Dynamics of particulate organic matter in a coastal system characterized by the occurrence of marine mucilage & A stable isotope study (PDF Download Available)
See all >4 CitationsSee all >68 ReferencesSee all >3 Figures
7.98Université Bordeaux 14.88Insitut des Milieux Aquatiques+ 92.6433.06University of BordeauxShow more authorsAbstractIn coastal systems, particulate organic matter (POM) originates from various autochthonous and allochthonous organic matter sources. Also, some coastal systems are characterized by the occurrence of large amounts of mucilaginous material of biologic origin (i.e. phytoplankton, bacteria), which aggregates and potentially traps other organisms and particles present in the water column. This study focuses on POM origin and spatio-temporal dynamics in the South-East coast of the Bay of Biscay, an area subject to mucilage occurrence. In order to investigate POM quantitative and qualitative (C and N elemental and isotopic ratios) characteristics, sampling was performed over an annual cycle at two sites experiencing different mucilage occurrence and river influence. Contribution of phytoplankton, terrestrial POM and anthropogenic POM to coastal-POM composition was calculated using a three-sources mixing model. Overall, phytoplankton dominated the coastal-POM composition at all seasons, sites and most of the depths (71.6 ± 24.2%). Terrestrial-POM contribution was moderate (22.7 ± 21.8%) and anthropogenic-POM contribution was usually negligible (5.7 ± 7.4%). Both sites mainly exhibited similar vertical and temporal variations in terms of POM origin and dynamics: terrestrial-POM contribution increased with depth and was higher in winter at all depths and in autumn in bottom waters, compared to other seasons. The main differences between both sites were related to the vertical dynamics of the terrestrial contribution to the coastal POM. Horizontal, vertical and temporal variation of POM composition was linked to processes driving the sedimentary hydrodynamics: the river flow, the direction of the river plume and events of sediment resuspension/deposition. During the study period, the mucilage occurred only as flocs (small aggregates). The mucilage was of autochthonous origin and did not trap detectable amount of allochthonous material.Discover the world's research14+ million members100+ million publications700k+ research projectsFigures
Dynamics of particulate organic matter in a coastal system characterizedby the occurrence of marine mucilage –A stable isotope studyCamilla Liénarta,b,?, Nicolas Susperreguic,d, Vanessa Rouaudc,e, Joana Cavalheiroc,e, Valérie Davida,b,Yolanda Del Amoa,b, Robert Duranc,e,BéatriceLaugac,e,MathildeMonperrusc,e, Thierry Pigotc,e,Sabrina Bichona,b, Karine Charliera,b,NicolasSavoyea,baUniv. Bordeaux, Laboratoire d&Environnements et Paléoenvironnements Océaniques et Continentaux (EPOC), UMR 5805, Allée GeoffroySaint-Hilaire, 33615 Pessac Cedex, FrancebCNRS, Laboratoire d&Environnements et Paléoenvironnements Océaniques et Continentaux (EPOC), UMR 5805, Allée Geoffroy Saint-Hilaire, 33615 Pessac Cedex, FrancecUniv. de Pau etdes Pays de l&Adour, Institut des sciences analytiques et de Physicochimie Pour l&Environnement etles Matériaux (IPREM), UMR CNRS 5254, 2 Avenue du Président PierreAngot,64000 Pau, FrancedInstitut des Milieux Aquatiques (IMA), 1 rue de Donzac, 64100 Bayonne, FranceeUniv. de Pau et des Pays de l&Adour, Fédération MIRA, FR4155, Allée du parc Montaury, 64600 Anglet, Franceabstractarticle infoArticle history:Received 28 January 2016Received in revised form 22 July 2016Accepted 3 August 2016Available online 8 August 2016In coastal systems, particulate organic matter (POM) originates from various autochthonous and allochthonousorganic matter sources. Also, some coastal systems are characterized by the occurrence of large amounts of mu-cilaginous material of biologic origin (i.e. phytoplankton, bacteria), which aggregates and potentially traps otherorganisms and particles present in thewater column. This study focuses on POM origin and spatio-temporal dy-namics in the South-Eastcoast of the Bay of Biscay, an area subject to mucilage occurrence. In orderto investigatePOM quantitative and qualitative (C and N elemental and isotopic ratios) characteristics, sampling was per-formed over an annual cycle at two sites experiencing different mucilage occurrence and river influence. Contri-bution of phytoplankton, terrestrial POM and anthropogenic POM to coastal-POM composition was calculatedusing a three-sources mixing model. Overall, phytoplankton dominated the coastal-POM composition at all sea-sons, sites and most of the depths (71.6 ± 24.2%). Terrestrial-POM contribution was moderate (22.7 ± 21.8%)and anthropogenic-POM contribution was usually negligible (5.7 ± 7.4%). Both sites mainly exhibited similarvertical and temporal variations in terms of POM origin and dynamics: terrestrial-POM contribution increasedwith depth and was higher in winter at all depths and in autumn in bottom waters, compared to other seasons.The main differences between both sites were related to the vertical dynamics of the terrestrial contribution tothe coastalPOM. Horizontal, vertical and temporal variation of POM composition was linked to processes drivingthe sedimentary hydrodynamics: the river flow, the direction of the river plume and events of sediment resus-pension/deposition. During the study period, the mucilage occurred only as flocs (small aggregates). The muci-lage was of autochthonous origin and did not trap detectable amount of allochthonous material.(C) 2016 Elsevier B.V. All rights reserved.Keywords:Particulate organic matterStable isotopesMarine mucilageCoastal systemsBay of Biscay1. IntroductionCoastal systems are defined as the portion of global ocean wherephysical, biological and biogeochemical processes are directly affectedby land (Gattuso and Smith, 2007). Coastal zones are among the mostproductive ecosystems of the planet: although they cover only 5% ofthe surface area of the global ocean, they account for about 12% of oce-anic primary production, 71% of organic matter burial (Dunne et al.,2007) and offer 60 times more ecosystem services per surface areaunit than the open ocean (de Groot et al., 2012). Especially, these sys-tems act as dynamical land/sea interfaces by regulating numerousfluxes of matter and energy. Coastal systems are under the influenceof different sources of organic matter (OM) (Bode et al., 2006). Indeed,the pool of particulate organic matter (POM) in coastal systems is a mix-ture of organic particles as living organisms and/or detritus originatingfrom autochthonous (in situ pelagic and benthic primary producers)as well as allochthonous (continental) reservoirs (Savoye et al., 2012;Tesi et al., 2007). All these POM sources potentially fuel the coastalJournal of Sea Research 116 (2016) 12–22?Corresponding author at: Univ. Bordeaux, Laboratoire d&Environnements etPaléoenvironnements Océaniques et Continentaux (EPOC), UMR CNRS 5805, StationMarine d&Arcachon, 2 rue du Pr. Jolyet, 33120 Arcachon Cedex, France.E-mail addresses: camilla.lienart@u-bordeaux.fr (C. Liénart),ima.susperregui@wanadoo.fr (N. Susperregui), vanessa.rouaud@univ-pau.fr (V. Rouaud),joana.cavalheiro@univ-pau.fr (J. Cavalheiro), valerie.david@u-bordeaux.fr (V. David),yolanda.del-amo@u-bordeaux.fr (Y. Del Amo), robert.duran@univ-pau.fr (R. Duran),beatrice.lauga@univ-pau.fr (B. Lauga), mathilde.monperrus@univ-pau.fr (M.Monperrus),thierry.pigot@univ-pau.fr (T. Pigot), sabrina.bichon@u-bordeaux.fr (S. Bichon),karine.charlier@u-bordeaux.fr (K. Charlier), nicolas.savoye@u-bordeaux.fr (N. Savoye).http://dx.doi.org/10.1016/j.seares.1/(C) 2016 Elsevier B.V. All rights reserved.Contents lists available at ScienceDirectJournal of Sea Researchjournal homepage: /locate/seares
