source, diagenesis, and fluxes of particulate organic carbon … publications/1066... · source,...

15
Source, diagenesis, and uxes of particulate organic carbon along the western Adriatic Sea (Mediterranean Sea) T. Tesi a, b, , L. Langone a , M. Giani c , M. Ravaioli a , S. Miserocchi a a ISMAR-CNR UOS di Bologna Istituto di Scienze Marine P. Gobetti 101 40129 Bologna Italy b ITM Department of Applied Environmental Science Stockholm University via Svante Arrhenius väg 8 SE-11418 Stockholm Sweden c OGS Istituto Nazionale di Oceanograa e di Geosica Sperimentale Dipartimento di Oceanograa Biologica via Auguste Piccard 54 34151 Trieste Italy abstract article info Article history: Received 21 April 2012 Received in revised form 24 February 2013 Accepted 2 March 2013 Available online 15 March 2013 Communicated by G.J. de Lange Keywords: Adriatic Sea carbon cycling organic carbon ux and burial diagenesis fossil terrigenous and marine carbon In this study, we investigated the modern organic carbon (OC) cycling along the clinoform-shaped deposit that developed after the attainment of the modern sea-level in the Adriatic Sea (~5.5 kyr cal BP). Newly acquired data were combined with published results to characterize the (i) origin, (ii) diagenesis, and (iii) uxes of OC along the Adriatic clinoform. δ 13 C, Δ 14 C, and lignin phenols were used to constrain the composition of OC accumulating in surface sediments. Sediment cores collected at different water depths were used to describe the early diagenesis during burial in different regions. In addition, on the basis of an extensive number of accumulation rates and OC data, we assessed the ux of OC to the seabed and its burial. Our results showed that terrigenous OC is the dominant OC source in the Po prodelta mainly in the form of pre-aged soil-derived OC and vascular plant fragments. Along the clinoform, both Δ14C and the concentration of lignin-derived phenols decreased with increasing distance from the Po prodelta indicating the inuence of an additional pool of aged OC that gradually becomes more important because of its selective preservation during the sediment transport. As a result, degradation rates (k) decreased along the clinoform as a function of the sediment oxidative history. The calculated half-life of reactive OC (t 1/2 ) was ~14.6 yrs in the Po prodelta whereas topset/forest deposits south of this region exhibited higher values, ~100 yrs, indicating the presence of refractory material. In the distal bottomset region, the t 1/2 was particularly high ranging from ~255 to ~912 yrs. Because of the signicant southward component of the sediment transport, the OC deposition in the southern surface sediments exceeded the local OC input via rivers (ratio deposition/input 1.2). Conversely, the northern Adriatic was characterized by a marked imbalance (ratio deposition/input 0.30.5). According to our calculations, the OC ux to the seabed along the clinoform was ~309 Gg of C per year whereas the OC burial was ~180 Gg of C per year, corresponding to an overall burial efciency of ~59%. © 2013 Elsevier B.V. All rights reserved. 1. Introduction Clinoform-shaped deposits are ubiquitous sedimentological bodies of modern continental margins, including both carbonate and silicoclastic platforms. They formed after the attainment of the modern sea level high-stand (mid-late Holocene) when river outlets and shoreline migrated landward. Their shape and thickness are affected by a series of factors including relative sea level, sediment supply, depositional regime, and sediment type (Pirmez et al., 1998). Typical clinoforms developing along continental margins consist of a prograding body capped by aggrading topsets that become thinner upwards. Clinoforms formed over the last few thousands of years were described along the inner shelf of diverse settings: tectonically passive margins, such as the Amazon prodelta (Nittrouer et al., 1986), active margins, such as the GangesBrahmaputra setting (Goodbred et al., 2003) and several epicontinental-shelves (Alexander et al., 1991). As clinoform-shape deposits are essential building blocks of the inll of sedimentary basins (Mitchum et al., 1977; Vail et al., 1977), they are sites of intense organic carbon (OC) deposition and account for a signicant fraction of the OC burial in the ocean during intergla- cial periods. In addition to the high deposition rates, the OC burial in these deposits is promoted by the relatively low reactivity of the land-derived material being diagenetically pre-altered and matrix- protected against degradation (Mayer, 1994; Mead and Goñi, 2008). Furthermore, hypopycnal coastal plumes experience intense new primary productivity constituting another pool of organic biomass accumulating along the clinoform body (Lohrenz et al., 1990; Campanelli et al., 2011). However, in high energy environments, some clinoforms can act as efcient incinerators where OC burial is limited by the prolongated residence of particles in reuxing suboxic mobile mud (e.g. Fly river delta, Gulf of Papua; (Aller and Blair, 2004). In this biogeochemical study, we focused on sigmoidal clinoforms that are generally associated with low-energy environments (Pirmez Marine Geology 337 (2013) 156170 Corresponding author at: ITM, Department of Applied Environmental Science, Stockholm University, Svante Arrhenius väg 8, SE-11418 Stockholm, Sweden. Tel.: +46 8 674 7245. E-mail addresses: [email protected], [email protected] (T. Tesi). 0025-3227/$ see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.margeo.2013.03.001 Contents lists available at SciVerse ScienceDirect Marine Geology journal homepage: www.elsevier.com/locate/margeo

Upload: doancong

Post on 15-Feb-2019

217 views

Category:

Documents


0 download

TRANSCRIPT

Marine Geology 337 (2013) 156–170

Contents lists available at SciVerse ScienceDirect

Marine Geology

j ourna l homepage: www.e lsev ie r .com/ locate /margeo

Source, diagenesis, and fluxes of particulate organic carbon along the westernAdriatic Sea (Mediterranean Sea)

T. Tesi a,b,⁎, L. Langone a, M. Giani c, M. Ravaioli a, S. Miserocchi a

a ISMAR-CNR UOS di Bologna Istituto di Scienze Marine P. Gobetti 101 40129 Bologna Italyb ITM Department of Applied Environmental Science Stockholm University via Svante Arrhenius väg 8 SE-11418 Stockholm Swedenc OGS Istituto Nazionale di Oceanografia e di Geofisica Sperimentale Dipartimento di Oceanografia Biologica via Auguste Piccard 54 34151 Trieste Italy

⁎ Corresponding author at: ITM,Department ofApplied EnUniversity, Svante Arrhenius väg 8, SE-11418 Stockholm, Sw

E-mail addresses: [email protected], tom

0025-3227/$ – see front matter © 2013 Elsevier B.V. Allhttp://dx.doi.org/10.1016/j.margeo.2013.03.001

a b s t r a c t

a r t i c l e i n f o

Article history:Received 21 April 2012Received in revised form 24 February 2013Accepted 2 March 2013Available online 15 March 2013

Communicated by G.J. de Lange

Keywords:Adriatic Seacarbon cyclingorganic carbon flux and burialdiagenesisfossilterrigenous and marine carbon

In this study, we investigated the modern organic carbon (OC) cycling along the clinoform-shaped depositthat developed after the attainment of the modern sea-level in the Adriatic Sea (~5.5 kyr cal BP). Newlyacquired data were combined with published results to characterize the (i) origin, (ii) diagenesis, and(iii) fluxes of OC along the Adriatic clinoform. δ13C, Δ14C, and lignin phenols were used to constrain thecomposition of OC accumulating in surface sediments. Sediment cores collected at different water depthswere used to describe the early diagenesis during burial in different regions. In addition, on the basis of anextensive number of accumulation rates and OC data, we assessed the flux of OC to the seabed and its burial.Our results showed that terrigenous OC is the dominant OC source in the Po prodelta mainly in the form ofpre-aged soil-derived OC and vascular plant fragments. Along the clinoform, both Δ14C and the concentrationof lignin-derived phenols decreased with increasing distance from the Po prodelta indicating the influence of anadditional pool of aged OC that gradually becomes more important because of its selective preservation duringthe sediment transport. As a result, degradation rates (k) decreased along the clinoform as a function of thesediment oxidative history. The calculated half-life of reactive OC (t1/2) was ~14.6 yrs in the Po prodeltawhereas topset/forest deposits south of this region exhibited higher values, ~100 yrs, indicating the presenceof refractory material. In the distal bottomset region, the t1/2 was particularly high ranging from ~255 to~912 yrs. Because of the significant southward component of the sediment transport, the OC deposition inthe southern surface sediments exceeded the local OC input via rivers (ratio deposition/input 1.2). Conversely,the northern Adriatic was characterized by a marked imbalance (ratio deposition/input 0.3–0.5). According toour calculations, the OC flux to the seabed along the clinoformwas ~309 Gg of C per year whereas the OC burialwas ~180 Gg of C per year, corresponding to an overall burial efficiency of ~59%.

© 2013 Elsevier B.V. All rights reserved.

1. Introduction

Clinoform-shaped deposits are ubiquitous sedimentological bodiesof modern continental margins, including both carbonate andsilicoclastic platforms. They formed after the attainment of the modernsea level high-stand (mid-late Holocene) when river outlets andshoreline migrated landward. Their shape and thickness are affectedby a series of factors including relative sea level, sediment supply,depositional regime, and sediment type (Pirmez et al., 1998).Typical clinoforms developing along continental margins consist of aprograding body capped by aggrading topsets that become thinnerupwards. Clinoforms formed over the last few thousands of yearswere described along the inner shelf of diverse settings: tectonicallypassive margins, such as the Amazon prodelta (Nittrouer et al., 1986),

vironmental Science, Stockholmeden. Tel.: +46 8 674 [email protected] (T. Tesi).

rights reserved.

active margins, such as the Ganges–Brahmaputra setting (Goodbred etal., 2003) and several epicontinental-shelves (Alexander et al., 1991).

