Progress in Oceanography 129 (2014) 149–158

Contents lists available at ScienceDirect

Progress in Oceanography

journal homepage: www.elsevier .com/ locate /pocean

Benthic food web structure in the Comau fjord, Chile (�42�S):Preliminary assessment including a site with chemosynthetic activity

http://dx.doi.org/10.1016/j.pocean.2014.03.0050079-6611/� 2014 Elsevier Ltd. All rights reserved.

⇑ Corresponding author. Address: Departamento de Biología Marina, UniversidadCatólica del Norte, Larrondo #1281, Coquimbo, Chile. Tel.: +56 9 89915604.

E-mail address: zapata.bm@gmail.com (G. Zapata-Hernández).

Germán Zapata-Hernández a,⇑, Javier Sellanes a, Christoph Mayr b,c, Práxedes Muñoz a

a Departamento de Biología Marina, Facultad de Ciencias del Mar, Universidad Católica del Norte, Coquimbo, Chileb Institut für Geographie, Friedrich-Alexander-Universität Erlangen-Nürnberg, Erlangen, Germanyc GeoBio-Center and Department für Geo- und Umweltwissenschaften, Ludwig-Maximilians Universität München, Germany

a r t i c l e i n f o

Article history:Available online 25 March 2014

a b s t r a c t

Using C and N stable isotopes we analyzed different trophic aspects of the benthic fauna at two sites inthe Comau fjord: one with presence of venting of chemically reducing fluids and extensive patches ofbacterial mats (XH: X-Huinay), and one control site (PG: Punta Gruesa) with a typical fjord benthic hab-itat. Due to the widespread presence of such microbial patches in the fjord and their recognized trophicrole in reducing environments, we hypothesize that these microbial communities could be contributingto the assimilated food of consumers and transferring carbon into high trophic levels in the food web.Food sources in the area included macroalgae with a wide range of d13C values (�34.7 to �11.9‰),particulate organic matter (POM, d13C = �20.1‰), terrestrial organic matter (TOM, d13C = �32.3‰ to�27.9‰) and chemosynthetic filamentous bacteria (d13C = ��33‰). At both sites, fauna depicted typicalvalues indicating photosynthetic production as a main food source (>�20‰). However, at XH selectedtaxa reported lower d13C values (e.g. �26.5‰ in Nacella deaurata), suggesting a partial use of chemosyn-thetic production. Furthermore, enhanced variability at this site in d13C values of the polyplacophoranChiton magnificus, the limpet Fissurella picta and the tanaid Zeuxoides sp. may also be responding to theuse of a wider scope of primary food sources. Trophic position estimates suggest three trophic levelsof consumers at both sites. However, low d15N values in some grazer and suspension-feeder speciessuggest that these taxa could be using other sources still to be identified (e.g. bacterial films, microalgaeand organic particles of small size-fractions). Furthermore, between-site comparisons of isotopic nichewidth measurements in some trophic guilds indicate that grazers from XH have more heterogenic trophicniches than at PG (measured as mean distance to centroid and standard deviation of nearest neighbordistance). This last could be ascribed to the utilization of a mixture of photosynthetic and chemosyntheticcarbon sources. In addition, corrected standard ellipses area (SEAc) values in suspension-feeders andcarnivores at both sites suggest a similar magnitude of exploitation of food sources. However, grazersfrom XH have a greater expansion of their isotopic niche (SEAc), probably explained by the presence ofspecies with low d13C and d15N values, and directly associated to chemosynthetic carbon incorporation.

� 2014 Elsevier Ltd. All rights reserved.

1. Introduction

The Chilean Patagonia (�41–55�S) contains one of the majorfjord regions in the world, with a highly heterogeneous landscapeincluding channels, inner seas and inlets (Sánchez et al., 2011) thatprovide important ecosystem services which have been scarcelymeasured or valued (Iriarte et al., 2010). The Comau fjord, located�80 km SE of the city of Puerto Montt (�42�S), is about 34 km longand 5 km wide, with a maximum depth of about 600 m inits mouth and steep slopes characterizing the fjord margins

(Haussermann and Försterra, 2009). Characteristic of the area is alow salinity layer that may attain a thickness of up to 7 m and astrong tidal amplitude (4–6 m) which have pronounced effectsover the benthic community structure above 10 m depth (Melzeret al., 2006). On the other hand, hard-bottom benthic communitiesof the southern Chilean fjords are generally among the least stud-ied worldwide, due to remoteness, difficult access and inhospitableconditions of the region (Schwabe et al., 2006).

In the Comau fjord, particular benthic communities have beenreported such as shallow water scleractinian cold-water coralbanks, brachiopod banks and extensive patches of chemoautotro-phic bacteria, these last ones preliminarily associated to the seep-age of groundwater loaded with reduced sulfur compounds(Häussermann et al., 2012). As it will be shown later, although iso-