food webs (Bode et al., 2006). Indeed, primary producers and detritalorganic matter contribute to a large part of ‘energy channels’in foodwebs (Moore et al., 2004) leading to consider POM as one of the struc-turing compartments for coastalecosystemfunctioning. In order tobet-ter understand ecological and biogeochemical functioning of coastalsystems, it is thus essential to determine theorigin and ideally quantifythe relative contribution of these sources to the POM pool in thesesystems.Worldwide in the ocean, exacerbated and evolved stages of marinesnow (Fogg, 1995) appearwithin the water column, forming large mu-cilaginous aggregates of various sizes and shapes (Bongiorni et al., 2007;Giani et al., 1992; Stachowitsch et al., 1990). This phenomenon waspar-ticularly reported in the Mediterranean sea, mainly during summermonths (Bongiorni et al., 2007; Giani et al., 1992; Precali et al., 2005;Rinaldi et al., 1995). These mucilaginous aggregates are composed of or-ganic matter, together with a significant inorganic fraction (Giani et al.,2005), and are supposed to be initially generated by different aquaticmicroorganisms such as phytoplankton or bacteria (Azam et al., 1999;Giani et al., 2005; Najdek et al., 2002; Pompei et al., 2003). Biologicaland physico-chemical factors are responsible for mucilaginous aggre-gate occurrence, but this phenomenon also depends on the nature ofthe aggregating organic matter, environmental conditions of the siteof formation and transformation during aging (Giani et al., 2005).When mucilage aggregates, it creates microhabitats that host variousorganic particles and organisms (bacteria, phytoplankton,zooplankton,marine and terrestrial detritus) as well as inorganic particles (Ki?rboe,2000; Passow, 2002; Silver et al., 1984; Simon et al., 2002). Mucilagi-nous aggregates are thus an important vector for vertical carbon exportto deep ocean (Passow, 2002; Passow et al., 2001) but also have an im-pact on food web and plankton community structures (Passow, 2002;Riley, 1963; Verdugo et al., 2004).Many tools exist to characterize and quantify the POM compositionin coastal systems. Among them, mass, element, and isotope ratios arethe more conventional. Particulate organic carbon to chlorophyll aratio (POC:Chla) discriminates living phytoplankton(POC:Chla≤100 g·g-1) or phytoplankton-dominated POM(POC:Chlab200 g·g-1) from other material (POC:ChlaN200 g·g-1).C:N ratio discriminates heterotrophs (C:N ≈3–6 mol·mol-1)fromphytoplankton (C:N ≈6–10 mol·mol-1) and from terrestrial detritus(C:NN12 mol·mol-1). Carbon isotope ratio (δ13C) discriminates conti-nental POM (i.e. terrestrial POM, freshwater phytoplankton and/or an-thropogenic POM; δ13Cb-26‰) from marine POM (δ13CN-24‰)in temperate systems (Savoye et al., 2003, and references therein). Ni-trogen isotope ratio (δ15N) of POM exhibits no universal values but itis largely used, in combination to the above ratios, to characterize andquantify POM composition in coastal systems (e.g. Berto et al., 2013).Quantification of POM composition is usually performed using a mixingmodel, which is a system of mass balance equations based on the aboveparameters (e.g. Dubois et al., 2012).Within this context, the aims of the present study are 1) to charac-terize POM composition and its spatio-temporal dynamics over an an-nual cycle in a coastal system subject to mucilage occurrence, thesoutheastern Bay of Biscay, and 2) to investigate the interaction be-tween POM and mucilage using isotopic and elemental tools. Only fewstudies investigated POM-mucilage interaction using this approach(Faganeli et al., 2009; Giani et al., 2006) and it is the first study in thisoceanic region.2. Material and methods2.1. Study areaThe study area is located in the coastal zone of the South-East of theBay of Biscay (Fig. 1). In this area, mucilage occurrence has been report-ed for fifteen years by fishermen (d&Elbée et al., 2016). Because of geo-morphology characterized by no embayment, strait coastline andsandy sea floors, this area is homogenized by alongshore oceanic cur-rents (Ferrer et al., 2009). As a general functioning in the Bay of Biscay,the water circulation over the continental shelf is mainly controlled bywinds and thus highly variable over the seasons (Koutsikopoulos andLe Cann, 1996). In the South-East of the Bay of Biscay, freshwater inputsmainly come from the Adour River (mean annual flow of ca. 300 m·s-1,Puillat et al., 2004;Stoichev et al., 2004). The Adour River plume mainlyflows toward the south, south-west (Petus et al., 2014; Sagarminaga etal., 2005). However, it is a very reactive system and its direction is main-ly controlled by river discharge rates and modulated byseasonal preva-lence of winds (Ferrer et al., 2009; Petus et al., 2014; Petus et al., 2010).The southern coastal waters are exposed to the Adour River plume(Petus et al., 2014). Exposure increases under light wind conditions(Dailloux, 2008), whereas under south or south-west wind conditionsand when river discharge increases, the plume exposure is toward thenorth and attached to the coast (Petus et al., 2014). Low-salinity watersdue to the Adour River discharge can be observed 15–20 km off thecoast line and down to 50 m depth (Ferrer et al., 2009).2.2. Sampling strategyTwo study sites were selected in the coastal zone of this area: &Biar-ritz& site (off the city of B 1°36′1″W, 43°29′14″N) and &Tarnos&site (off thecity of T1°32′8″W, 43°32′52″N) (Fig. 1), located re-spectively 3 km and 1.8 km from the shore, and 8 km South and 2.7 kmNorth of the Adour River mouth. Biarritz site is more influenced by theAdour plume than Tarnos site (Petus et al., 2010; Sagarminaga et al.,2005). At both sites, sampling was performed at three depths in surface(-1 m), intermediate (ca. -8 m) and bottom (ca. --23 m) waters.At both sites, coastal water was sampled on a monthly basis fromJune 2013 to June 2014. A more intensive sampling (ca. every threedays) was performed at Biarritz site, which is more subject to mucilageoccurrence (Susperregui, unpublished data), during targeted periods ofmucilage occurrence in spring and autumn. Also, sampling of mucilagi-nous material was performed at Biarritz site during these periods fromOctober 2013.In order to determine the origin and composition of costalPOM, pos-sible sources (i.e. end-members) of POM were also sampled (Fig. 1):surface riverine water was monthly sampled at Urt Village in theAdour R water of three sewage treatment plants (STP) effluentswas sampled downstream the effluent treatment before dilution intosurrounding environment. STP 1 (Saint Frédéric) collects industrial ef-fluent whereas STP 2 (Pont l&aveugle) and 3 (Saint Bernard) collecturban effluents. The three STP flow directly into the Adour Estuary.Sampling of STP was performed during the two official operatingmodes defined according to the flow entering the STP: the dry-weatheroperating mode (D; all the inflow is treated), which is the usual mode,and the wet-weather operating mode (W; when the inflow is higherthan an upper limit, the additional inflow is only pretreated and thenby-passed to the river). During extreme rain events, part of the rainwater is directly by-passed. Such events were very rare during thestudy period and never appeared during the sampling dates. Since‘pure’phytoplankton (i.e. phytoplankton cells without miscellaneousparticles) cannot be sampled or extracted from bulk POM, its signaturewas estimated from our dataset considering that POM of low POC:Chlaratio is phytoplankton-dominated (see Section 4.2.1).Seawater, freshwater, STP water and mucilaginous material weresampled for particulate organic carbon (POC), particulate nitrogen(PN) and isotopic ratios (δ13Candδ15N). Seawater and freshwater,were also sampled for chlorophyll a(Chla). Water temperature and sa-linity were measured in situ at all sites and dates.2.3. Data base, sampling, processing and storage proceduresDaily river-flow data were provided by Banque Hydro (www.hydro.eaufrance.fr/). Temperature and salinity were measuredusing a HORIBA13C. Liénart et al. / Journal of Sea Research 116 (2016) 12–22