As clinoform-shape deposits are essential building blocks of theinfill of sedimentary basins (Mitchum et al., 1977; Vail et al., 1977),they are sites of intense organic carbon (OC) deposition and accountfor a significant fraction of the OC burial in the ocean during intergla-cial periods. In addition to the high deposition rates, the OC burial inthese deposits is promoted by the relatively low reactivity of theland-derived material being diagenetically pre-altered and matrix-protected against degradation (Mayer, 1994; Mead and Goñi, 2008).Furthermore, hypopycnal coastal plumes experience intense newprimary productivity constituting another pool of organic biomassaccumulating along the clinoform body (Lohrenz et al., 1990;Campanelli et al., 2011). However, in high energy environments,some clinoforms can act as efficient incinerators where OC burial islimited by the prolongated residence of particles in refluxing suboxicmobile mud (e.g. Fly river delta, Gulf of Papua; (Aller and Blair, 2004).

In this biogeochemical study, we focused on sigmoidal clinoformsthat are generally associated with low-energy environments (Pirmez

157T. Tesi et al. / Marine Geology 337 (2013) 156–170

et al., 1998; Cattaneo et al., 2007). In particular, we characterized themodern accumulation, degradation, and burial of OC along thelate-Holocene sigmoid in the Western Adriatic Sea (MediterraneanSea) (Fig. 1). This sedimentary body consists of a mud wedge recog-nizable on seismic profiles as a progradational unit lying on top ofthe maximum flooding surface that marks the time of maximumlandward shift of the shoreline attained around 5.5 kyr cal BP(Correggiari et al., 2001; Cattaneo et al., 2003). Along shore-normalsections, this mud wedge exhibits quasi flat topsets in shallow watersand a gradual increase in the slope of foreset beds (0.5°, Correggiari etal., 2001) (Fig. 1b,c). In the last two decades, several projects have in-vestigated sediment dynamics and organic geochemistry along theAdriatic mud wedge (PRISMA 1, PALICLAS, EURODELTA, US/EUEuroSTRATAFORM, PASTA, VECTOR) (Langone et al., 1996; Eppingand Helder, 1997; Giani et al., 2001; Giordani et al., 2002; Ravaioliet al., 2003; Frignani et al., 2005; Palinkas and Nittrouer, 2006; Tesiet al., 2006; Miserocchi et al., 2007; Palinkas and Nittrouer, 2007;Tesi et al., 2007, 2008a; Giani et al., 2009, 2010; Tesi et al., 2011;Weltje and Brommer, 2011). All these studies increased our under-standing of strata formation and organic matter cycling in thisepicontinental margin. With this study, our overarching goal was tocombine the results gained during these projects with newly acquireddata to (i) further characterize the origin of sedimentary OC accumu-lating in surface sediments, (ii) investigate the reactivity of OC duringburial along the Adriatic muddy deposit, and (iii) assess fluxes toseabed and burial of OC along the uppermost strata. Furthermore, be-cause the accumulation is not necessarily linked to a specific river dueto sediment transport along the shelf (Frignani et al., 2005; Palinkasand Nittrouer, 2006; Cattaneo et al., 2007), another important goalof this study was to characterize the spatial distribution of OC deposi-tion and burial along the Adriatic clinoform. Our study benefited froman extensive number of radionuclide-based sediment accumulationrates (210Pb and 137Cs) and numerous biogeochemical data of surfacesediments and sediment cores. Newly acquired data were obtained tofill up gaps in the existing biogeochemical dataset in order to have asynoptic overview of the Adriatic mud-wedge. Specific details regard-ing the source of the data used in this study will be provided in thetext.

2. Background

2.1. The late-Holocene mud wedge

The Adriatic shelf has three major features: (1) epicontinentalmargin characterized by a microtidal regime, (2) clastic sources mainlylocated on the western side where a series of rivers discharge intothe sea (i.e., line-source system, Palinkas and Nittrouer, 2006) and(3) thermoaline cyclonic circulation (Poulain, 2001) that transportssediments southwards along the Italian coast (Fain et al., 2007)(Fig. 1a).

As a result of these forcings, a continuous belt of deltaic andshallow-marine deposits forms the late-Holocence mud wedgealong the western Adriatic shelf (Fig. 1b). This shallow depositreaches up to 35 m in thickness (north of the Gargano promontory)and encompasses three connected depositional elements (Cattaneoet al., 2003): (a) the Po delta system, (b) the central Apenninefine-grained deposit fed by numerous steep rivers characterized byhigh sediment yields (Apennine Rivers), and (c) the Gargano sub-aqueous delta in the southern region, away from any direct riverinput. In volume this deposit is the major component of thelate-Holocene Highstand System track formed after the attainmentof the present sea level highstand (ca. 5.5. cal kyr BP, Correggiari etal., 2001). On seismic reflection profiles (Fig. 1c) the late-Holocenemud wedge exhibits a clinoform-shaped architecture characterizedby three distinctive elements: (1) “topset” beds, shallow andlow-angle deposits, (2) “foreset” beds, the central and steepest strata

characterized by relatively high accumulation, and (3) “bottomset”beds, gently inclined strata in the deepest region of the clinoform.Thickness and slope of these elements vary along the mud wedgebased on environmental forcings such as oceanographic conditions,sediment supply, and accommodation space (Cattaneo et al., 2003).Sedimentation outside the clinoform is negligible and not recogniz-able using seismic profiles.

Anthropogenic- and climate-induced changes affected bothgrowth and internal architecture of the Adriatic clinoform throughoutthe late-Holocene (Cattaneo et al., 2003). In its initial stage, after theattainment of the present sea level highstand, clinform progradationwas relatively low. Changes in pollen abundances showed at leasttwo significant intervals of deforestation since the late Bronze Age(ca. 3700 yrs before present) that resulted in increased soilerosion and sediment supply to the shelf. Subsequently, significantprogradation occurred during the Little Ice Age (ca. 500–100 yrs BP)because of considerable precipitations that characterized this period(Cattaneo et al., 2003). Since War World II, natural and artificial sub-sidence, riverbed excavation, and increased reservoir constructionshave resulted in slowing down the rate of progradation. In spite ofthe hydropower management, sediment supply to the Po prodelta isstill highly episodic because of flood events that ensure a rapid supplyand deposition of land-derived material in the coastal ocean(Palinkasand Nittrouer, 2006; Wheatcroft et al., 2006). Conversely, the sedi-ment delivery along the Apennine stretch suffered particularly fromreservoir constructions and discharge regulation. As a result, most ofthe upper sediments lack laminated beds typical of event-driven de-position (Palinkas and Nittrouer, 2006; Wheatcroft et al., 2006).According to a recent sediment budget (Frignani et al., 2005), roughlyone-fourth of the material enters the Adriatic Sea via the Po River(12.2 Tg of sediments). The remaining material is supplied by north-ern rivers draining the eastern Alps (3.2 Tg of sediments) and short,steep rivers draining the Apennine Mountains (29.7 Tg of sediment)(Fig. 1a).

3. Datasets

3.1. Data from the literature

3.1.1. Sediment accumulation rates (SARs) and mass accumulation rates(MARs)

Three recent 100-yr sediment budgets along the late-Holocenemud wedge were carried out by Frignani et al. (2005) and Palinkasand Nittrouer (2006, 2007). All budgets are based on radioisotopegeochronology (mainly 210Pb and 137Cs). In this study we combinedthese datasets to gain a better spatial data distribution. The final col-lection of accumulation rates is composed of 231 point measurementsspread all over the western continental shelf (Fig. 2a).

3.1.2. Biogeochemical data of surface sedimentsA significant number of surface sediments were collected along

the Adriatic shelf in the last decade. Tesi et al. (2006, 2007) and(Ogrinc et al., 2005) analyzed the elemental composition of surfacesediments collected along the western and northern Adriatic shelf.Additional published data of surface sediments collected in variousregions of the Adriatic margin (Epping and Helder, 1997; Giani etal., 2001; Giordani et al., 2002; Miserocchi et al., 2007; Giani et al.,2009, 2010) were added to obtain a final dataset of 383 measure-ments of OC (Fig. 2b). All sediment samples used in this study werecollected via either box-corer or light gravity core built to preserve in-tact the sediment/water interface. Most of the surface sediments(258) are 1 cm thick (0–1 cm sediment interval; Epping and Helder,1997; Tesi et al., 2006; Miserocchi et al., 2007; Tesi et al., 2007). Therest of the surface sediments encompass the uppermost 2 cm(0–2 cm sediment interval). Finally, some of the aforementionedstudies presented additional biogeochemical data including carbon

158 T. Tesi et al. / Marine Geology 337 (2013) 156–170

159T. Tesi et al. / Marine Geology 337 (2013) 156–170

stable isotopes (δ13C), radiocarbon measurements (Δ14C), and con-centration of lignin phenols (Tesi et al., 2007, 2008a). These resultswere combined with newly radiocarbon measurements to furthercharacterize the origin of the OC accumulating along the clinoform.

3.1.3. Biogeochemical data sediment coresPublished down-core profiles of OC were used to characterize the

reactivity of OC in the bottomset beds of the clinoform (water depth>140 m) (Trincardi et al., 1996; Asioli et al., 2001). The location ofthese sediment cores is shown Fig. 3 (cores PAL94-8 and CM92-43).

3.1.4. Po River samplesδ13C, Δ14C, and biomarkers data of suspendedmaterial collected in

the Po river during a flood event (April 2009) were obtained by Tesiet al. (2011). Bulk and biomarkers analyses were carried out onboth coarse (>63 μm) and fine (b63 μm) fractions.

3.2. Newly acquired data

3.2.1. Surface sedimentsThere are relatively many radiocarbon data of surface sediments

collected in the Po prodelta (Tesi et al., 2008a). Conversely, at thebest of our knowledge, any radiocarbon measurements were carriedout on particulate OC along the Adriatic mud wedge. For this reason10 surface sediments collected along the topset region of the Adriaticclinoform were selected for radiocarbon measurements (Fig. 3a). Theanalytical procedure will be described in the Method section. Onthese sediment samples, several analyses were already performed in-cluding δ13C and biomarker measurements such as lignin phenols(Tesi et al., 2007).

