Horton BP, Edwards RJ, Lloyd JM. 1999. UK intertidal foraminiferaldistributions: implications for sea-level studies. Marine Micropale-ontology 36: 205–223.

Horton BP, Edwards RJ, Lloyd JM. 2000. Implications of a microfos-sil-based transfer function in Holocene sea-level studies. In Holo-cene Land-Ocean Interaction and Environmental Change Aroundthe North Sea, Shennan I, Andrews JE (eds). Geological Society ofLondon: London; 41–54.

Horton BP, Long AJ, Donnelly JP. 2009. Quaternary ice sheet-oceaninteractions and landscape responses. Quaternary Science Reviews28: 1570–1572.

Troels-Smith. 1955. Characterisation of unconsolidated sediments.Dammarks Geol Undersogelse Raekke IV: 38–73.

Jackson D, Harkness DD, Mason CF, et al. 1986. Spartina anglica asa carbon source for slat-marsh invertebrates: a study usingd13C values. Oikos 46: 163–170.

Juggins S. 1992. Diatoms in the Thames Estuary, England: ecology,palaeoecology, and salinity transfer function. BibliothecaDiatomologia.

Keeley JE, Sandquist DR. 1992. Carbon: freshwater plants. Plant Celland Environment 15: 1021–1035.

Kemp AC, Nelson AR, Horton BP. 2013a. 14.31 radiocarbon datingof plant macrofossils from tidal-marsh sediment. In Treatise onGeomorphology, Shroder JF (ed.). Academic Press: San Diego;370–388.

Kemp AC, Telford RJ, Horton BP, et al. 2013b. ReconstructingHolocene sea level using salt-marsh foraminifera and transferfunctions: lessons from New Jersey, USA. Journal of QuaternaryScience 28: 617–629.

Kemp AC, Vane CH, Horton BP, et al. 2010. Stable carbon isotopesas potential sea-level indicators in salt marshes, North Carolina,USA. The Holocene 20: 623–636.

Kemp AC, Vane CH, Horton BP, et al. 2012. Application of stablecarbon isotopes for reconstructing salt-marsh floral zones andrelative sea level, New Jersey, USA. Journal of Quaternary Science27: 404–414.

Kramer CY. 1956. Extension of multiple range tests to group meanswith unequal numbers of replications. Biometrics 12: 307–310.

Lamb AL, Vane CH, Wilson GP, et al. 2007. Assessing d13C and C/Nratios from organic material in archived cores as Holocene sealevel and palaeoenvironmental indicators in the Humber Estuary,UK. Marine Geology 244: 109–128.

Lamb AL, Wilson GP, Leng MJ. 2006. A review of coastal palae-oclimate and relative sea-level reconstructions using d13C and C/Nratios in organic material. Earth-Science Reviews 75: 29–57.

Lambeck K, Smither C, Johnston P. 1998. Sea-level change, glacialrebound and mantle viscosity for northern Europe. GeophysicalJournal International 134: 102–144.

Li H, Yang SL. 2009. Trapping effect of tidal marsh vegetation onsuspended sediment, Yangtze Delta. Journal of Coastal Research254: 915–924.

Long AJ. 1992. Coastal responses to changes in sea-level in the EastKent Fens and southeast England, UK over the last 7500 years.Proceedings of the Geologists’ Association 103: 187–199.

Long AJ. 1995. Sea-level and crustal movements in the Thamesestuary, Essex and east Kent. In The Quaternary of the Lower



Reaches of the Thames. Quaternary Research Association FieldGuide, Bridgeland DR, Allen P, Haggart BA (eds). QRA: London.

Long AJ, Innes JB. 1995. The backbarrier and barrier depositionalhistory of Romney Marsh, Walland Marsh and Dungeness, Kent,England. Journal of Quaternary Science 10: 267–283.

Long AJ, Scaife RG, Edwards RJ. 1999. Pine pollen in intertidalsediments from Poole Harbour, UK; implications for late-Holocenesediment accretion rates and sea-level rise. Quaternary Internation-al 55: 3–16.

Mackie EAV, Leng MJ, Lloyd JM, et al. 2005. Bulk organic d13C andC/N ratios as palaeosalinity indicators within a Scottish isolationbasin. Journal of Quaternary Science 20: 303–312.

Mackie EAV, Lloyd JM, Leng MJ, et al. 2007. Assessment of d13C andC/N ratios in bulk organic matter as palaeosalinity indicators inHolocene and Lateglacial isolation basin sediments, northwestScotland. Journal of Quaternary Science 22: 579–591.

Malamud-Roam F, Ingram BL. 2001. Carbon isotopic compositions ofplants and sediments of tide marshes in the San Francisco Estuary.Journal of Coastal Research 17: 17–29.