U53G probe. Seawater was sampled using a 20 L Niskin bottle andstored inthe dark in a pre-cleaned container beforesub-sampling for fil-tration. Freshwater Adour River was sampled 10 m far from the shorefrom the end of a pontoon, with a 5 L pre-cleaned bottle attached to apole. After-treatments effluent from STP was sampled from the 24 h re-frigerated composite sampler with 1 L pre-cleaned glass bottles. Themucilaginous material was collected thanks to its sticky property: itwas aggregated using a net (200 μm mesh size) in order to collectenough material for isotopic and elemental ratios. The net was rinsedwith seawater. 1 L of water from net collector was poured in a pre-cleaned container, stored in the dark and kept cold before filtration.Back to the laboratory, water samples were gently filtrated throughpre-combusted (4 h - 450 °C) GF/F filters (47 mm?) for POC, PN, C:Nratio, δ13C and δ15N. All parameters were analyzed on the same filter.Another GF/F filter was used for Chla.Afterfiltration, coastal and river-ine filters were rinsed using pre-filtered in situ water and STP filtersusing Q-water. Chlafilters were stored frozen at -80 °C. Filters forthe other parameters were dried overnight at 50 °C and then stored ina desiccator. 1 L of the water collected from the net was filtrated on aNitrex nylon filtration cloth (60 μm meshsize) and the mucilaginous ag-gregates retained on the filter were collected and stored in a cryotube at-80 °C.2.4. Samples analysisChlorophyll awas extracted using 90% acetoneand analyzed by fluo-rescence (Turner Design 10-AU Fluorometer) following Yentsch andMenzel (1963). Filters for the other parameters were decarbonated bycontact with HCl vapour (4 h for seawater filters, 8 h for freshwaterand STP fi Lorrain et al., 2003). One piece of each filter waspunched (11 mm?) and analyzed for POC and PN using an elementalanalyzer(Thermo Finnigan Flash EA 1112 analyzer). The rest of thefilterwas analyzed for isotopic and elemental ratios. Mucilage samples werefreeze-dried, and ground into powder using a ball mill. Powders wereweighed into tin cups for the determination of N isotopic composition,and in silver cups for the determination of C isotopic composition. Thelatter were in-cup decarbonated using 1.2 N HCl (Kennedy et al.,2005). Isotopic and elemental ratios were analyzed using an ElementalAnalyzer (ThermoFisher Scientific Flash 2000) connected to an IsotopeRatio Mass Spectrometer (Isoprime, GV instruments). According to therecommendations of the International Union of Pure and AppliedChemistry (IUPAC), isotopic data are expressed in the conventionaldelta notation (Coplen, 2011):δ13Csample or δ15 Nsample
1/4 Rsample=Rstandard??–1??where R=13C/12Cor15N/14N, and the references were Vienna Pee DeeBelemnite (VPDB) for δ13C and atmospheric N2for δ15N. Analytical un-certainties were ≤0.2‰for δ13C and δ15N and ≤0.2 mol.mol-1for C:Nratio.2.5. Statistical analysis and mixing modelAll statistical analyses were performed with the R software (http://cran.r-project.org/, R development core team 2009). Because normalityand homoscedasticity were not observed forthe isotopic and elementalparameters, all data were analyzed using non-parametric tests.Paired by time non-parametric ANOVA (Friedman test and post-hoctests) were performed to assess differences between depths for eachsite and parameter, and differences between sites for each depth andFig. 1. Location of the studiedsites, Biarritz and Tarnos(diamonds), in the southeasterncoast of the Bay of Biscay (France)and the sites sampled fororganic matter sources(triangles): Urtvillage in the Adour River and three sewage treatment plants (STP 1, 2 and 3) flowing in the Adour estuary.14 C. Liénart et al. / Journal of Sea Research 116 (2016) 12–22
parameter. A Kruskal-Wallis test was performed to assess, for surface andintermediate data pooled by date, the differences in δ13C, δ15N and C:Nratio of POM sampled during and in absence of mucilage occurrence.In order to define seasonal groups over theannual cycle, a constraintagglomerative hierarchical classification (CAHC, chclust function, riojapackage) method, which preserves chronological order of the data,was performed on δ13C, δ15N and C:N data. Surface and intermediatedata at Biarritz and Tarnos were averaged in order to get one singlevalue per sampling date. Then, data were standardized and the Euclide-an distance was used to build the association matrix before applying theCAHC. Groups were defined using a broken-stick model (Bennett, 1996;bstick function, rioja package).To estimate the organic matter composition of coastal POM, each se-lected sources of organic matter was incorporated into a mixing modelusing a Bayesian approach (package SIAR; Parnell et al., 2010). Thismodel was used to quantify the relative contribution of organic mattersources (phytoplankton, terrestrial POM and anthropogenic POM) tocoastal POM composition using three variables (δ13C, δ15NandN:Cra-tios). The absolute uncertainty associated to the model outputs wasusually close to 10%.A qualitative index,based on visual net observations, was defined toqualify mucilage amount: a clear net without mucilage wasqualified as‘nil’, a net with some mucilage was qualified as ‘moderate’,andfinally anet clogged by large amount of brown mucilage was qualified as ‘large’.3. Results3.1. Environmental characteristicsWater temperature varied from ca. 11 to ca. 23 °C over the annualcycle at both sites. As a general pattern, mean temperature was higherfor surface than bottom waters with ca. 1.5 °C difference. For bothsites, salinity was 33.2 ± 1.9 (mean ± standard deviation). Salinity in-creased with depth by about 2 salinity units. Episodically, salinities of28–29 were recorded in surface waters. The Adour River flow was lowfrom August to October with values ranging from 80 to 300 m3·s-1(Fig. 2). Pulsed high river flows occurred from November to late spring.Maximum flow (N3000 m3·s-1) was observed in January.Over the studied annual cycle, the mucilage appeared in June, Sep-tember, October, and December 2013 and from March to May 2014,i.e. mainly during low river flow (Fig. 2) and calm weather conditions(not shown). Duration of the mucilage occurrence varied from fewdays in March 2014 to at least one month in September–October 2013.3.2. Spatio-temporal variability of coastal-POM characteristicsFor most of the studied parameters, there are no significant differ-ences (p-value N0.05) between Biarritz and Tarnos sites, whatever thedepth (Supplementary Table 1a). Only the C:N ratio was significantlydifferent (p-value b0.05) in surface + intermediate waters mainly be-cause of two dates (27/11/2013 and 07/03/2014; SupplementaryTable 2) where Biarritz exhibited higher values than Tarnos (9–10 mol·mol-1vs ca. 7.5 mol·mol-1). Thus, the overall description ofthe results below (Sections 3.2.1 and 3.2.2;Figs. 3 to 5) is valid forboth sites. The only few specific differences are pointed out.3.2.1. Vertical variabilityChlorophyll aconcentration varied usually from 0.1 to ca. 3.5 μg·L-1with only few higher values of 4–6μg·L-1in surface waters and one ex-treme value
values exhibited an overall decrease withdepth (Fig. 3a). POC concentration mainly ranged between 100 and500 μg·L-1with only few higher values in surface and bottom waters(Fig. 3b). Overall, POC concentrations were higher in surface watersthan intermediate and bottom waters (Fig. 3b). Consequently POC:Chlaratios were usually higher in bottom waters than in surface and inter-mediate waters (Fig. 3c). Especially, 47%, 44%, and 22% of the POC:Chlavalues were lower than 200 g·g-1in surface, intermediate and bottomwaters, respectively. δ13C, δ15N and C:N ratios were very variable(-26.5 ≤δ13C≤-19.1‰;1.7≤δ15N≤9.8‰;5.3≤C:N ≤15.5 mol·mol-1,Fig. 3d–f). In average, δ13C values were lower in bottom than surfaceand intermediate waters (Fig. 3d). This was statistically significant (p-value b0.05) for Biarritz site (Supplementary Table 1b). C:N values usu-ally ranged between 6 and 10 mol·mol-1, but were sporadicallthey roughly increased with depth (Fig. 3f).Overall, surface waters were characterized by higher concentrationsof chlorophyll aand POC, and by lower values of POC:Chlaand C:N ra-tios, whereas bottom waters were characterized by lower concentra-tions of chlorophyll aand POC, higher values of POC:Chlaand C:Nratios and lower δ13C values. Intermediate waters usually exhibitedvalues similar to that observed at surface waters or showed intermedi-ate values between surface and bottom waters. These overall trendswith depth were usually significant in Biarritz site. In Tarnos site,these trends were valid for part of the dates (e.g. 12/06//3/2014; Supplementary Tables 1 and 2).3.2.2. Temporal variabilityA constraints agglomerative hierarchical clustering analysis (Supple-mentary Fig. 1) revealed four groups corresponding to four seasons:‘summer’(12/06/13 to 25/09/13), ‘autumn’(27/09/13 to 14/10/13),‘winter’(27/11/13 to 12/03/14) and ‘spring’(14/03/14 to 19/06/14).Each season exhibited different POM characteristics (Fig. 4a–f). In sum-mer, chlorophyll aand POC concentrations were low (with the excep-tion of few high values), the range of POC:Chlavalues was very large(75 ≤POC:Chla≤1380), δ13C exhibited mean values (-23.2 ± 1.4‰),δ15N was high (6.1 ± 1.7‰) and C:N ratio relatively low (7.4 ±1.2 mol·mol-1). In autumn, chlorophyll aand POC concentrationsFig. 