3.2.2. Sediment coresTo describe the reactivity of OC during burial in shallow marine

sediments (water depth b50 m), 7 new sediment cores were selectedand analyzed to measure the OC content. These cores were collectedusing gravity corer in different areas of the Adriatic mud-wedgeincluding the Po prodelta as well as topset and forest strata of theApennine clinoform (Fig. 3, cores E25, D18, N19, ANA35, GW37,2 F47/1, 2 F28/1). Sediment and mass accumulation rates (SARs andMARs) of these cores were obtained by Frignani et al. (2005) andPalinkas and Nittrouer (2006, 2007).

3.2.3. River samplesTo better constrain the composition of the material entering the

Adriatic Sea, 4 suspended samples of the Po River and 4 surface sedi-ments of Apennine Rivers (Reno, Esino, Sangro, and Biferno) wereanalyzed (Fig. 1a). Po River samples were collected in November2011 during high river discharge (flow ranged from ~2500 to~5500 m3 s−1). Samples were taken at Pontelagoscuro gauging station(~90 km from the sea). Samples of Apennine Rivers were collected insummer 2004 during low discharge. The location of these rivers isshown in Fig. 1a. In addition, to investigate the chemical heterogeneityof autochthonous OC, 2 operationally-defined chemical and physicalfractions were obtained from each river sample according to themethods developed by Song et al. (2002) (kerogen//black carbon mix-ture) and Gustafsson et al. (2001) (pyrogenic carbon). Specific analyt-ical details will be given in the next section. OC and δ13C were carriedout on both bulk material and isolated fractions.

Fig. 1. (a) Map of the study area. Arrows show the surface circulation. Open symbols show theand Biferno from north to south) (b) Adriatic Sea (dashed line in panel a). The contour maps shthe Adriatic clinoform (A, B, C, D, E). (c) Chirp profile along a shore normal transect (solid line insurface (mfs).

4. Method

4.1. Elemental composition and carbon stable isotopes

The elemental composition of all samples used in this study wasdetermined via an elemental analyzer. Organic carbon values arereported as percent relative to the dry weight. Prior to analysis, inor-ganic carbon was removed via acidification with HCl. For specificdetails concerning the analysis (e.g., elemental analyzer model, acidconcentration, standards, etc.) in existing data sets see the referencesin the Table 1. In this study, we used a similar analytical method(Nieuwenhuize et al., 1994) to characterize the elemental composi-tion of sediment cores and river samples. OC was measured at theIstituto di Scienze Marine (ISMAR-CNR) using a FISONS NA2000Element Analyzer. Prior to analysis, carbonates were removed viadissolution with 1.5 M HCl. Values are reported as percent of thedry mass weight. The average standard deviation, determined byreplicate analyses of the same sample, was ±0.07 and ±0.01%(1 SD) for carbon and nitrogen, respectively.

All δ13C published results presented in this study were obtained atthe Istituto di Scienze Marine (ISMAR-CNR). New sediment cores andriver samples were analyzed using the same method (Tesi et al.,2008a). Carbon stable isotope composition was determined using aFINNIGAN Delta Plus mass spectrometer, which was directly coupledto the FISONS NA2000 EA by means of a CONFLO interface for continu-ous flowmeasurements. Uncertainties were lower than±0.2‰ (1 SD),as determined from routine replicate measurements of the IAEA refer-ence sample IAEA-CH7 (polyethylene, −31.8‰ vs Vienna-PDB).Data are reported as parts per thousand (‰) relative to Vienna-PDBstandard.

4.2. Radiocarbon measurements (Δ14C)

All radiocarbon analyses were carried out by the National OceanSciences Accelerator Mass Spectrometry Facility (NOSAMS, WoodsHole Oceanographic Institution). Briefly, CO2 samples were obtainedby the combustion of bulk OC from pre-acidified sediments. Subse-quently, the CO2 was purified cryogenically and converted to graphiteusing hydrogen reduction with an iron catalyst. The graphite wasthen pressed into targets, which were analyzed on the acceleratoralong with standards and process blanks. Oxalic Acid II (NIST-SRM-4990C) was the primary standard used for all 14C measurements(Olsson, 1970). Radiocarbon results are expressed as parts per thou-sand (‰) relative to the standard (Stuiver and Polach, 1977).

4.3. Organic matter isolation

The organic matter delivered by rivers encompasses organic resi-dues in various stages of decay and reactivity which are the resultsof biological/physical processes and geological cycles. In this study,two fractionation techniques were used to isolate two pools of carbonknown for having relatively low degradation rates compared to bulkmaterial (Shindo, 1991; Middelburg et al., 1999; Schmidt andNoack, 2000; Burdige, 2007). The first method is a wet chemical pro-cedure and it was used to isolate a kerogen/black carbon (BC) mixturevia acid demineralization, solvent extraction, and base hydrolysis(Song et al., 2002). The second method was used to obtain pyrogenicBC (highly condensed products and residues of incomplete combus-tion of biomass and fossil fuel) through thermal oxidation and HCltreatment (Gustafsson et al., 2001).

location of Apennine Rivers where surface sediments were collected (Reno, Esino, Sangro,ow the thickness of the late-Holocene mud deposit. Polygons display 5 sub-regions alongFig. 1b) (image courtesy of Dr. A. Correggiari). Dashed line shows the maximum flooding

Fig. 2. (a) Collection of mass and sediment accumulation rates along the Adriatic clinoform (filled area). (b) Collection of surface OC content data along the Adriatic clinoform (filledarea).

160 T. Tesi et al. / Marine Geology 337 (2013) 156–170

4.3.1. Fraction 1: kerogen/BC mixtureKnown amounts of finely ground samples were placed in pretared

Teflon tubes with HCl (6 M) to remove the inorganic carbon (60 °Covernight). After the digestion, residues were centrifuged to removethe supernatant. Solid residues were then demineralized at 60 °C for20 h in a concentrated HCl (6 M) and HF (22 M) mixture (ratio1:2 v/v). The remaining residue was centrifuged and rinsed withMilliQwater. The solid phase was then left in HCl (6 M) for 10 h to removeminerals formed during the demineralization process (e.g. fluorite).Prior to the solvent extraction, samples were rinsed three times withMilliQ water and oven-dried at 55 °C. Solvent extraction was

performed using a methanol and methylene chloride mixture (ratio1:2 v/v) in a sonicator bath for 30 min. After the extraction sampleswere centrifuged to remove the supernatant. The extraction was re-peated until the supernatant became colorless. The non-extractablema-terial was then oven-dried at 55 °C. Base extraction was performedusing a NaOH solution (0.5 M, N2 gas-purged) for 12 h. After the ex-traction, samples were centrifuged and the supernatant removed. Thebase extraction was repeated several times until the supernatant be-came colorless. Finally, the solid residue was rinsed for three timeswith MilliQ water and oven-dried at 55 °C. Prior to biogeochemicalanalyses, samples were ground to homogenize the kerogen/BC mixture.

Fig. 3.Map of sediment cores and surface sediments collected along the (a) Adriatic shelf and (b) Po Prodelta. The figure also shows the radiocarbon age of surface sediments (0–1)cm. Radiocarbon data off the Po river mouth refer to samples collected in April 2002 (a subset from Tesi et al., 2008a, 2008b).

161T. Tesi et al. / Marine Geology 337 (2013) 156–170

4.3.2. Fraction 2: pyrogenic BCAbout 50 mg of finely ground samples were weighed into

pre-tared porcelain crucibles. Thermal oxidation was carried out at375 °C for 20 h in the presence of excess of air. Samples were thenreweighed to assess the amount of material loss (non pyrogenic or-ganic matter). Sub-samples (~10 mg) were weighed into Ag cupsand acidified with HCl (1.5 M) to remove the carbonate fraction.Acidified samples were then oven-dried to remove the excess ofacid prior to elemental and carbon stable isotope analyses.

4.4. Down-core degradation

The degradation of OC (i.e., proportion of bulk OC removed byearly diagenetic processes) below the mixing surface layer (SML)was studied in sediment cores collected in several regions of theAdriatic mud-wedge (Fig. 3) according to the diagenetic model intro-duced by (Berner, 1980):

OC zð Þ ¼ OC0 � e−kz=w þ OCburied ð1Þ

where OC(z) is the sediment-normalized concentration of OC at anydepth z, OC0 is the initial concentration of reactive OC, and OCburiedis the concentration of non-reactive OC; k is the first-order degrada-tion rate constant (yr−1) and w is the sediment accumulation rate(cm yr−1).

On the basis of the OC concentration in surface sediments (z =0.5 cm), we assessed the OC content at t = 0 using Eq. (1) that corrects

Table 1Summary of sedimentological and biogeochemical data used in this study.

Sample Measurement

Sediment cores Mass accumulation ratesPo River suspended materials δ13C, Δ14C, and concentration of lignin-derived phSurface sediments OC content

δ13C, Δ14C, and concentration of lignin-derived phSediment cores OC content

for mixing processes that occur in the uppermost cm (0–1 cm). If thethickness of the mixing layer was well known, it would be possible tocorrect for any homogenizing effect using Eq. (1). However, becausethe thickness of the surface mixing layer for all OC data (383 measure-ments) is not known, we preferred to limit our correction to the topcm rather than introducing an error that would derive from an uncon-straint variable. As a result, at those sites were the surface mixinglayer is >1 cm, the OC0 can be underestimated. However, it is importantto understand that the sediment accumulation rate exerts first-ordercontrol on the OC flux magnitude. This is because the sedimentaccumulation can vary several orders of magnitude compared to theOC that instead exhibits a narrower range of values. Therefore, theerror associated with the lack of mixing layer data is relatively smallcompared to the error associated with mass accumulation rateestimates.