Malamud-Roam F, Ingram BL. 2004. Late Holocene d13C and pollenrecords of paleosalinity from tidal marshes in the San FranciscoBay estuary, California. Quaternary Research 62: 134–145.

Massey AC, Gehrels WR, Charman DJ, et al. 2008. Relative sea-levelchange and postglacial isostatic adjustment along the coast ofsouth Devon, United Kingdom. Journal of Quaternary Science 23:415–433.

McVean DN. 1956. Ecology of Alnus glutinosa (L.) Gaertn: V. Noteson some British alder populations. Journal of Ecology 44: 321–330.

Metcalfe SE, Ellis S, Horton BP, et al. 2000. The Holocene evolutionof the Humber Estuary: reconstructing change in a dynamicenvironment. In Holocene Land-Ocean Interaction and Environ-mental Change Around the North Sea, Shennan I, Andrews JE (eds).Geological Society of London: London; 97–118.

Meyers PA. 1994. Preservation of elemental and isotopic sourceidentification of sedimentary organic matter. Chemical Geology114: 289–302.

Middelburg JJ, Herman PMJ. 2007. Organic matter processing in tidalestuaries. Marine Chemistry 106: 127–147.

Middelburg JJ, Nieuwenhuize J, Lubberts RK, et al. 1997. Organiccarbon isotope systematics of coastal marshes. Estuarine Coastaland Shelf Science 45: 681–687.

Milker Y, Horton BP, Vane CH, et al. 2015. Seasonal distribution ofintertidal foraminifera, bulk sediment stable carbon isotope and C/N geochemistry in an Oregon tidal marsh. Journal of ForaminiferalResearch 45: 146–155.

Milne GA, Peros M. 2013. Data-model comparison of Holocene sea-level change in the circum-Caribbean region. Global and PlanetaryChange 107: 119–131.

Mitchell S, Akesson L, Uncles R. 2012. Observations of turbidity inthe Thames estuary, United Kingdom. Water and EnvironmentJournal 26: 511–520.

Moore PD, Webb JA, Collinson ME. 1991. Pollen Analysis. BlackwellPublishing: Oxford.

M€uller A, Mathesius U. 1999. The palaeoenvironments of coastallagoons in the southern Baltic Sea, I. The application of sedimenta-ry Corg/N ratios as source indicators of organic matter. Palae-ogeography, Palaeoclimatology, Palaeoecology 145: 1–16.

M€uller A, Voss M. 1999. The palaeoenvironments of coastal lagoonsin the southern Baltic Sea, II. d13C and d15N ratios of organicmatter — sources and sediments. Palaeogeography, Palaeoclima-tology, Palaeoecology 145: 17–32.

Murray JW. 2006. Ecology and Applications of Benthic Foraminifera.Cambridge University Press: Cambridge.

Ostrowska A, Porebska G. 2012. Assessment of TOC-SOM and SOM-TOC conversion in forest soil. Polish Journal of EnvironmentalStudies 21: 1767–1775.

Patrick R, Reimer CW. 1966 Diatoms of the United States, Mono-graph 13. Academy of the Natural Sciences of Philadelphia.

Patrick R, Reimer CW. 1975. Diatoms of the United States,Monograph 13. Academy of the Natural Sciences of Philadelphia.

Preuss H. 1979. Progress in computer evaluation of sea level datawithin the IGCP Project no. 61. In 1978 International Symposium

on Coastal Evolution in the Quaternary. Universidade Sao Paulo:Sao Paulo; 104–134.

Ranwell DS. 1972. Ecology of Salt Marshes and Sand Dunes.Chapman & Hall: London.

Reimer P. 2013. IntCal13 and Marine13 radiocarbon age calibrationcurves 0-50,000 years cal BP. Radiocarbon 55: 1869–1887.

Riding JB, Kyffin-Hughes JE. 2004. A review of the laboratorypreparation of palynomorphs with a description of an effectivenon-acid technique. Revista Brasileira de Paleontologia 7: 13–44.

Roberts DH, Chiverrell RC, Innes JB, et al. 2006. Holocene sea levels,Last Glacial Maximum glaciomarine environments and geophysicalmodels in the Northern Irish Sea Basin, UK. Marine Geology 231:113–128.

Sawai Y, Horton BP, Nagumo T. 2004. The development of a diatom-based transfer function along the Pacific coast of eastern Hokkaido,northern Japan—an aid in paleoseismic studies of the Kurilsubduction zone. Quaternary Science Reviews 23: 2467–2483.

Schlesinger WH. 1997. Biogeochemistry: an Analysis of GlobalChange. Elsevier: Amsterdam.

Schumacher BA. 2002. Methods for the determination of total organiccarbon (TOC) in soils and sediments. Ecological Risk AssessmentSupport Center , NCEA-C1283; 1–23.