2. Adour River flow from 01/06/2013 to 30/06/2014. POM sampling dates are represented by thevertical full lines (dates without mucilage occurrence) and dotted lines (dates withmucilage occurrence).15C. Liénart et al. / Journal of Sea Research 116 (2016) 12–22
were low, POC:Chlavalues were low and exhibited a narrow range(284 ± 131 g·g-1), δ13C(-22.7 ± 1.0‰)andδ15N (5.0 ± 0.8‰)alsoexhibited narrow ranges and intermediate values and C:N ratio(7.0 ± 0.8 mol·mol-1) was the lowest. In winter, chlorophyll aandPOC concentrations were low, the range of POC:Chlavalues was verylarge (71 ≤POC:Chla≤1396), δ13C(-24.8 ± 1.2‰) and δ15N (3.5 ±0.8‰) were low and C:N ratio (9.2 ± 2.5 mol·mol-1) was relativelyhigh. In spring, chlorophyll aconcentration, POC concentration andδ13C(-21.5 ± 1.4‰) were high, POC:Chlaratio was low and exhibiteda narrow range (221 ± 126 g·g-1), and C:N ratio was relatively low(8.0 ± 0.8 mol·mol-1).3.3. Characteristics of OM sources and of mucilageAdour POM (Table 1) showed constant signatures over the annualcycle with, comparatively to other OM sources, low δ13C(-26.5 ±surface intermed. bottom0123456Chla(-1)0 500 POC ( -1)surface intermed. bottom0 500 POC:Chla(g.g-1)surface bottomintermed.a. b. c.130 254616445086-26 -24 -22 -20δ13C /‰surface intermed. bottom2 4 6 8 10δ15N/‰surface intermed. bottom6810 12 14C:N (mol.mol-1)surface intermed. bottomd. e. f.Fig. 3. Concentrations of chlorophyll a(Chl a) (a.) and of particulate organic carbon (POC) (b.), POC:Chlaratio (c.), stable isotopic ratio of carbon (δ13C) (d.) and nitrogen (δ15N) (e.), andelemental carbon to nitrogen ratio (C:N) (f.) for Biarritz and Tarnos sites (pooled data sets) at the three depths (surface, intermediate, bottom). Median (bold line) and mean (crosses)values are shown in the boxplot. Hinges are 25th and 75th percentiles. Whiskers are 5th and 95th percentiles. The values below the vertical arrow are high values not shown on theboxplot scale.681012 14C:N (mol.mol-1)winter springsummer autumnwinter springsummer autumn-26 -24 -22 -20δ13C /‰winter springsummer autumn246810δ15N/‰d. e. f.winter springsummer autumn0123456Chla-1)a.130150005001000POC:Chla(g.g-1)winter springsummer autumnc.50861500winter springsummer autumnb.0500 100025461644POC (-1)Fig. 4. Concentrations of chlorophyll a(Chl a) (a.) and of particulate organic carbon (POC) (b.), POC:Chlaratio (c.), stable isotopic ratio of carbon (δ13C) (d.) and nitrogen (δ15N) (e.), andelemental carbon to nitrogen ratio (C:N) (f.) for Biarritz and Tarnos sites(pooled data sets) for the fourseasons. Median (bold line) and mean (crosses) values are shown in theboxplot.Hinges are 25th and 75th percentiles. Whiskers are 5th and 95th percentiles. The values below the vertical arrow are high values not shown on the boxplot scale.16 C. Liénart et al. / Journal of Sea Research 116 (2016) 12–22
h.C:N (mol.mol -1)Adour River POMphytoplanktonAnthropogenic POM (D)Anthropogenic POM(W)Surface waterIntermediate waterBottom waterδ15N/‰-30 -28 -26 -24 -22 -20 -18 -16 -30 -28 -26 -24 -22 -20 -18 -16-2024681012-30 -28 -26 -24 -22 -20 -18 -16δ15N/‰-30 -28 -26 -24 -22 -20 -18 -16springwinter winterspring-202468101224681012e. f.24681012C:N (mol.mol -1)g.-2024681012-2024681012-30 -28 -26 -24 -22 -20 -18 -16 -30 -28 -26 -24 -22 -20 -18 -16δ15N/‰-30 -28 -26 -24 -22 -20 -18 -16δ15N/‰-30 -28 -26 -24 -22 -20 -18 -16autumnsummer summerautumn24681012C:N (mol.mol -1)a. b.d.24681012C:N (mol.mol -1)δ13C/‰ δ13C/‰c.Fig. 5. Biplots showing δ13C, δ15N and C:N ratio of coastal POM ( Biarritz and Tarnos data sets), terrestrial POM from the Adour River, phytoplankton, and anthropogenic POMsampled in sewage treatment plants functioning in dry (D) and wet (W) modes. Error bars refer to standard deviations.Table 1Estimatedelemental andisotopic signatures of phytoplankton (a.)and measured elemental and isotopic signatures of mucilaginous materialsampled using anet (a.), terrestrialPOM fromthe AdourRiver (b.) and anthropogenic POM from sewage treatment plantssampled during wet (W)and dry (D) operatingmodes (b.). Data are reported as average ± standard deviation.Note that phytoplankton signatures were estimated from coastal-POM data of low POC:Chlaratio (see Section 4.2.1).a. Summer Autumn Winter SpringPhytoplankton δ13C(/‰)-22.3 ± 1.6 -21.0 ± 0.4 -24.4 ± 0.9 -20.6 ± 1.0δ15N(/‰) 5.7 ± 1.5 5.1 ± 0.5 3.4 ± 0.6 3.9 ± 1.4C:N (mol·mol-1) 6.4 ± 0.6 6.6 ± 0.2 7.5 ± 0.6 7.8 ± 0.8Summer Autumn Winter Early spring (March) Late spring (April/May)Mucilaginous material δ13C(/‰)-19.5 ± 0.2 -22.5 -22.5 ± 0.1 -18.6 ± 1.0δ15N(/‰)–5.6 ± 0.1 3.0 3.9 ± 0.3 6.7 ± 0.3C:N (mol·mol-1)–6.5 ± 0.04 4.6 5.6 ± 1.1 8.4 ± 1.9b. Adour River Anthropogenic POM (W) Anthropogenic POM (D)δ13C(/‰)-26.5 ± 0.7 -25.5 -26.3 ± 0.2δ15N(/‰) 4.6 ± 0.9 4.4 6.6 ± 0.6C:N (mol·mol-1) 9.0 ± 0.9 6.1 6.5 ± 0.217C. Liénart et al. / Journal of Sea Research 116 (2016) 12–22
0.7‰), mean δ15N (4.6 ± 0.9‰) and relatively high C:N ratio (9.0 ±0.9 mol.mol-1), except for the POC:Chlaratio that was always highbut also very variable (1300 ± 1200 g.g-1; not shown). AnthropogenicPOM (sewage treatment plants, Table 1)showedlowδ13C and C:N ratio(average ± standard deviation of the two operating modes: -25.9 ±0.5‰and 6.3± 0.3 mol·mol-1, respectively). Anthropogenic δ15N var-ied according to the operating mode: δ15N was high during dry weathercondition (6.6 ± 0.6‰) and low during wet weather condition (4.4‰).Phytoplankton signatures were estimatedfrom coastal-POM dataof lowPOC:Chlaratio (see Section 4.2.1) for each season. Values exhibitedlarge variations over time for δ13C and, to a lesser extent, for δ15N.δ13Crangedfrom-24.2 ± 0.9‰in winter to -20.6 ± 1.0‰in sprδ15Nrangedfrom3.4±0.6‰in winter to 5.7 ± 1.5‰in summer. C:Nratio was less variable: it ranged from 6.4 ± 0.6mol·mol-1in summerto 7.8 ± 0.8 mol.mol-1in spring (Table 1). δ13C, δ15N and C:N values ofmucilaginous material (Table 1) varied over time and were lower inwinter and early spring (δ13C≈-22.5‰;δ15N≈3.5‰;C:N ≈5 mol·mol-1) than in spring and autumn (δ13C≈-19.5‰;δ15N≈6‰;C:N≈6.5–9mol·mol-1).Overall, elementaland/or isotopic signatures of Adour POM, anthro-pogenic POMand phytoplankton were usually differentto each other atall seasons (Fig. 5).4. Discussion4.1. Dynamics of particulate organic matterOne of the main patterns of coastal POM characteristics is an overallsimilarity between Biarritz and Tarnos sites. Indeed, a date-to-date hor-izontal comparison of POM concentrations and signatures reveals theabsence of significant differences between Biarritz and Tarnos sites forPOC and Chlaconcentrations, POC:Chlaratio, δ13C and δ15N (Supple-mentary Table 1). Thus, both sites are likely to have the same generalfunctioning regarding POM origin and dynamics. Thestudied area expe-riences wind-induced alongshore currents that lead to horizontally-ho-mogenized water masses and its associated particles (Ferrer et al., 2009;Fontán et al., 2006; González et al., 2008). Nevertheless, some differ-ences were observed between both sites for C:N ratio at some winterdates in surface and intermediate waters. They were characterized byhigher values at