4.5. Assessment of fluxes and budgets

Sediment fluxes along the Adriatic mud-wedge were calculatedbased on mass accumulation rates published by Frignani et al.(2005) and Palinkas and Nittrouer (2006, 2007) (Fig. 2a). Calculationswere carried in MATLAB using the GRIDDATA function. Griddata fits asurface in non-uniformly spaced data based on triangle-based linearinterpolation. Griddata interpolates this surface at the points speci-fied by uniformly spaced matrices (Long and Lat matrices) to producea new matrix Z. The new surface always passes through these points.Long and Lat matrices form an uniform grid whose cells are

Reference

Frignani et al. (2005); Palinkas and Nittrouer (2006, 2007)enols This study; Tesi et al. (2011)

Epping, and Helder (1997); Giani et al. (2001, 2009, 2010);Miserocchi et al. (2007); Ogrinc et al. (2005); Tesi et al. (2006, 2007);

enols This study; Tesi et al. (2007, 2008a, 2008b)This study; Asioli et al. (2001); Trincardi et al. (1996).

Fig. 4. Example of data interpolation and calculations performed in the Po prodelta. Black areas correspond to Not-a-Number value (NaN).

162 T. Tesi et al. / Marine Geology 337 (2013) 156–170

1 km × 1 km in the UTM projection (zone 33 T WGS84) (Fig. 4). Forthe sediment budget, prior to data fitting, original mass accumulationdata (g cm−2 yr−1) were converted in g km−2 yr−1. Thus, the cu-mulative sum of all cells in the grid corresponds to the annual sedi-ment deposition. The surface OC distribution along the Adriatic shelfwas estimated using the same approach and maintaining the samemeshgrid (i.e., same Lat and Long matrixes). The source of the OCdata used to perform the interpolation was previously discussed(Fig. 2b).

A schematic representation and an example of calculationsperformed in the Po prodelta are given in Fig. 4. After the interpola-tion, both sediment accumulation and OC content matrices were fil-tered to leave only those cells inside the polygonal region whosevertices are specified by the latitude and longitude of the raw data.Outside points were set as NaN (i.e., Not-a-Number) (blank areas).As a result, the shape of these Z matrixes is slightly different onceplotted in the UTM projection. Finally, the OC flux to the seabed wascalculated multiplying the matrix of sediment accumulation rates bythe matrix of sediment-normalized OC data (Fig. 4). Cells of thisthird matrix contained values only when accumulation and OC datageographically overlapped. If this condition was not satisfied, thecell value was set as NaN (blank in Fig. 4). In addition, in order to ac-count for the degradation, prior to calculations, the OC contents ofsurface sediments were corrected to estimate the OC content at thesediment/water interface. Specific details about this correction willbe given in the text. As for the sediment budget, the cumulativesum of all elements of the OC flux matrix corresponds to the OC annu-ally accumulating in the seabed.

The accuracy of data interpolation was assessed via the analysisof the root-mean of squared residuals (Root Mean Square Error,RMSE). RMSE is a statistical method used to describe how wellmodeled data represent the observed data (Weber and Englund,1994). Among the interpolation methods applied, triangle-basedlinear interpolation had the lower RMSE. In addition, in ordercompare interpolated data having different units, the RMSE was

normalized to the mean of observed values to obtain the percentRMSE (PRMSE).

5. Results and discussion

The next sections are intended to address the following points inline with the objectives of the paper: (i) source, (ii) degradation,and (iii) flux and burial of OC along the Adriatic clinoform. Newlyacquired data were merged with previously published study and,therefore, results and discussion will be presented together to bettercontextualize the new data with respect to published results.

5.1. OC source

An early attempt to characterize the source of the organic materialaccumulating in the Adriatic sea was carried out by Faganeli et al.(1994) on the basis of a two-end member mixing model between13C depleted terrigenous organic matter and 13C enriched marinephytoplankton. Based on this model, the authors inferred that theconcentration of terrigenous OC in the Adriatic Sea increases west-wards because of the influence of the Po and Apennine rivers alongthe Italian coast. A similar approach was used by Miserocchi et al.(2007) who investigated the OC composition in surface sedimentsand flood deposits in the Po prodelta using elemental composition(TN/OC) and carbon stable isotopes (δ13C). This study revealed thatthe Po sediments are dominated by terrigenous material whereasthe phytoplankton is a minor fraction. Subsequently, Tesi et al.(2007) specifically addressed the OC composition along westernHolocene clinoform including both Po prodelta and Apenninemud-wedge. The authors assessed the source of the organic materialpresent in 30 surface sediments using a 4-end member mixingmodel based on δ13C, TN/OC, and concentration of lignin-derivedphenols. According to this model, terrigenous material is the domi-nant OC source in the Po prodelta mainly in the form of lignin-rich or-ganic material. South of the Po region, the lignin content decreases

163T. Tesi et al. / Marine Geology 337 (2013) 156–170

whereas the δ13C increases with increasing distance from theprodelta suggesting the increment of the autochthonous OC as thematerial moves southwards (Tesi et al., 2007). Subsequently, Tesi etal. (2008a) analyzed the 14C-age, lignin content, and sedimentgrain-size along a shore-normal transect in the Po prodelta. Basedon these analyses, the authors inferred that the OC composition inthe surface Po prodeltaic deposits is largely terrigenous dominatedand it can be described as a mixing between pre-aged soil OC, mainlyassociated with the inorganic matrix of fine particles (b63 μm), andyoung vascular plant fragments abundant in the coarse fraction(>63 μm) (Fig. 5).

To further investigate the composition of organic matter in surfacesediments south of the Po prodelta, in this study we carried out addi-tional radiocarbon analyses on a few, well characterized sedimentsamples previously analyzed by Tesi et al. (2007) to determine theconcentration of terrigneous biomarkers such as lignin phenols.These new radiocarbon analyses revealed that the 14C-age of sedi-mentary OC further decreases along the Apennine mud-wedge(Δ14C = −328 ± 35‰) (Figs. 3 and 5). This indicate that the OCcomposition of surface sediments along the Apennine clinoform ismore complicated than a simple dilution/replacement of pre-agedterrigenous material with modern phytodetritus as previouslythought (Faganeli et al., 1994; Miserocchi et al., 2007; Tesi et al.,2007). Instead, the 14C vs lignin plot (Fig. 5) suggests the presenceof a third pool of aged OC (and lignin-free) in addition to pre-agedsoil-derived OC. By contrast, the contribution of modern marine OC

Fig. 5. Biogeochemical composition of Po river suspended sediments (triangles), Po prodelta(diamonds). The composition of 4 potential OC sources (vascular plant fragments, soil OC, mthe samples. Source of the data: lignin data (Tesi et al., 2007, 2008a), radiocarbon data in thstudy).

(Δ14C > 0‰) seems to be extremely low. If the concentration offresh phytodetritus gradually increased with distance from the Poprodelta as suggested by lignin and δ13C data (Tesi et al., 2007), weshould have measured an increase of 14C values as observed inlate-spring along the southern Adriatic slope. In this region, the com-position of suspended material collected via sediment traps (~25 mabove the seabed) alternates periods characterized by significant lat-eral advection of aged OC from outer-shelf deposits with periodsinfluenced by the export of modern marine phytodetritus from theupper water column (Tesi et al., 2008b) (Fig. 5).

Sedimentation along the western Adriatic results from a series ofdistributed fluvial sources (i.e., a line source system, Po and Apenninerivers, Fig. 1a) that feed the clinoform and therefore it is likely thatthis aged material found in surface sediments enters the Adriatic viarivers in the form of refractory OC pools (e.g.. bedrock-derived fossilcarbon, black carbon, aged organic geopolymers, etc.). Subsequently,throughout the sediment transport pathway, the refractory materialis likely to be better preserved compared to the labile OC (Zonneveldet al., 2010). As a result the refractory, 14C-depleted fraction becomesrelatively more important as diagenesis progresses. Considering thatstudies in surface marine sediments suggest the presence of aged al-lochthonous pools of carbon having a marine-like δ13C signature(Dickens et al., 2004), we questioned whether the enrichment of δ13Cvalues observed south of the Po prodelta by Tesi et al. (2007) couldbe the result of the selective preservation of allochthonous refractoryorganic material. To further investigate this aspect, in this study we

surface sediments (square), Apennine sediments (circles), and sediment trap samplesarine phytoplankton, and fossil OC) are plotted to visually show their contribution to

e Po prodelta (Tesi et al., 2008a), radiocarbon data along the Adriatic mud-wedge (this

164 T. Tesi et al. / Marine Geology 337 (2013) 156–170

separated 2 operationally-defined chemical and physical refractoryfractions from 4 suspended samples collected in the Po river(November 2011 flood event) and 4 surface sediments from the Apen-nine rivers (see Tesi et al., 2006). The locations of these rivers are plot-ted in Fig. 1a. Organic fractions were isolated according to methodsdeveloped by Song et al. (2002) and Gustafsson et al. (2001) (seeMethod section for further details). The first procedure was used to iso-late a kerogen/black carbon (BC) mixture through several steps includ-ing acid demineralization, solvent extraction, and alkaline hydrolysis.The second method was used to obtain pyrogenic BC through thermaloxidation and HCl treatment. Fig. 6 and Table 2 show the bulk carbonisotope composition of these 8 river samples and relative isolated frac-tions. Significant differences were observed between bulk samples andseparated fractions for all Po river samples. Pyrogenic BC exhibited themost 13C enriched isotopic composition (−23.9 ± 0.8‰) whereas thekerogen/BCmixture (−25.5 ± 0.6‰) exhibited values relatively closerto the bulk material (−27.3 ± 0.2‰). These fractions accounted for~15% and ~58% of the overall OC pool respectively. Similar resultswere observed for the Apennine river samples although trends wereless marked than the Po river samples. The reason for these differencesmight be due to the different nature of the samples analyzed (e.g.,suspended samples vs surface sediments, geographical differences,etc.). Regardless of this discrepancy whose origin falls outside thescope of the paper, our analyses highlight the presence of OC pools,most likely refractory, having an isotopically enriched signature com-pared to the bulk river-borne material. Therefore, caution should beexercised when assessing the relative proportion of allochthonousand autochthonous material in this region if δ13C trends are exclusivelyinterpreted as a simple dilution/replacement of isotopically depletedallochthonous material with 13C enriched marine OC. This is particular-ly important in light of the new 14C data that indicated the influence ofaged OC in surface sediments that most likely has an allochthonousorigin.