Scrimshaw MD, Lester JN. 1995. Organochlorine contamination insediments of the inner Thames Estuary. Water and EnvironmentJournal 9: 519–525.

Sen Gupta BK (ed.). 2002. Modern Foraminifera. Springer: London.Shennan I. 1982. Interpretation of Flandrian sea-level data from theFenland, England. Proceedings of the Geologists’ Association 93:53–63.

Shennan I. 1986. Flandrian sea-level changes in the Fenland. I: Thegeographical setting and evidence of relative sea-level changes.Journal of Quaternary Science 1: 119–153.

Shennan I. 1987. Sea-level changes in the North Sea. In Sea-LevelChanges, Tooley MJ, Shennan I (eds). Basil Blackwell: Oxford;109–151.

Shennan I. 1989. Holocene crustal movements and sea-level changesin Great Britain. Journal of Quaternary Science 4: 77–89.

Shennan I, Andrews JE. 2000. Holocene Land-Ocean Interaction andEnvironmental Change Around the North Sea. Geological Societyof London: London.

Shennan I, Bradley S, Milne G, et al. 2006. Relative sea-levelchanges, glacial isostatic modelling and ice-sheet reconstructionsfrom the British Isles since the Last Glacial Maximum. Journal ofQuaternary Science 21: 585–599.

Shennan I, Horton BP. 2002. Holocene land- and sea-level changesin Great Britain. Journal of Quaternary Science 17: 511–526.

Shennan I, Long A, Horton BP. 2015. Handbook of Sea-LevelResearch. Wiley: Chichester.

Sidell J. 2003. Holocene sea level change and archaeology in theinner Thames Estuary, London, UK. Doctoral thesis, DurhamUniversity.

Sidell J, Scaife RG, Wilkinson KN, et al. 2000. The HoloceneEvolution of the London Thames. MoLAS Monograph 5. Museumof London: London.

Smith BN, Epstein S. 1970. Biogeochemistry of the stable isotopes ofhydrogen and carbon in salt marsh biota. Plant Physiology 46:738–742.

Smith BN, Epstein S. 1971. Two categories of 13C/12C ratios forhigher plants. Plant Physiology 47: 380–384.

Smith DE, Firth CR, Cullingford RA. 2002. Relative sea level trendsduring the early-middle Holocene along the eastern coast ofmainland Scotland, UK. Boreas 33: 185–202.

Smith DE, Hunt N, Firth CR, et al. 2012. Patterns of Holocene relativesea level change in the north of Britain and Ireland. QuaternaryScience Reviews 54: 58–76.

Smol JP, Stoermer EF, editors. 2010. The Diatoms: Applications forthe Environmental and Earth Sciences. Cambridge University Press:Cambridge.

Sokal RR, Rohlf FJ. 1969. Introduction to Biostatistics. W. H.Freedman: New York.

Stuiver M, Polach HA. 1977. Reporting of 14C data�discussion.Radiocarbon 19: 355–363.



Tooley MJ. 1982. Sea-level changes in northern England. Proceedingsof the Geologists’ Association 93: 43–51.

Tooley MJ. 1985. Sea levels. Progress in Physical Geography 9: 113–120.

Tukey JW. 1953. Some selected quick and easy methods of statisticalanalysis. Transactions of the New York Academy of Sciences 16:88–97.

Twiddy EJ. 1996. Applications of stable carbon and oxygen isotopeanalysis to some aspects of coastal environmental change. PhDThesis, Durham University.

Tyson RV. 1995. Sedimentary Organic Matter: Organic Facies andPalynofacies. Chapman & Hall: London.

van de Plassche O. 1982. Sea-level change and water-level move-ments in the Netherlands during the Holocene. Mededelingen RijksGeologische Dienst 36: 1–93.

van de Werff H, Huls H. 1958-1966. Diatomeeënflora van Neder-land, 8 parts. Published privately by van de Werff, De Hoef, TheNetherlands.

van der Wal D, Pye K. 2004. Patterns, rates and possible causes ofsalt marsh erosion in the Greater Thames area (UK). Geomorpholo-gy 61: 373–391.

van Groenigen JW, van Kessel C. 2002. Salinity-induced patterns ofnatural abundance 13C and 15N in plant and soil. Soil ScienceSociety of America Journal 66: 489–498.

Vane CH. 2003. The molecular composition of lignin in sprucedecayed by white-rot fungi (Phanerochaete chrysosporium andTrametes versicolor) using pyrolysis-GC–MS and thermochemolysiswith tetramethylammonium hydroxide. International Biodeteriora-tion and Biodegradation 51: 67–75.

Vane CH, Drage TC, Snape CE, et al. 2005. Decay of cultivatedapricot wood (Prunus armeniaca) by the ascomycete Hypocreasulphurea, using solid state 13C NMR and off-line TMAH thermo-chemolysis with GC-MS. International Biodeterioration and Bio-degradation 55: 175–185.