Biarritz than Tarnos sites (Supplementary Table 2).These high C:N values (N9mol·mol-1) in Biarritz site are mainly asso-ciated withlow δ13C(≤ca. -24.5‰). This point out a higher influence ofterrestrial material to the POM composition in Biarritz compared toTarnos sites at these depths. Indeed, the turbid plume of the AdourRiver is usually oriented toward the South, South-West (Petus et al.,2010; Sagarminaga et al., 2005).The vertical dynamics of POM is also similar between both sites withan overall decrease in chlorophyll aand POC concentrations, an overallincrease in POC:Chlaand C:N ratios and an overall decrease in δ13Cwithdepth (Fig. 3). In ocean systems, the vertical decrease in chlorophyll aand POC concentrations as well as increase in POC:Chlaand C:N ratiosare common features (Boyd et al., 1999; Martin et al., 1987). In coastalsystems, the latter indicates phytoplankton remineralization through-out the water column and/or the input of terrestrial POM at depth. In-deed these ratios increase with phytoplankton decay due to apreferential remineralization of chlorophyll aand PN over POC, andboth ratios are high for terrestrial POM (e.g. Savoye et al., 2003,andref-erences therein). However δ13C increases with phytoplanktonremineralization (Savoye et al., 2003) but decreases with terrestrial in-puts (Savoye et al., 2012). In the present study, the δ13C decrease withdepth indicates thus a higher proportion of terrestrial materialin coastalPOM at depth in the studied system.POM characteristics also vary over time following common featuresfor coastal temperate systems (Fig. 4): the contrasted seasons werespring (high concentrations of chlorophyll aand POC, high δ13C, andlow POC:Chlaratio) and winter (relatively low concentrations ofchlorophyll aand POC, lowest δ13C and δ15N, relatively high POC:Chlaand highest C:N ratios). Summer and autumn usually exhibited inter-mediate characteristics but are nevertheless characterized by low con-centration of chlorophyll aand POC, and by the lowest C:N ratio. Thelow C:N ratio associated with intermediate to high POC:Chl aratiomay indicate the presence of bacteria and/or heterotrophs in thewater column. Spring POM reveals overall characteristics of phyto-plankton-dominated POM, whereas winter POM characteristics indicat-ed a higher influence of terrestrial material.Such horizontal, vertical and temporal variations of POM character-istics at local or regional scales are common features in coastal temper-ate systems (Cresson et al., 2012; Tesi et al., 2007). Especially theinfluence of terrestrial POM usually depends on the proximity of theriver mouth and the river flows (Berto et al., 2013; Miller et al., 2013).In the present study, the terrestrial influence characterized by lowδ13C and high C:N ratio was observed at both sites and all depths, butmainly in autumn and winter, and more frequently in bottom than sur-face waters and in Biarritz site than in Tarnos site. Indeed, the Adourplume usually flows toward the South, South-West (Sagarminaga etal., 2005) where is located Biarritz site. As an example,the terrestrial in-fluence was observed throughout the water column at Biarritz site butonly in the bottom water at Tarnos site on the 07/03/14, and thenonly in Biarritz bottom water on the 12/03/14 (Supplementary Table2). This influence was observed after the flood crest of the 05/03/14(Fig. 2). In contrast, the terrestrial influence was also observed duringperiods of low river flow, as in late summer and early autumn (Figs. 2and 5). During these periods, direct export of terrestrial POM from theAdour River does not explain this terrestrial influence to coastal POM.It is likely due to processes of resuspension-deposition of terrestrialPOM deposited on the sediment surface and redirected into the watercolumn by bottom currents. Indeed, alongshore oceanic currents andwind regimes are able to promote horizontal POM exchanges (Tesi etal., 2007) and to induce continuous resuspension-deposition processesof sediment OM in shallow dynamic systems (Dubois et al., 2012).Such processes homogenize sediment organic matter at the systemscale in a shallow semi-enclosed system(Dubois et al., 2012). Resuspen-sion events are illustrated by the highest values of POC concentration inbottom water (Fig. 3b).4.2. Origin and composition of particulate organic matterCoastal POM is most often the result of the mixing of numerous OMsources originating from the continent, and from benthic and pelagicreservoirs (Berto et al.,2013; Tesi et al., 2007). However, outside estuar-ies and river plumes, coastal POM is usually mainly composed of phyto-plankton (Cressonet al., 2012; Tesi et al., 2007). Inthe present study C:Nratio mainly ranged between 6 and 9 mol·mol-1, indicatingthat phyto-plankton should be a major contributor to the coastal POM. Especiallyca. 45% of surface and intermediate coastal POM exhibited POC:Chlalower than 200 g·g-1. Thus coastal POM composition was often domi-nated by phytoplankton. Nevertheless, episodic high C:Nratios associat-ed to low δ13Cindicatedtheinfluence of terrestrial POM to coastal POM(Section 4.1;Figs. 3–5). In order to estimate the contribution of organicmatter sources to the coastal POM composition, a mixing model wasperformed.4.2.1. Choice of the OM sources for running the mixing modelIn coastal systems, OM sources (i.e. end-members) that are com-monly used for running mixing models are riverine POM, that consistsof terrestrial POM and/or freshwater phytoplankton, ‘marine’POM, usu-ally considered as marine phytoplankton, benthic macrophytes,microphytobenthos, sediment organic matter, anthropogenic POMorig-inating from treated/untreated effluents of sewage treatment plants,etc. (e.g. Berto et al., 2013; Cresson et al., 2012).Sediment organic matter being itself a mixture composed of theother above-cited OM sources (e.g. Dubois et al., 2012), it was not18 C. Liénart et al. / Journal of Sea Research 116 (2016) 12–22
considered as an end-member in the present study. Also, because thewater column of the study sites was ca. 30 m deep and because of thestrong bottom currents and waves and the sandy nature of the sedi-ment, benthic primary producers likely were not able to grow there.Thus, no benthic primary producers were considered as end-members.Pure phytoplankton end-member could not be sampled or extractedfrom bulk particulate organic matter. Thus, elemental and isotopic sig-natures of in situ phytoplankton could not be measured. In many stud-ies, authors used the signature of off-shore POM as a proxy of coastalphytoplankton signature. This assumes that 1) off-shore POM is only/mainly composed of phytoplankton, and 2) the signatures of off-shoreand coastal phytoplankton are similar in space and time. This assump-tion is usually not valid and has led to erroneous outputs of mixingmodels and ecological misinterpretations (Miller and Page, 2012). An-other option for estimating phytoplankton elemental and isotopic sig-nature is to consider that POM of low POC:Chlais a proxy ofphytoplankton. Indeed phytoplankton exhibits low POC:Chlaratio (ca.40–140 g·g-1) and it is considered that the POM of POC:Chlaratiolower than 200 g·g-1is dominated by phytoplankton (Savoye et al.,2003, and references therein). Thus it was considered here that the ele-mental and isotopic ratios of the POM of POC:Chlaratio lower than200 g·g-1was the best estimate for the elemental and isotopic ratiosof phytoplankton. Such an approach has already been used in previousstudies (e.g., Dubois et al., 2014; Dubois et al., 2012; Savoye et al.,2012). In temperate systems, phytoplankton δ13Candδ15N usually ex-hibit large variation over an annual cycle because of the large variationof environmental conditions, such as light conditions, temperature, nu-trient availability, which also lead to large variation in cell productivityand isotopic fractionation (Fry, 1996; Lowe et al., 2014; Savoye et al.,2003). This was the case in the present study: δ13C varied by 6‰andδ15Nby4‰throughout the study period. Thus, for running the mixingmodel, phytoplankton elemental and isotopic signatures were discrim-inated