In addition to our indirect evidence from the sediments, the drain-age basin feature itself could explain the supply of fossil carbon hav-ing a marine-like δ13C signature, in particular along the Apennine

Fig. 6. Carbon stable isotope composition of bulk OC, black carbon (BC), and kerogen/BCsediments).

mountains that exhibit the highest uplift rates and sediment yieldsof the Italian drainage basin (~1400 tons km−2 yr−1; Cattaneo etal., 2003). The Apennine orogenesis developed through several tec-tonic phases, mostly during the Cenozoic Era and came to a climaxin the Miocene and Pliocene. The majority of geologic units of the Ap-ennines are made up of marine sedimentary rocks that were deposit-ed over the southern margin of the Tethys Sea (Picotti and Pazzaglia,2008). At the present time, outcropped laminated strata are commonalong the Apennine thrust-belt and they are used by climate-paleoscientists to investigate anoxic/hypoxic events that occurred in theancient basin. In the northern Apennine watershed, for instance,(Capozzi and Picotti, 2003) analyzed the biostratigraphy and carbonstable isotope composition of outcropped laminated sapropeliticdeposits and bioturbated strata formed during the Mid-Pliocene(~3 Myr) (Fiumana outcrop). Throughout the outcrop, the authorsindicated the presence of diatom-rich material as well as relativelyenriched isotopic composition in both laminated and bioturbatedstrata (δ13C up to −21.9‰, average −23.6 ± 0.5‰). Therefore, con-sidering the high sediment yields (Cyr and Granger, 2008) and theshort distance from the sea, steep Apennine Rivers might potentiallyand rapidly transfer a significant pool of this sedimentary carbon dur-ing episodic discharge in direct response to precipitation (Palinkasand Nittrouer, 2006). The resulting short transit time could promotethe fast transfer of easily-erodible sedimentary rocks to the sea limit-ing the stabilization in soils as observed in other systems (Blair et al.,2003; Komada et al., 2004; Leithold et al., 2006). Given the lack of hy-drological and biogeochemical data along the Apennine watershed,more work is needed in this direction to better constrain the origin,composition, and age of the organic material supplied to the AdriaticSea particularly in response to event-driven input.

5.2. Down-core OC degradation

The loss of sedimentary OC as a result of diagenetic mineralizationbelow themixing surface layer was studied in sediment cores collectedin several regions of the Adriatic clinoform (Fig. 3) using the diagenetic

mixture from the Po River (suspended samples) and the Apennine Rivers (surface

Table 2Biogeochemical composition of river samples collected in the Po River and Apennine Rivers.

Sample Type River Discharge TSM Bulk river sample Kerogen/BC mixture BC

m3/s mg/l OC (wt.%) δ13C (‰) δ13C (‰) %a δ13C (‰) %a

Po River Po7-Nov-2011

Suspended matter 2625 744.1 1.08 −27.4 −25.9 51.2 −24.8 11.0

Po8-Nov-2011

Suspended matter 3818 943.2 1.35 −27.4 −26.0 54.9 −24.4 17.6

Po9-Nov-2011

Suspended matter 4551 614.2 1.37 −27.1 −24.9 65.6 −23.4 16.9

Po10-Nov-2011

Suspended matter 5459 410.7 1.33 −27.2 −24.9 58.7 −23.1 11.5

Apennine Rivers RenoJune-2005

Surface sediment NA NA 1.09 −27.2 −26.3 29.3 −25.2 10.3

EsinoJune-2005

Surface sediment NA NA 1.22 −26.4 −27.0 26.0 −25.7 17.6

SangroJune-2005

Surface sediment NA NA 1.05 −27.2 −26.0 55.6 −23.9 6.1

BifernoJune-2005

Surface sediment NA NA 1.20 −26.7 −26.8 19.1 −24.6 14.6

NA = not available.TSM = Total suspended matter.BC = Black Carbon.

a Percent refers to the fraction of the total OC.

165T. Tesi et al. / Marine Geology 337 (2013) 156–170

model introduced by Berner (1980). An essential condition of thismodel is the steady deposition of organic matter. However, whereasthe accumulation of sediment along the Apennine is fairly steady(Frignani et al., 2005; Palinkas and Nittrouer, 2007), steady-state con-ditions are not always the case in the Po prodelta wheremost of the de-position is episodic because of floods (Palinkas et al., 2005; Wheatcroftet al., 2006; Tesi et al., 2011, 2012). In addition, the combination of fastprogradation with lobe-switching in the subaqueous delta makessequence stratigraphy complex (Correggiari et al., 2005). At siteswithin the Po prodelta where all these factors are important, 210Pb pro-files show non-steady long-term (decadal) sediment accumulation(Palinkas and Nittrouer, 2007). An example of this is given in Fig. 7where we show two Po prodelta sediment cores (cores D18 and E25)characterized by non-steady state conditions according to published210Pb data (Palinkas and Nittrouer, 2007). In these cores, the OCcontent fluctuates around a mean value until it reaches an older unit(>100 yrs, unsupported 210Pb activity), most likely formed under adifferent depositional setting. In these strata, the OC content suddenlydrops forming a step-like profile. However, in a few locations of thePo prodelta such as at N19 site (Fig. 3b), the OC content shows an

Fig. 7. OC (wt.%) down-core profiles of sediment cores D18 (circles) and E25 (squares)collected in the Po prodelta (see Fig. 3 for the location of sediment cores).

exponential decrease with sediment depth below the surface mixinglayer that suggests relatively steady deposition of river-borne material.The relatively steady-state conditions at this site were also confirmedby 210Pb data (Palinkas and Nittrouer, 2007).

Along the mud-wedge south of the Po prodelta sediment deposi-tion is relatively steady (Palinkas and Nittrouer, 2007). Sedimentcores collected (Fig. 3a) in this region are characterized by a widerange of sediment accumulation rates (from 0.03 to 1.11 cm yr−1;Frignani et al., 2005; Palinkas and Nittrouer, 2007). These sampleswere collected at different water depths including the deepest partof the mud-wedge such as the Mid-Adriatic depression (~250 m).On these cores and core N19 we applied the diagenetic model previ-ously described in the Method section. As a first step, we calculatedthe fraction of reactive OC (OCreact) at any depth as difference be-tween the down-core concentrations of bulk OC and the concentra-tion of non-reactive OC (OCburied). For the region south of the Poprodelta, OCburied was assessed on the basis of the lowermost stratain cores CM92-43 and PAL94-8 (OCburied = 0.48 wt.%) whereas inthe Po prodelta OCburied was calculated from cores N19, E25, andD18 as averaged values of strata having only supported 210Pb activity(>100 yr, OCburied = 0.64%). OC0 was calculated as the intercept atz = 0 of the linear regression between core depth and natural logof OCreact. Then, the natural log of the [OC(z) − OCburied]/OC0 ratioswas plotted against the accumulation time (depth/accumulationrate, i.e. z/w in Eq. (1)), and k (first-order degradation rate) was esti-mated from the slope of the linear relationship displayed by the data(Fig. 8). Sediment cores collected in the shallow waters (b90 m)south of the Po prodelta exhibited a similar slope and, thus, thesesamples were grouped together to obtain an average k value. Finally,down-core profiles of OC were calculated according to the model andplotted with the measured values in Fig. 8 to visually estimate the ac-curacy of modelled data.

The highest k value was observed in the Po prodelta (k =−0.0473 yr−1) indicating that the organic material in this region isrelatively more reactive (t1/2 = ~14.6 yr) compared to the other sites(Fig. 8). Along the Apennine stretch of the Adriatic mud-wedge, kvalues decrease with increasing water depth suggesting that the OCbecomes gradually less reactive with increasing distance from thecoast. The k values ranged from −0.0069 yr−1 (t1/2 = ~100 yr) inthe foreset region to −0.00076 yr−1 (t1/2 = ~912 yr) in the distalbottomset region. The overall trend of k values is consistent with thecomposition of OC accumulating along the Adriatic clinoform previous-ly described using 14C and lignin data (Fig. 5). In the Po prodelta, in

Fig. 8. (a) First-order kinetic decay of OC of sediment cores collected along the Adriatic clinoform (open diamonds core N19, open circles cores 2F47/1, ANA35, 2F28/1, GW37, as-terisks core PAL94-8, open triangle CM92-43). (b), (c), (d), (e), (f), (g), and (h) OC down-core profiles and modeled data represented by open squares. The sediment accumulationrates are displayed as cm yr−1. For the location of these sediment cores see Fig. 3.

166 T. Tesi et al. / Marine Geology 337 (2013) 156–170

spite of the influence of terrigenous material, organic material is rela-tively more reactive than sedimentary OC depositing along the centralmud-wedge where the concentration of refractory OC steadily in-creases because of its selective preservation along the sediment trans-port. As a result, the material that moves along the sediment dispersalsystem and that reaches the bottomset region of the clinoform gradu-ally becomes less reactive. Therefore, the residence time of a particlebefore its burial (i.e., transport-related oxygen exposure time; (Keil etal., 2004)) is likely to be a key element affecting the overall reactivityof organic material as observed in other continental margins such asthe Washington shelf. Along this margin, Keil et al. (2004) suggestedthe lateral transport of oxygen sensitive organic material characterizedby a slow degradation rate. However, it is important to highlight that inour study we compare sediment cores with marked different accumu-lation rates (Fig. 8) and therefore the lack of labile OC in distal stationsmight be potentially biased by the core sub-sampling. Specifically,when the time period, corresponding to the sediment interval ana-lyzed, is significantly higher than the half-life time of a reactive poolof carbon, the lack of this pool cannot be entirely attributed to theloss during the lateral transport as the down-core time resolution is

not sufficient to capture in-situ changes of highly reactive materials.However, our results are consistent with the across- and along-shelfdecrease of benthic respiration rates and oxygen demand calculatedvia oxygen profiles (Giordani et al., 2002) that confirm the presenceof less reactive material with increasing distance from the shorelineas well as from the Po prodelta. It is also important to highlight thatthe trophic conditions in the basin changed through time as a resultof human-induced changes in the watershed (Barmawidjaja et al.,1995). Therefore the contemporary marine carbon flux is relativelyhigher compared to a century ago. This could potentially affect our cal-culations although our 14C results clearly indicate that modern marinecarbon is not the dominant OC source in surface sediments.