Vane CH, Drage TC, Snape CE. 2003. Biodegradation of oak(Quercus alba) wood during growth of the shiitake mushroom(Lentinula edodes): A molecular approach. Journal of Agriculturaland Food Chemistry 51: 947–956.

Vane CH, Drage TC, Snape CE. 2006. Bark decay by the white-rotfungus Lentinula edodes: polysaccharide loss, lignin resistance andthe unmasking of suberin. International Biodeterioration andBiodegradation 57: 14–23.

Vane CH, Kim AW, Moss-Hayes V, et al. 2013. Degradation ofmangrove tissues by arboreal termites (Nasutitermes acajutlae) andtheir role in the mangrove C cycle (Puerto Rico): chemicalcharacterization and organic matter provenance using bulk d13C,C/N, alkaline CuO oxidation-GC/MS, and solid-state. Geochemis-try, Geophysics, Geosystems 14: 3176–3191.

Vane CH, Martin SC, Snape CE, et al. 2001. Degradation of lignin inwheat straw during growth of the oyster mushroom (Pleurotusostreatus) using off-line thermochemolysis with tetramethylammo-nium hydroxide and solid-state 13C NMR. Journal of Agriculturaland Food Chemistry 49: 2709–2716.

Veres DS. 2002. A comparative study between loss on ignition andtotal carbon analysis on minerogenic sediments. Studia Universita-tis Babes-Bolyai, Geologia XLVII: 171–182.

Vos PC, de Wolf H. 1993. Diatoms as a tool for reconstructingsedimentary environments in coastal wetlands; methodologicalaspects. Hydrobiologia 269–270: 285–296.

Walker D. 1970. Direction and rate in some British post-glacialhydroseres. In Studies in the Vegetational History of the BritishIsles, Walker D, West RG (eds). Cambridge University Press:Cambridge; 117–139.

Waller MP, Binney HA, Bunting MJ, et al. 2005. The interpretationof fen carr pollen diagrams: pollen-vegetation relationships withinthe fen carr. Review of Palaeobotany and Palynology 133: 179–202.

Waller MP, Grant MJ. 2012. Holocene pollen assemblages fromcoastal wetlands: differentiating natural and anthropogenic causesof change in the Thames estuary, UK. Journal of QuaternaryScience 27: 461–474.

Watcham EP, Shennan I, Barlow NLM. 2013. Scale considerations inusing diatoms as indicators of sea-level change: lessons fromAlaska. Journal of Quaternary Science 28: 165–179.

Wilkinson KN, Scaife RG, Sidell J. 2000. Environmental and sea-levelchanges in London from 10 500 BP to the present: a case studyfrom Silvertown. Proceedings of the Geologists’ Association 111:41–54.

Wilson GP, Lamb AL. 2012. An assessment of the utility of regionaldiatom-based tidal-level transfer functions. Journal of QuaternaryScience 27: 360–370.

Wilson GP, Lamb AL, Leng MJ, et al. 2005a d13C and C/N aspotential coastal palaeoenvironmental indicators in the Merseyestuary, UK. Quaternary Science Reviews 24: 2015–2029.

Wilson GP, Lamb AL, Leng MJ, et al. 2005b Variability of organicd13C and C/N in the Mersey Estuary, U.K. and its implications forsea-level reconstruction studies. Estuarine Coastal and Shelf Sci-ence 64: 685–698.

Wolaver TG, Hutchinson S, Marozas M. 1986. Dissolved andparticulate organic carbon in the North Inlet Estuary, SouthCarolina: what controls their concentrations? Estuaries 9: 31–38.

Woodroffe SA, Long AJ. 2010. Reconstructing recent relative sea-level changes in West Greenland: local diatom-based transferfunctions are superior to regional models. Quaternary International221: 91–103.

Woodworth PL, Teferle FN, Bingley RM, et al. 2009. Trends in UKmean sea level revisited. Geophysical Journal International 176:19–30.

Zong Y, Horton BP. 1998. Diatom zones across intertidal flats andcoastal saltmarshes in Britain. Diatom Research 13: 375–394.

Zong Y, Horton BP. 1999. Diatom-based tidal-level transfer functionsas an aid in reconstructing Quaternary history of sea-level move-ments in the UK. Journal of Quaternary Science 14: 153–167.

Zong Y, Yu F, Huang G, et al. 2010. Sedimentary evidence of LateHolocene human activity in the Pearl River Delta, China. EarthSurface Processes and Landforms 35: 1095–1102.



Documents

The application of δ13C, TOC and C/N geochemistry to … · 2015-10-24 · The application of d13C, TOC and C/N geochemistry to reconstruct Holocene relative sea levels and paleoenvironments