on a seasonal basis (Fig. 5,Table 1).Adour POM was also considered as an end-member. As the AdourRiver signatures exhibited low seasonal variability (δ13C=-26.5 ±0.7‰,δ15N = 4.6 ± 0.9‰, C:N = 9.0 ± 0.9 mol·mol-1) the averagevalues were used for running the mixing model. Adour river POM wasmainly of terrestrial origin) and almost never phytoplankton-dominated,as deduced by high POC:Chlaand C:N ratios. In the following the term‘terrestrial POM’will be used for describing the Adour POM.Anthropogenic POM is rarely considered as a contributor to coastalPOM.However,itcouldcontributetoupto50%ofPOMcompositioninsome coastal systems (e.g. Berto et al., 2013). POM originating fromthreesewagetreatmentplants(STP) was sampled in the present study.Their elemental and isotopic signatures varied according to 1) the sewagetreatment characterized by the type of effluent (urban vs industrial), theresidence time, the availability of tertiary treatment, etc., and/or 2) theweather conditions leading to different operating mode, which affectSTP functioning. Consequently, anthropogenic POM signatures were dis-criminated according to the treatment and/or the weather condition(Table 1;Fig. 5) and STP were considered as an end-member.The different sources of coastal POM were well discriminated bytheir elemental and/or isotopic signatures (Table 1;Fig. 5). For instance,anthropogenic POM and terrestrial POM exhibited similar δ13C valuesbut were well discriminated by their δ15N and/or C:N ratio values.Also, even if the C:N ratio and δ13C of the three anthropogenic POM atboth functioning (wet and dry conditions) were very similar, theirδ15N allowed their discrimination. Most of the coastal POM elementaland isotopic values were similar to phytoplankton signatures or inter-mediate between the phytoplankton and terrestrial signatures. This in-dicated that the latter sources are the main contributors to coastal POMcomposition in this area. However there were few departures from thisoverall pattern indicating that anthropogenic POM also episodicallycontributed to this composition (e.g. in autumn, Fig. 5c,d).Finally, the mixing model was run for each sampling date, site anddepth using terrestrial POM, phytoplankton and anthropogenic POMas end-members (Table 1). The isotopic and elemental values used forrunning themixing model were: the annual average values for terrestri-al POM, the seasonal average values for phytoplankton, and the averageof the operating mode values for anthropogenic POM.4.2.2. Composition of particulate organic matterAt annual time scale, considering the two sites and the three depths,the results of the mixing model (averages ± standard deviations) indi-cated that coastal POM was dominated by phytoplankton (71.6 ±24.2%), that the contribution of terrestrial POM was moderate(22.7 ± 21.8%), and that the contribution of anthropogenic POM wasusually negligible (5.7 ± 7.4%) (Fig. 6). The dominance of phytoplank-ton contribution to coastal POM composition was recorded at all sea-sons, sites and depths, except in Tarnos bottom water during autumn(Fig. 6). This contribution was higher than 70% in surface and interme-diate waters of Tarnos site at all seasons and of Biarritz site from springto autumn (Fig. 6). Such pattern is commonly observed in coastal sys-tems (e.g. Cresson et al., 2012; Savoye et al., 2003; Tesi et al., 2007).The contribution of terrestrial POM to the coastal POM compositionwas generally higherin bottom waters than in surface and intermediatewaters (in average, ca. +20%), and, for the bottom waters, in autumnand winter than in spring and summer (in average, ca. 40% comparedto 20%; Fig. 6). This higher contribution of terrestrial POM in bottom wa-ters suggests that this refractory material accumulated on the sedimentsurface and was resuspended due to the bottom currents in autumn andwinter, when the weather condition goes toward stronger wind- andwave-induced currents. The only large contribution of terrestrial POMto surface and intermediate waters was recorded in winter at the Biar-ritz site (in average, 25%), i.e. at the season where the Adour flow wasthe highest (Fig. 2) and at the site that is the more located in the vicinityof the Adour plume (Petus et al., 2010). Similar trends of higher contri-bution of terrestrial POM in wet versus dry seasons and in bottom ver-sus surface waters were observed in other coastal systems (e.g Gao etal., 2014; Tesi et al., 2014).Finally, the contribution of the anthropogenic POM to the coastalPOM was usually negligible (b10%) (Fig. 6), i.e. within the range of themixing model uncertainty (see Section 2.5). Anthropogenic POM contri-butions higher than 10% were only encountered in Biarritz bottom wa-ters, i.e., at the site the more in the vicinity of the Adour plume (Fig. 6).As for the terrestrial POM, the refractory anthropogenic POM may accu-mulate on the sediment surface and be resuspended in the bottomwater. Anthropogenic POM is usually not considered as a source forcoastal POM and when it is considered, its contribution is usually low(e.g. b20%; Cresson et al., 2012). However, it has been estimated as con-tributingto 25–50% of the POM composition in the Venice lagoon (Bertoet al., 2013). Nonetheless, the Venice lagoon is a microtidal semi-enclosed system whereas the southern Bay of Biscay is a mesotidalopen system. Thus, anthropogenic POM input is largely diluted withinPOM of autochthonous origin in the latter system.4.3. Mucilage characteristics and relationship with POMBiotic and abiotic processes inducing the formation of mucilaginousaggregates has been extensively studied (Mecozzi et al., 2005; Passow,2002; Turk et al., 2010; Verdugo et al., 2004). The production of exu-dates by phytoplanktonic and/or bacterial cells (Passow, 2002; Simonet al., 2002) is usually related to abrupt salinity changes, or nutrientpulses or depletion in coastal waters (Degobbis et al., 1999; Giani etal., 2012; Mecozzi et al., 2005). Meteorological conditions (i.e. winds,waves, etc.) are also liable to influence aggregation or break up of themucilaginous material (Degobbis et al., 1999; Precali et al., 2005). Re-petitivecollision and sticking process leadto different degrees of muci-lage aggregation and determine their ability to trap particles andorganisms in the water column.When it starts to aggregate, mucilage can be morphologically classi-fied based on visual observations (Precali et al., 2005; Stachowitsch et19C. Liénart et al. / Journal of Sea Research 116 (2016) 12–22
al., 1990). Studies separate mucilage macroaggregates in categoriesbased on a set of microbial or biochemical parameters (Bongiorni etal., 2007). Usually, flocs (b1 cm) are present at the beginning, beforemassive mucilage events. Then, macroflocs (1–5 cm) and stringers (2–25 cm) appear after a maximum abundance of flocs (Precali et al.,2005). Larger aggregates such as ribbons (10–20 cm to over 1 m), cob-webs (complex structure of few meters) and clouds (0.5 to 3–4m)floatin the upper water column and can form large and dense layers of ag-gregates in intermediate water masses or cover the sediment. Also,layers of different size, thickness and with creamy or gelatinous consis-tency may appear at the water surface. During the study year, visual ob-servations indicated that the mucilage was only present as flocs, incontrast to the previous years where field observations indicated thatmucilage reached higher degrees of aggregation (Susperregui, pers. d&Elbée et al., 2016).In the literature mucilage is described as “composed of various or-ganic and inorganic materials depending largely on the given systemand environmental conditions”(Simon et al., 2002). Particles trappedby the aggregates and thus mucilage composition largely depend onwhat is present in water column. This induces different functions,such as attachment surfaces, refuges for predation, substrate, etc., anddegradation processes of the mucilage (Verdugo et al., 2004). In orderto assess the composition of mucilage, a previous study was conductedon the Basque coast (d&Elbée et al., 2016). It consisted in trapping muci-lage using