5.3. OC fluxes to the seabed and OC burial

The fluxes of sediment and OC were calculated for 5 sub-regions tobe consistent with the previous literature (Frignani et al., 2005; Tesiet al., 2007). The rationale for subdividing the depositional system isthat sediment transport along the clinoform has an important south-ward component due to the overall cyclonic circulation (Poulain,

167T. Tesi et al. / Marine Geology 337 (2013) 156–170

2001). As a result, the sediment deposition along the Adriaticmud-wedge is disconnected from the local river input (Cattaneo etal., 2003; Frignani et al., 2005). Therefore, the comparison betweeninput and accumulation in each sub-region allowed us to characterizethose areas where sediment supply exceeds burial (bypassing zone)and vice versa. The location of these 5 regions is displayed in Fig. 1bwhereas Table 3 shows the relative sediment supply (data fromFrignani et al., 2005) and burial (this study). Analogous calculationswere carried out to quantify the supply of terrigenous OC combiningsediment input with the average OC concentration of river-borneparticles and top soil samples from the drainage basin (this study;Pettine et al., 1998; Jones et al., 2005; Tesi et al., 2011).

Fig. 9a shows the filled contour plot of interpolated sediment ac-cumulation rates. The pattern of sediment deposition is consistentwith previous sediment flux studies carried out by Frignani et al.(2005) and Palinkas and Nittrouer (2006, 2007). The highest sedi-ment accumulation rates were observed in the Po prodelta wheresediments accumulate preferentially in two depocenters. Other im-portant sites of deposition are the regions located south of Ancona(central region) and north of the Gargano promontory (southern re-gion). While the coherence between our results and previous studiesshould not be a surprise as we essentially re-used their sediment ac-cumulation data, our analysis of root-mean of squared residuals indi-cates that the error associated with the interpolation of these datasetswas not negligible. According to our calculations, the PRMSE (percentroot mean square error, see Method section) of mass accumulationrates was ~23%. However, this error can be considered typical giventhe nature of non-uniformly distributed data and it is almost certainthat the interpolation accuracy in the aforementioned sediment bud-get studies was even lower due to the minor number of data used.

Percentage error associated with OC data interpolation was rela-tively low (~13%) compared to the mass accumulation rate interpola-tion. Interpolated surface OC concentrations are highest in the Poprodelta (Fig. 9b). Along the mud wedge, the highest OC contents areprimarily located along the foreset/bottomset regions although thedeepest areas of the mud-wedge exhibited particularly low concentra-tions, likely reflecting the age and the extent of diagenesis of these sur-face sediments as previously described. The fact that topset exhibitsrelatively lower concentrations is likely due to the high-energy condi-tions that characterize these shallow sediments compared to foresetand bottomset sediments (Cattaneo et al., 2007) and that results in acoarser sediment texture (Palinkas and Nittrouer, 2006; Tesi et al.,2006).

Table 3Input, fluxes, and accumulation of sediment and OC in 5 sub-regions along the western Ad

A B C D

NorthernAdriatic

Po prodelta Po-Ancona Ancon

Drainage basin sedimentyield (T km−2 yr−1)

222 201 1400 1400

Sediment input (Tg) 3.2 12.2 16.9 7.Area (km2)a 394 1,792 4,242 5,828Terrigenous OC input (Gg)b 35.2 134.2 185.9 85.

Marine OC input (Gg)c

Sediment deposition (Tg) 2.35 6.64 9.14 14.OC deposition (Gg) 9.4 68.1 68.8 95.OC Burial (Gg) 5.86 37.44 38.55 60.Sediment deposition/input 0.7 0.5 0.5 1.OC deposition/terrigenous OC input 0.27 0.51 0.37 1.Burial effciency (%) 62.1 54.9 56.1 62.

a Area refers to the region along the shelf where OC fluxes to the seabed were calculatedb Calculated based on sediment input (Frignani et al., 2005) and average concentration of

drainage basin (Pettine et al., 1998; Jones et al., 2005; Giani et al., 2009; Tesi et al., 2011).c Calculated based on average primary productivity along the Adriatic coast (Bosc et al.,

Before calculating the annual flux of OC to the seabed, in order toaccount for the degradation occurring in surface sediments, we esti-mated the initial concentration of OC at the sediment/water interface.This was necessary because, for instance, the uppermost cm couldspan from ~1 yr in the Po prodelta (e.g., N19) to ~35 yrs in thedeepest region of the clinoform (e.g., CM92-43) due to differentsediment accumulation rates. Therefore, if diagenesis is not takeninto account in areas characterized by low sediment accumulation,the potential risk is to calculate an annual flux on the basis of surfacesediments that experienced not negligible degradation. With this inmind, the concentration of the reactive OC at the sediment/water in-terface (z = 0 cm, OC0 in Eq. (1)) was assessed using the first-orderkinetic decay model previously used to described the down-core deg-radation. In this model (Eq. (1)), OC(z) is the sediment-normalized OCcontent of surface sediments (z = 0.5 cm) whereas k and OCburiedvalues were defined on the basis of the previous calculations(Fig. 8). For the Po prodelta, k value was obtained from the N19 sed-iment core (Fig. 8b). For shallow clinoform sediments (water depthb90 m) we used the average k value obtained from 6 sedimentcores collected between 20 and 70 m water depth (2F 47/1, ANA35,2 F 28/1, GW37, Fig. 8c,d,e,f) whereas in the deepest part of theclinoforms (water depth >90 m) we used an averaged k value ofcore CM92-43 and PAL98-08 (Fig. 8g,h). Once we obtained the OCconcentrations at the sediment/water interface, data were interpolat-ed and used to calculate the annual flux of OC on the shelf (Fig. 9c)according to the rationale described in the Method section.

Table 3 shows the annual flux to the seabed of sediment and OCfor each subarea. According to our calculations, the northern regionsof the Adriatic shelf are bypassing areas of sediment consistent withcurrent literature (Correggiari et al., 2001; Frignani et al., 2005). Bycontrast, the mud-wedge below Ancona is affected by significant lat-eral advection of river-borne material from the north (deposition ex-ceeds sediment input in regions D and E). This is coherent with theoverall thickness of the Holocene mud-wedge (Fig. 1b) as well as con-sistent with previous calculations carried out by Frignani et al.(2005). The material that does not accumulate in these 5 regions(i.e., total input − total deposition = ~3.5 Tg of sediment account-ing for ~8% of total input) keeps moving southwards and most likelyit feeds the subaqueous Gargano delta that is away from any directriver input (Cattaneo et al., 2003; Fig. 1b). Lastly, the material thatescapes sedimentation in the shelf is laterally exported duringgravity-driven processes, such as dense water cascading, occurringalong the slope in the southern Adriatic (Trincardi et al., 2007;

riatic clinoform.

E Total Reference

a-P.Penna P.Penna-Gargano

1400 Cattaneo et al. (2003)

8 5 45.1 Frignani et al. (2005)2,337 14,593

8 55 496.1 Frignani et al. (2005); Giani et al. (2009);Jones et al. (2005); Pettine et al. (1998);Tesi et al. (2011);

4377.9 Bosc et al. (2004)29 9.19 41.6 This study9 67.4 309.6 This study26 40.72 182.8 This study8 1.8 0.9 This study12 1.23 0.62 This study9 60.4 59.0 This study

(see Fig. 4).OC in suspended river samples, surface river sediments, and surface soil samples in the

2004) and total area where we calculated the OC fluxes to the seabed.

Fig. 9. (a) Sediment fluxes to the seabed (g cm−2 yr−1), (b) OC content in surface sediments (wt.%), and (c) OC flux to the seabed (mg cm−2 yr−1).

168 T. Tesi et al. / Marine Geology 337 (2013) 156–170

Turchetto et al., 2007). The calculated OC flux to the seabed is~309 Gg of OC per year (Table 3) accounting for roughly 62% of theannual allochthonous OC input from the western Italian rivers(corrected for the mass loss due to lateral export) and 7% of the pri-mary productivity (calculated based on the annual average of 4-yearSeaWiFS Time Series of coastal observations, 300 g OC m−2 yr−1,(Bosc et al., 2004) (Table 3). Along the clinoform, we observe amarked difference between input and deposition of OC compared tothe sediment budget indicating that, in addition to lateral advection,the degradation prior to final deposition has an additional importanteffect on the OC accumulation. Only the region located south of Anco-na (regions D and E, Fig. 1b) has a sediment deposition that slightlyexceeds the input, most likely because the material laterally advectedfrom the north counterbalances the loss caused by degradation.