vertically anchoring nets (mesh size 750 μm) deployed at dif-ferent depth of the water column during six months(April–September).Mucilage was recovered weekly. A wide variety of both planktonic andbenthic components composed of 111 taxonomic units, and exogenousmatter of terrestrial origin, such as seeds, mammal&s hairs and pollenswas found associated to the mucilaginous aggregates (d&Elbée et al.,2016). However, this sampling method provided a substrate for theorganisms and led to an artificial ecosystem, which allowed the devel-opment of benthic organisms. This was not representative of the phe-nomenon that occurs in the water column. In contrast, samplingmucilage directly in the water column was intended in the presentstudy.When macroaggregates are large enough, in situ collection of muci-lage is usually performed by scuba diving using syringe or peristalticpump (Giani et al., 2006), or by centrifugation of water samples(Faganeli et al., 2009). These methods allow sampling the mucilagewithout sampling large amounts of organisms or particles that are notdirectly associated with mucilage. Such sampling strategies could notbe implemented in the present study, essentially because mucilagewas not forming large enough aggregates. In order to investigate the re-lationship between mucilaginous material and coastal POM, and to de-termine whether mucilaginous material trapped autochthonous and/or allochthonous particles, two approaches were performed: 1) thecomparison of isotopic and elemental signatures between the mucilag-inous material trapped using a net (mesh size of 200 μm) and the coastalPOM (see Section 2.3), and 2) the comparison of coastal POM isotopicand elemental signatures between periods of occurrence and of absenceof mucilage.Mucilaginous material collected by the net exhibited overall values(average ± standard deviation) of -19.9 ± 1.9‰for the δ13C, 5.7 ±1.4‰for the δ15N and 7.4 ± 1.9 mol·mol-1for the C/N ratio. Thesevalues were similar to the scarce values reported in the literature formucilage (δ13C=-19.2 ± 0.6‰(Giani et al., 2006); δ13C=-19‰and δ15N=5‰(Faganeli et al., 2009)). δ13C of the mucilaginous mate-rial was very different from the terrestrial and anthropogenic δ13Cbutsimilar to phytoplankton δ13C(Table 1)andδ13C of surrounding coastalPOM (Supplementary Table 2). This indicates that this material wasmainly of autochthonous origin and did not aggregate detectablesummer autumn winter spring Year020406080100)%(noitubirtnocecruoSBiarritz - Surface + Intermediate waters Tarnos - Surface + Intermediate watersBiarritz bottom water Tarnos bottom watersummer autumn winter spring Yearsummer autumn winter spring Yearsummer autumn winter spring Year020406080100020406080100020406080100Anthropogenic POM Phytoplankton Terrestrial POMFig. 6. Composition of the coastal POM (results of mixing model) for each of the four seasons and for the whole year. As there were no significant differences (p-value N0.05) in POMisotopic and elemental ratios between surface and intermediate waters (see Section 3.2.1), mixing model results were pooledfor surface and intermediate waters.20 C. Liénart et al. / Journal of Sea Research 116 (2016) 12–22
amounts of terrestrial and anthropogenic POM. Date-to-date compari-son of isotopic values between coastal POM or phytoplankton and netmaterial showed that, in most cases, isotopic values of the latter werehigher than those of the formers. Taking into account the fact thatδ13Candδ15N increase with the trophic level of organisms(Middelburg, 2014), this comparison indicates that the mucilaginousmaterial collected by the net may have been influenced by organismsof higher trophic levels. The net material was observed with a stereomi-croscope and among the mucilaginous matrix, phytoplankton and zoo-plankton organisms were identified. The presence of zooplankton inthis material explained the higher values of δ13Candδ15Ninthismate-rial compared to the coastal POM or to the phytoplankton. These organ-isms may have been trapped by the mucilage in the water column.However, the sampling method used in the present study may alsohave collected organisms that were not in situ associated with themucilage.Isotopic and elemental values of coastal POM (surface and interme-diate water) were also compared between periods with and withoutoc-currence of mucilage within each season. Whatever the season, therewas no significant difference in isotopic or elemental values of coastalPOM between periods with and without mucilage occurrence exceptfor the δ13CinspringasPOMδ13C was in average 1‰higher during mu-cilage events. This overall lackof difference indicatedthat the mucilagehad an elemental and isotopic signature similar to that of coastal POM,which was dominated by phytoplankton in this system (Section4.2.2). Interestingly, in average, phytoplankton contribution increasedfrom 66% to 78% and 86% when the mucilage amount was ‘nil’,‘moder-ate’or ‘large’(see Section 2.5 for the definition of the mucilage amountand the dedicated index). This illustrates the tight relationship betweenmucilage and phytoplankton, as reported in other coastal regions(Flander-Putrle and Malej 2008; Najdek et al., 2002).The above results indicate that themucilage present in the waters ofthe south-eastern Bay of Biscay was of autochthonous origin, did not ag-gregate detectable amounts of terrestrial or anthropogenic material andhad likely an isotopic and elemental signature similar to that ofphytoplankton.5. ConclusionThis study investigated coastal POM composition and its interactionwith mucilaginous material by using C and N elemental and isotopic ra-tios and a three-sources mixing model.Coastal POM composition of the South-East of the Bay of Biscay re-vealed a phytoplankton-dominated POM at all seasons, sites and mostof the depths (71.6 ± 24.2%). Terrestrial POM contribution was moder-ate (22.7 ± 21.8%) and anthropogenic POM contribution was usuallynegligible (5.7 ± 7.4%) and remained ca. constant over the year. Bothsites mainly exhibited similar vertical and temporal variations interms of POM origin and dynamics. Terrestrial POM contribution in-creased with depth and was higher in winter at all depths and in au-tumn in bottom waters, compared to other seasons. This wasconsistent with the literature focused on coastal areas. The main differ-ences between the two sites were: 1) the higher contribution of terres-trial POM in surface and intermediate waters and the highercontribution of anthropogenic POM to the bottom water for the site lo-cated south of the Adour River mouth (Biarritz), whereas 2) the contri-bution of terrestrial POM was higher in the bottom water of the sitelocated north of the Adour River mouth (Tarnos). Horizontal, verticaland temporal variations of coastal POM composition were linked tothe dynamics of the Adour plume and events of sediment resuspen-sion/deposition.Finally, mucilaginous material was mainly of autochthonous phyto-plankton origin and trapped no detectable amounts of allochthonousterrestrial or anthropogenic POM.There was a tight relationship be-tween phytoplankton relative contribution and mucilage. During thestudy period, and contrary to previous years where larger aggregateswere observed, mucilage was only present at floc stage. Consequently,a direct comparison of elementaland isotopic signatures between muci-lage, coastal POM and POM sources was not possible, making the inves-tigation of the relationship between coastal POM and mucilage difficultand incomplete. Such an investigation would be of high interest if itwould be performed during the whole process of mucilage aging, fromflocs to higher levels of aggregation.AcknowledgementsAuthors thank Nathalie Anglade, Laurent Dubois, Josiane Popowskyand Maxime Saudemont for their help on the field, and PacoRodriguez-Tress for computer skills. The work was supported by theprogram PERMALA-LIGA, founded by the Aquitaine Regional Council,and by the French national program EC2CO-DRIL MOSLIT.Appendix A. Supplementary dataSupplementary data to this article can be found online at http://dx.doi.org/10.1016/j.seares..ReferencesAzam, F., Fonda Umani, S., Funari, E., 1999. Significance of bacteria in the mucilage phe-nomenon in the northern Adriatic Sea. Ann. Ist. Super. Sanita 35, 411–419.Bennett, K., 1996. Determination of the number of zones in a biostratigraphical sequence.New Phytol. 132, 155–170. http://dx.doi.org/10.1111/j..1996.tb04521.x.Berto, D.,Rampazzo, F., Noventa, S., Cacciatore, F., Gabellini, M., Aubry, F.B., Girolimetto, A.,Brusà, R.B., 2013. Stable carbon and nitrogen isotope ratios as tools to evaluate thenature of particulate organic matter in the Venice lagoon. Estuar. Coast. Shelf Sci.135, 66–76. http://dx.doi.org/10.1016/j.ecss.2013.06.021.Bode, A., Alvarez-Ossorio, M.T., Varela, M., 2006. Phytoplankton and macrophyte contri-butionsto littoral food webs in theGalician upwellingestimated from stableisotopes.Mar. Ecol. Prog. Ser. 318, 89–102. http://dx.doi.org/10.3354/meps318089.Bongiorni, L., Armeni, M., Corinaldesi, C., Dell&Anno, A., Pusceddu, A., Danovaro, R., 2007.Viruses, prokaryotes and biochemical composition of organic matter in differenttypes of mucilage aggregates. Aquat. Microb. Ecol. 49, 15–23. http://dx.doi.org/10.3354/ame01126.Boyd, P.W., Sherry, N.D., Berges, J.A., Bischop, J.K.B., Calvert, S.E., Charette, M.A.,Giovannoni, S.J., Goldblatt, R., Harrison, P.J., Moran, S.B., Roy, S., Soon, M., Strom, S.,Thibault, D., Vergin, K.L., Whitney, F.A., Wong, C.S., 1999. Transformation of biogenicparticles from the pelagic to deep ocean realm. Deep-Sea Res. II 46, 2761–2792.Coplen, T.B., 2011. Guidelines and recommended terms for expression of stable-isotope-ratio and gas-ratio measurement results. Rapid Commun. Mass Spectrom. 25,2538–2560. http://dx.doi.org/10.1002/rcm.5129.Cresson, P., Ruitton, S., Fontaine, M.F., Harmelin-Vivien, M., 2012. Spatio-temporal varia-tion of suspended and sedimentary organic matter quality in the bay of Marseilles(NW Mediterranean) assessed by biochemical and isotopic analyses. Mar. Pollut.Bull. 64, 1112–1121. http://dx.doi.org/10.1016/j.marpolbul.2012.04.003.Dailloux, D., 2008. Video measurements of the Adour plume dynamic and its surfacewater optical characteristics. Ph.D thesis. Université de Pau et des Pays de l&Adour.de Groot, R., Brander, L., van der Ploeg, S., Costanza, R., Bernard, F., Braat, L., Christie, M.,Crossman, N., Ghermandi, A., Hein, L., Hussain, S., Kumar, P., McVittie, A., Portela, R.,Rodriguez, L.C., ten Brink, P., van Beukering, P., 2012. Global estimates of the valueof ecosystems and their services in monetary units. Ecosyst. Serv. 1, 50–61. http://dx.doi.org/10.1016/j.ecoser.2012.07.005.Degobbis, D., Malej, A., Fonda Umani,S., 1999. The mucilagephenomenon in the northernAdriatic Sea. A critical review of the present scientific hypotheses. Ann. Ist. Super.Sanita 373–381.d&Elbée, J., Lalanne, Y., Susperregui, N., Soulier, L., Castège, I., D&Amico, F., 2016. Biologicalcharacterization of aggregates clogging fishing nets on the Basque coastal waters ofFrance. Estuar. Coast. Shelf Sci. http://dx.doi.org/10.1016/j.ecss..Dubois, S.,Savoye, N., Grémare,A., Plus, M., Charlier,K., Beltoise, A., Blanchet, H., 2012. Or-igin and composition of sediment organic matter in a coastal semi-enclosed ecosys-tem: An elemental and isotopic study at the ecosystem space scale. J. Mar. Syst. 94,64–73. http://dx.doi.org/10.1016/j.jmarsys.2011.10.009.Dubois, S., Blanchet, H., Garcia, A., Massé, M., Galois, R., Grémare, A., Charlier, K., Guillou,G., Richard, P., Savoye, N., 2014. Trophic resource use by macrozoobenthic primaryconsumers within a semi-enclosed coastal ecosystem: Stable isotope and fatty acidassessment. J. Sea Res. 88, 87–99. http://dx.doi.org/10.1016/j.seares..Dunne, J.P., Sarmiento, J.L., Gnanadesikan, A., 2007. A synthesis of global particle exportfrom the surface ocean and cycling through the ocean interior and on the seafloor.Glob. Biogeochem. Cycles 21, 1–16. http://dx.doi.org/10.GB002907.Faganeli,J., Ogrinc, N., Kovac, N., Kukovec, K., Falnoga, I., Mozetic, P.,Bajt, O., 2009. Carbonand nitrogen isotope composition of particulate organic matter in relation to muci-lage formation in the northern Adriatic Sea. Mar. Chem. 114, 102–109. http://dx.doi.org/10.1016/j.marchem.2009.04.005.21C. Liénart et al. / Journal of Sea Research 116 (2016) 12–22
Ferrer, L., Fontán, A., Mader, J., Chust, G., González, M., Valencia, V., Uriarte, A., Collins,M.B., 2009. Low-salinity plumes in the oceanic region of the Basque Country. Cont.Shelf Res. 29, 970–984. http://dx.doi.org/10.1016/j.csr.2008.12.014.Flander-Putrle, V., Malej, A., 2008. The evolution and phytoplankton composition of mu-cilaginous aggregates in the northern Adriatic Sea. Harmful Algae 7, 752–761. http://dx.doi.org/10.1016/j.hal.2008.02.009.Fogg, G.E., 1995. Some speculations on the nature of the pelagic mucilage community ofthe northern Adriatic Sea. Sci. Total Environ. 165, 59–63. http://dx.doi.org/10.1016/(95)04543-A.Fontán, A., Mader, J., Gonzalez, M.,Uriate, A., Gyssels, P., Collins,M.B., 2006. Marinehydro-dynamics between San Sebastián and Hondarribia (Guipúzcoa, northern Spain): fieldmeasurements and numerical modelling. Sci. Mar. 51–63. http://dx.doi.org/10.3989/scimar.2006.70s151.Fry, B., 1996.12C/13C fractionation by marine diatoms.Mar. Ecol. Press Ser. 134,283–294.Gao, L., Li, D., Ishizaka, J., 2014. Stable isotope ratios of carbon and nitrogen in suspendedorganic matter: Seasonal and spatial dynamics along the Changjiang (Yangtze River)transport pathway. J. Geophys. Res. Biogeosci. 119, 557–566. http://dx.doi.org/10.JG002433.Gattuso, J.-P., Smith, S.V., 2007. Coastal zone. In: Cleveland, C.J. (Ed.), Encyclopedia ofEarth. Environmental Information Coalition, National Council for Science and the En-vironment, Washington D. C.Giani, M., Cicero, A.M., Savelli, F., Bruno, M., Donati, G., Farina, A., Veschetti, E., Volterra, L.,1992. Marine snow in the Adriatic Sea: a multifactorial study. Sci. Total Environ.539–549 (suppl).Giani, M., Berto, D., Zangrando, V., Castelli, S., Sist, P., Urbani, R., 2005. Chemical character-ization of different typologies of mucilaginous aggregates in the northern AdriaticSea. Sci. Total Environ. 353, 232–246. http://dx.doi.org/10.1016/j.scitotenv.2005.09.027.Giani, M., Zangrando, V., Berto, D., 2006. 13C/12C isotope ratio in the organic matterforming the mucilaginous aggregates in the Northern Adriatic Sea. Isot. Environ.Stud. Aquat. 93–96 (Forum C&S Papers).Giani, M., Sist, P., Berto, D., Serrazanetti, G.P., Ventrella, V., Urbani, R., 2012. The organicmatrix of pelagic mucilaginous aggregates in the Tyrrhenian Sea (MediterraneanSea). Mar. Chem. 132-133, 83–94. http://dx.doi.org/10.1016/j.marchem.2012.01.002.González, M., Ferrer, L., Uriarte, A., Urtizberea, A., Caballero, A., 2008. Operational ocean-ography system applied to the prestige oil-spillage event. J. Mar. Syst. 72, 178–188.http://dx.doi.org/10.1016/j.jmarsys.2007.07.005.Kennedy, P., Kennedy, H., Papadimitriou, S., 2005. The effect of acidification on the deter-mination of organic carbon, total nitrogen and their stable isotopic composition inalgae and marine sediment. Rapid Commun. Mass Spectrom. 19, 1063–1068. http://dx.doi.org/10.1002/rcm.1889.Ki?rboe, T., 2000. Colonization of marine snow aggregates by invertebrate zooplankton:abundance, scaling, and possible role. Limnol. Oceanogr. 45, 479–484. http://dx.doi.org/10.4319/lo.2000.45.2.0479.Koutsikopoulos, C., LeCann, B., 1996. Physical processes andhydrological structures relat-ed to the Bay of Biscay anchovy. Eur. anchovy its Environ. 60, 9–19.Lorrain, A., Savoye, N., Chauvaud, L., Paulet, Y.M., Naulet, N., 2003. Decarbonation andpreservation method for the analysis of organic C and N contents and stable isotoperatios of low-carbonated suspended particulate material. Anal. Chim. Acta 491,125–133. http://dx.doi.org/10.1016/S)00815-8.Lowe, A.T., Galloway, A.W.E., Yeung, J.S., Dethier, M.N., Duggins, D.O., 2014. Broad sam-pling and diverse biomarkers allow characterization of nearshore particulate organicmatter. Oikos 123, 1341–1354. http://dx.doi.org/10.1111/oik.01392.Martin, J.H., Knauer, G.A., Karl, D.M., Broenkow, W.W., 1987. VERTEX: carbon}