The OC burial in these 5 regions was assessed based on sedimentaccumulation rates and the sediment-normalized concentration ofburied OC (OCburied) that we previously used to characterize the

degradation in sediment cores collected along the mud-wedge(Fig. 8). According to our calculations (Table 3), all regions combinedbury ~180 Gg of OC per year accounting for roughly 37% of the annualallochthonous OC input from the rivers and 4% of the primary produc-tivity (Table 3). The region that exhibits the highest OC burial is locat-ed between Ancona and Punta Penna (region D, ~60 Gg of OC).Considering the significant lateral advection that characterizes thewestern Adriatic, it is clear that sediment focusing promotes the buri-al of organic material in this area. However, in terms of burial efficien-cy, calculated as the ratio between OC burial and flux of OC to theseabed, all 5 regions exhibited similar values (~55–60%). Similarestimates in the Po prodelta were published by Giordani et al.(2002) although the same authors reported lower burial efficiencies(20–30%) in regions along the clinoform characterized by relativelylow sediment accumulation rates. The reason for these differenceslikely lies in the fact that Giordani et al. calculated the OC flux at thesediment/water interface on the basis of high resolution in-situ

169T. Tesi et al. / Marine Geology 337 (2013) 156–170

oxygen profiles (100-μm intervals) and fluxes of buried carbon(i.e., flux = respired + buried). Consequently, whereas their studyinvestigated the diagenesis of the highly reactive material at sub-millimeter scales, our analysis focused on relatively less reactive OCpools having a half-life that range between 101 and 103 yrs (Fig. 8)(see previous paragraph). Therefore, although our k values can accu-rately describe the diagenesis along the top 1–2 m of the sedimentcolumn (Fig. 8), they are probably too low to characterize the diagen-esis of highly reactive OC that occurs at the sediment/water interface.

6. Conclusions

Along the western Adriatic shelf, a shore-parallel muddy clinoformdeveloped after the attainment of the modern sea-level and it repre-sents the major present-day site of OC burial in the Adriatic Sea.Using published sediment accumulation rates, we estimated themodern flux in the surface sediments and (~309 Gg of OC per year)and burial (~180 Gg of OC per year) of OC along this deposit. Most ofthe OC accumulating in the Po prodeltaic sediments has an allochtho-nous origin mainly made up of soil-derived OC and vascular plant frag-ments. Our results show that, as the material moves southwardfollowing the overall cyclonic circulation, the organic material insurface sediments becomes gradually 14C-depleted. This suggestedthe selective preservation of refractory, aged pools of OC along thesediment transport pathway. By contrast, the accumulation of freshmarine organic matter accounts for a minor fraction of the materialdepositing in surface deposits. Likely this labile material is efficientlydegraded at the sediment/water interface whereas the organic materialthat is currently accumulating along the Adriatic mud-wedge is agedand exhibits a relatively high half-life (t1/2) on the order of 101–

103 yrs. In the Po prodelta, despite the influence of terrigenous OC,the organic material is relatively reactive (t1/2 = ~14.6 yr) comparedto the central mud-wedge where the higher concentration of aged OClikely is responsible for the lower degradation rates (t1/2 = ~100 yr).The fraction of refractory material further increases with distancefrom the coast. In the bottomset region, degradation rates are particu-larly low (t1/2 = ~255–912 yr) suggesting that the oxidative exposurehistory controls the overall reactivity of OC depositing along the shelf.Finally, because sediment transport has an important lateral compo-nent due to the overall cyclonic circulation that moves the sedimentsouthwards, fluxes and burial of OC are disconnected from the localsediment input. As a result, in the northern region a significant fractionof organic material is lost because of the sediment transport as well aslow burial efficiency (ratio deposition/input ~0.3-0.5). Conversely, inthe southern region the OC accumulation exceeds the local input(ratio deposition/input ~1.2) because of sediment focusing.

Acknowledgments

We wish to acknowledge the projects which provided data:Prisma1_Fondali1 (Progetto di ricerca e Sperimentazione per laSalvaguardia del Mare Adriatico), STEP (Science and Technology forthe Environmental Protection, CT90-0067), MATER (MAss Transferand Ecosystem Response, contract CT-96-0051), ANEMIE, US andEU_EuroStrataform, PASTA (European STRATA FORMation, Grant No.N00014-03-01-0154 and Contract EVK3-CT-2002-00079) and Ogrincet al., 2005. We also thank Gabriella Rovatti and Fabio Savelli for labassistance. Financial support for this studywasprovided by the interdis-ciplinary Italian government research project VECTOR, sub-action6.1.8 (VulnErabilità delle Coste e degli ecosistemi marini italiani aicambiamenti climaTici e loro ruolO nei cicli del caRboniomediterraneo)and the EU-PERSEUS, WP2 (Policy-oriented marine EnvironmentalResearch in the Southern European Seas, GrantNo. 287600). This is con-tribution number 1786 of CNR-ISMAR of Bologna.

References

Alexander, C.R., DeMaster, D., Nittrouer, C., 1991. Sediment accumulation in a modernepicontinental-shelf setting: the Yellow Sea. Marine Geology 98 (1), 51–72.

Aller, R.C., Blair, N.E., 2004. Early diagenetic remineralization of sedimentary organic C in theGulf of Papua deltaic complex (Papua New Guinea): net loss of terrestrial C and diage-netic fractionation of C isotopes. Geochimica et Cosmochimica Acta 68 (8), 1815–1825.

Asioli, A., Trincardi, F., Lowe, J., Ariztegui, D., Langone, L., Oldfield, F., 2001. Sub-millennial scale climatic oscillations in the central Adriatic during theLateglacial: palaeoceanographic implications. Quaternary Science Reviews 20(11), 1201–1221.

Barmawidjaja, D.M., van der Zwaan, G.J., Jorissen, F.J., Puskaric, S., 1995. 150 years ofeutrophication in the northern Adriatic Sea: evidence from a benthic foraminiferalrecord. Marine Geology (122), 367–384.

Berner, R.A., 1980. Early Diagenesis: A Theoretical Approach. Princeton Univ Pr.Blair, N.E., Leithold, E.L., Ford, S.T., Peeler, K.A., Holmes, J.C., Perkey, D.W., 2003. The per-

sistence of memory: the fate of ancient sedimentary organic carbon in a modernsedimentary system. Geochimica et Cosmochimica Acta 67 (1), 63–73.

Bosc, E., Bricaud, A., Antoine, D., 2004. Seasonal and interannual variability in algalbiomass and primary production in the Mediterranean Sea, as derived from4 years of SeaWiFS observations. Global Biogeochemical Cycles 18 (1), GB1005.

Burdige, D.J., 2007. Preservation of organic matter in marine sediments: controls,mechanisms, and an imbalance in sediment organic carbon budgets? ChemicalReviews-Columbus 107 (2), 467–485.

Campanelli, A., Grilli, F., Paschini, E., Marini, M., 2011. The influence of an exceptionalPo River flood on the physical and chemical oceanographic properties of theAdriatic Sea. Dynamics of Atmospheres and Oceans 52, 284–297.

Capozzi, R., Picotti, V., 2003. Pliocene sequence stratigraphy, climatic trends andsapropel formation in the Northern Apennines (Italy). Palaeogeography,Palaeoclimatology, Palaeoecology 190, 349–371.

Cattaneo, A., Correggiari, A., Langone, L., Trincardi, F., 2003. The late-Holocene Garganosubaqueous delta, Adriatic shelf: sediment pathways and supply fluctuations.Marine Geology 193 (1–2), 61–91.

Cattaneo, A., Trincardi, F., Asioli, A., Correggiari, A., 2007. The Western Adriatic shelfclinoform: energy-limited bottomset. Continental Shelf Research 27 (3), 506–525.

Correggiari, A., Trincardi, F., Langone, L., Roveri, M., 2001. Styles of failure in lateHolocene highstand prodelta wedges on the Adriatic shelf. Journal of SedimentaryResearch 71 (2), 218–236.

Correggiari, A., Cattaneo, A., Trincardi, F., 2005. The modern Po Delta system: lobeswitching and asymmetric prodelta growth. Marine Geology 222, 49–74.

Cyr, A.J., Granger, D.E., 2008. Dynamic equilibrium among erosion, river incision, andcoastal uplift in the northern and central Apennines, Italy. Geology 36, 103–106.

Dickens, A.F., Gélinas, Y., Masiello, C.A., Wakeham, S., Hedges, J.I., 2004. Reburial offossil organic carbon in marine sediments. Nature 427 (6972), 336–339.

Epping, E.H.G., Helder, W., 1997. Oxygen budgets calculated fromin situ oxygenmicroprofiles for Northern Adriatic sediments. Continental Shelf Research 17(14), 1737–1764.

Faganeli, J., Pezdic, J., Ogorelec, B., Misic, M., Hajdek, M., 1994. The origin of sedimenta-ry organic matter in the Adriatic. Continental Shelf Research 14, 365–384.

Fain, A., Ogston, A., Sternberg, R.W., 2007. Sediment transport event analysis on thewestern Adriatic continental shelf. Continental Shelf Research 27, 431–451.http://dx.doi.org/10.1016/j.csr.2005.03.007.

Frignani, M., Langone, L., Ravaioli, M., Sorgente, D., Alvisi, F., Albertazzi, S., 2005. Fine-sediment mass balance in the western Adriatic continental shelf over a centurytime scale. Marine Geology 222, 113–133.

Giani, M., Boldrin, A., Matteucci, G., Frascari, F., Gismondi, M., Rabitti, S., 2001. Down-ward fluxes of particulate carbon, nitrogen and phosphorus in the north-westernAdriatic Sea. The Science of the Total Environment 266 (1–3), 125–134.

Giani, M., Berto, D., Rampazzo, F., Savelli, F., Alvisi, F., Giordano, P., Ravaioli, M., Frascari,F., 2009. Origin of sedimentary organic matter in the north-western Adriatic Sea.Estuarine, Coastal and Shelf Science 84 (4), 573–583.

Giani, M., Rampazzo, F., Berto, D., 2010. Humic acids contribution to sedimentary or-ganic matter on a shallow continental shelf (northern Adriatic Sea). Estuarine,Coastal and Shelf Science 90 (2), 103–110.

Giordani, P., Helder, W., Koning, E., Miserocchi, S., Danovaro, R., Malaguti, A., 2002.Gradients of benthic–pelagic coupling and carbon budgets in the Adriatic andNorthern Ionian Sea. Journal of Marine Systems 33, 365–387.

Goodbred, S.L., Kuehl, S.A., Steckler, M.S., Sarker, M.H., 2003. Controls on facies distri-bution and stratigraphic preservation in the Ganges–Brahmaputra delta sequence.Sedimentary Geology 155 (3–4), 301–316.

Gustafsson, Ö., Bucheli, T.D., Kukulska, Z., Andersson, M., Largeau, C., Rouzaud, J.N.,Reddy, C.M., Eglinton, T.I., 2001. Evaluation of a protocol for the quantification ofblack carbon in sediments. Global Biogeochemical Cycles 15 (4), 881–890.

Jones, R.J.A., Hiederer, R., Rusco, E., Montanarella, L., 2005. Estimating organic carbon in thesoils of Europe for policy support. European Journal of Soil Science 56 (5), 655–671.

Keil, R.G., Dickens, A.F., Arnarson, T., Nunn, B.L., Devol, A.H., 2004. What is the oxygenexposure time of laterally transported organic matter along the Washingtonmargin? Marine Chemistry 92 (1), 157–165.

Komada, T., Druffel, E.R.M., Trumbore, S.E., 2004. Oceanic export of relict carbon bysmall mountainous rivers. Geophysical Research Letters 31 (7), L07504.

Langone, L., Asioli, A., Correggiari, A., Trincardi, F., 1996. Age-depth modelling throughthe late Quaternary deposits of the central Adriatic basin. Memorie Istituto Italianodi Idrobiologia (55), 177–196.

Leithold, E.L., Blair, N.E., Perkey, D.W., 2006. Geomorphologic controls on the age ofparticulate organic carbon from small mountainous and upland rivers. GlobalBiogeochemical Cycles 20 (3), GB3022.

170 T. Tesi et al. / Marine Geology 337 (2013) 156–170

Lohrenz, S.E., Dagg, M.J., Whitledge, T.E., 1990. Enhanced primary production at theplume/oceanic interface of the Mississippi River. Continental Shelf Research 10(7), 639–664.

Mayer, L.M., 1994. Surface area control of organic carbon accumulation in continentalshelf sediments. Geochimica et Cosmochimica Acta 58 (4), 1271–1284.

Mead, R.N., Goñi, M.A., 2008. Matrix protected organic matter in a river dominatedmargin: a possible mechanism to sequester terrestrial organic matter? Geochimicaet Cosmochimica Acta 72 (11), 2673–2686.

Middelburg, J.J., Nieuwenhuize, J., van Breugel, P., 1999. Black carbon in marinesediments. Marine Chemistry 65 (3–4), 245–252.

Miserocchi, S., Langone, L., Tesi, T., 2007. Content and isotopic composition of organiccarbon within a flood layer in the Po River prodelta (Adriatic Sea). ContinentalShelf Research 27 (3), 338–358.

Mitchum, R., Vail, P., Sangree, J., 1977. Seismic stratigraphy and global changes of sealevel, part 6: stratigraphic interpretation of seismic reflection patterns in deposi-tional sequences. Seismic stratigraphy—applications to hydrocarbon exploration.Memoir—American Association of Petroleum Geologists (26), 117–133.

Nieuwenhuize, J., Maas, Y.E.M., Middelburg, J.J., 1994. Rapid analysis of organic carbonand nitrogen in particulate materials. Marine Chemistry 45 (3), 217–224.

Nittrouer, C., Kuehl, S., DeMaster, D., Kowsmann, R., 1986. The deltaic nature of Amazonshelf sedimentation. Geological Society of America Bulletin 97 (4), 444–458.

Ogrinc, N., Fontolan, G., Faganeli, J., Covelli, S., 2005. Carbon and nitrogen isotopecompositions of organic matter in coastal marine sediments (the Gulf of Trieste, NAdriatic Sea): indicators of sources andpreservation.Marine Chemistry 95 (3), 163–181.

Olsson, I.U., 1970. Radiocarbon Variations and Absolute Chronology.Palinkas, C., Nittrouer, C., 2006. Clinoform sedimentation along the Apennine shelf,

Adriatic Sea. Marine Geology 234 (1), 245–260.Palinkas, C., Nittrouer, C., 2007. Modern sediment accumulation on the Po shelf,

Adriatic Sea. Continental Shelf Research 27 (3), 489–505.Palinkas, C.M., Nittrouer, C.A., Wheatcroft, R.A., Langone, L., 2005. The use of Be-7 to

identify event and seasonal sedimentation near the Po River delta, Adriatic Sea.Marine Geology 222, 95–112.

Pettine, M., Patrolecco, L., Camusso, M., Crescenzio, S., 1998. Transport of carbon andnitrogen to the northern Adriatic Sea by the Po River. Estuarine, Coastal andShelf Science 46 (1), 127–142.

Picotti, V., Pazzaglia, F.J., 2008. A new active tectonic model for the construction of theNorthern Apennines mountain front near Bologna (Italy). Journal of GeophysicalResearch 113 (B8), B08412.

Pirmez, C., Pratson, L.F., Steckler, M.S., 1998. Clinoform development by advection–diffusionof suspended sediment: modeling and comparison to natural systems. Journal ofGeophysical Research 103 (B10) (24141–24124,24157).

Poulain, P.-M., 2001. Adriatic Sea surface circulation as derived from drifter databetween 1990 and 1999. Journal of Marine Systems 29, 3–32.

Ravaioli, M., Alvisi, F., Vitturi, L.M., 2003. Dolomite as a tracer for sediment transportand deposition on the northwestern Adriatic continental shelf (Adriatic Sea, Italy).Continental Shelf Research 23 (14–15), 1359–1377.

Schmidt,M.W.I., Noack, A.G., 2000. Black carbon in soils and sediments: analysis, distribution,implications, and current challenges. Global Biogeochemical Cycles 14 (3), 777–793.

Shindo, H., 1991. Elementary composition, humus composition, and decomposi-tion in soil of charred grassland plants. Soil Science & Plant Nutrition 37 (4),651–657.

Song, J., Ping'an Peng, a, Huang, W., 2002. Black carbon and kerogen in soils and sedi-ments. 1. Quantification and characterization. Environmental Science & Technology36 (18), 3960–3967.

Stuiver, M., Polach, H.A., 1977. Reporting of C-14 data—discussion. Radiocarbon 19 (3),355–363.

Tesi, T., Miserocchi, S., Langone, L., Boni, L., Guerrini, F., 2006. Sources, fate and distribu-tion of organic matter on the Western Adriatic Continental Shelf, Italy. Water, Air,and Soil Pollution: Focus 6 (5–6), 593–603.

Tesi, T., Miserocchi, S., Goni, M.A., Langone, L., Boldrin, A., Turchetto, M., 2007. Organicmatter origin and distribution in suspended particulate materials and surficialsediments from the western Adriatic Sea (Italy). Estuarine, Coastal and ShelfScience 73 (3–4), 431–446.

Tesi, T., Langone, L., Goni, M.A., Miserocchi, S., Bertasi, F., 2008a. Changes in the compo-sition of organic matter from prodeltaic sediments after a large flood event (PoRiver, Italy). Geochimica et Cosmochimica Acta 72 (8), 2100–2114.

Tesi, T., Langone, L., Goni, M.A., Turchetto, M., Miserocchi, S., Boldrin, A., 2008b. Sourceand composition of organic matter in the Bari canyon (Italy): dense water cascad-ing versus particulate export from the upper ocean. Deep Sea Research Part I:Oceanographic Research Papers 55 (7), 813–831.

Tesi, T., Miserocchi, S., Goni, M.A., Turchetto, M., Langone, L., De Lazzari, A., Albertazzi,S., Correggiari, A., 2011. Influence of distributary channels on sediment and organicmatter supply in event-dominated coastal margins: the Po prodelta as a study case.Biogeosciences 8 (2), 365–385.

Tesi, T., Langone, L., Goni, M.A., Wheatcroft, R.A., Miserocchi, S., Bertotti, L., 2012. Earlydiagenesis of recently deposited organic matter: a 9-yr time-series study of a flooddeposit. Geochimica et Cosmochimica Acta 83, 19–36.

Trincardi, F., Cattaneo, A., Asioli, A., Correggiari, A., Langone, L., 1996. Stratigraphy ofthe late-Quaternary deposits in the central Adriatic basin and the record ofshort-term climatic events. Memorie Istituto Italiano di Idrobiologia 55, 39–70.

Trincardi, F., Foglini, F., Verdicchio, G., Asioli, A., Correggiari, A., Minisini, D., Piva, A.,Remia, A., Ridente, D., Taviani, M., 2007. The impact of cascading currents on theBari Canyon System, SW-Adriatic Margin (Central Mediterranean). Marine Geology246 (2–4), 208–230.

Turchetto, M., Boldrin, A., Langone, L., Miserocchi, S., Tesi, T., Foglini, F., 2007. Particle trans-port in the Bari Canyon (southern Adriatic Sea). Marine Geology 246 (2–4), 231–247.

Vail, P., Mitchum, R., Thompson, S., 1977. Seismic stratigraphy and global changes ofsea level, part 3: relative changes of sea level from coastal onlap. Seismic stratigra-phy—applications to hydrocarbon exploration. Memoir—American Association ofPetroleum Geologists 26, 63–81.

Weber, D.D., Englund, E.J., 1994. Evaluation and comparison of spatial interpolators II.Mathematical Geology 26 (5), 589–603.

Weltje, G.J., Brommer, M.B., 2011. Sediment-budget modelling of multi-sourced basinfills: application to recent deposits of the western Adriatic mud wedge (Italy).Basin Research 23, 291–308.

Wheatcroft, R.A., Stevens, A.W., Hunt, L.M., Milligan, T.G., 2006. The large-scale distributionand internal geometry of the fall 2000 Po River flood deposit: evidence from digitalX-radiography. Continental Shelf Research 26 (4), 499–516.

Zonneveld, K., Versteegh, G., Kasten, S., Eglinton, T.I., Emeis, K.C., Huguet, C., Koch, B., deLange, G.J., De Leeuw, J., Middelburg, J.J., 2010. Selective preservation of organicmatter in marine environments; processes and impact on the sedimentary record.Biogeosciences 7, 483–511.