150 G. Zapata-Hernández et al. / Progress in Oceanography 129 (2014) 149–158

topic evidence indicates that this water is probably associated tothe plumbing system of the adjacent ‘Barranco Colorado’ volcano,no thermal anomalies compared with ambient seawater were de-tected. Furthermore, no bubbling of methane has been observedand only modest amounts of this gas have been measured. Giventhat the system does not have clear characteristics of the so called‘cold-seeps’ or ‘hydrothermal vents,’ we will refer to it henceforthas a ‘cold-vent’. Here, filamentous bacteria communities cover barerock but also overgrow animals like chitons, gastropods, musselsand dead corals from the intertidal to at least 100 m depth. Mor-phologically, these communities appear similar to those studiedat an intertidal methane seep site at Mocha Island, central Chile(Sellanes et al., 2011). Although detailed analyses are still under-way, the filamentous bacteria of Comau fjord preliminarily seemto be related to at least four genera of thiotrichales: Thiomargarita,Thiomicrospira, Thiothrix and Beggiatoa (Belmar et al., 2012). Fur-thermore, recent metagenomic analysis for the microbial commu-nity associated with these bacterial mats indicates a highproportion of members of the class Gammaproteobacteria (Ugaldeet al., 2013), which are known as important chemoautotrophs(Yamamoto and Takai, 2011), participating as catalysts of microbialsulfur oxidation and inorganic carbon fixation in marine sediments(Lenk et al., 2011). Additionally, due to their wide presence in thearea, we hypothesize that these microbial communities could becontributing to the assimilated food of benthic consumers, as ithas been reported for some heterotrophic consumers (e.g. gastro-pods, peracarids and polychaetes) from shallow methane seeps atMocha Island (Sellanes et al., 2011) as well as in deep-water reduc-ing environments where these chemosynthetic bacteria also thrive(Levin and Michener, 2002; Levin and Mendoza, 2007; Carlier et al.,2010; Zapata-Hernández et al., 2013).

Furthermore, the availability of autochthonous photosyntheticproduction (i.e. phytoplankton and macroalgae) and large inputsof forest litter and wood chips into the fjords result in a high avail-ability of terrestrial organic matter (TOM) in the sediments (Quirogaet al., 2012), which may also be exploited by some benthic consum-ers (McLeod and Wing, 2009). However, in spite of a general analysisof the fjord food web (Mayr et al., 2011), the role of chemosyntheticbacteria and TOM as potential food sources remains unclear for mostbenthic consumers in Comau fjord, precluding to establish a possibledependence on these basal sources and detect any changes on ben-thic communities associated with temporal changes in the availabil-ity of these and other food resources. In addition, a detailed analysisof the trophic structure of benthic communities inhabiting fjordareas in southern Chile is still lacking.

The study of nutritional pathways and trophic structure is afundamental requirement for understanding energy flow fromthe base to higher trophic levels in food webs (Nilsen et al.,2008). The analysis of natural biological tracers, such as stable iso-tope ratios of carbon (d13C) and nitrogen (d15N) constitute a robustapproach often used to trace energy flow pathways throughoutfood webs in reducing environments (Van Dover, 2008). Whiled13C is generally used to identify carbon fixation pathways in pri-mary producers and the food sources of heterotrophic consumers(DeNiro and Epstein, 1978), enrichment of d15N between twoadjacent trophic levels (�3.4‰; Minagawa and Wada, 1984; Post,2002) allows the estimation of trophic positions of different con-stituents of the food webs (Cabana and Rasmussen, 1996; Carlieret al., 2007) if basal signatures are known. Furthermore, stableisotopes analyses allow the study of trophic niches, providing fortime- and space-integrated representations of the trophic ecologyof organisms and are powerful tools for studying ecologicalresponses to anthropogenic impacts (Layman et al., 2007a, andreferences therein).

Therefore, based on the carbon and nitrogen stable isotopeanalysis of a large variety of organisms from two localities in the

fjord, the aims of this study are: (i) to trace the origin of trophicsources (e.g. photosynthetic, chemosynthetic and terrestrial origin)that support the benthic fauna, (ii) to characterize the trophicstructure of benthic consumers, and (iii) to analyze the isotopicniche width of selected benthic trophic guilds. Thus, this studymay also be considered as a base-line of the benthic food web,which added to the identification of the sources of heterogeneityin their trophic structure and might be useful for environmentalmanagement in the Comau fjord area. This kind of primary infor-mation is crucial for such a fragile ecosystem facing a scenario ofchanges in the availability of food sources associated to anthropo-genic alterations (e.g. salmon farming, hydroelectric projects andglobal climate change, among others) to the benthic ecosystem.

2. Materials and methods

2.1. Site characteristics and sample collection

Two campaigns were performed during October 2012 andMarch 2013 in the vicinity of the Huinay Scientific Field Station(HSFS). Two locations were studied in the Comau fjord (Fig. 1):‘X-Huinay’ (XH; 42�23.2790S, 72�27.6350W), located on the westside of the fjord and at the base of ‘Barranco Colorado’ volcanowhere chemically reduced fluids (sulfide smelling), which are visu-alized by the hazy plume they produce due to density differenceswith seawater, are venting from cracks in the fjord wall. Theseareas are covered by extensive chemosynthetic bacterial patcheswhich are distributed from the intertidal zone (Fig. 2A) to �90 mdepth (and probably even greater depths), and also cover numer-ous invertebrate taxa (Fig. 2B), as it has been observed from otherseeps (e.g. bivalves, crustacean, gastropods; Bailey et al., 2011). Asa control, we selected a location at the opposite side of the fjord,locally known as ‘Punta Gruesa’ (PG; 42�24.5750S, 72�25.4620W).At this site, morphologically similar to XH, there is no evidenceof groundwater fluxes or profuse development of chemosyntheticbacterial mats. At both sites preliminary inspections, down to90 m depth, were performed using a remotely operated vehicle(mini ROV ‘‘Gnom’’) and posteriorly macro- (body size between0.3 and 1 cm) and megafauna (body size >1 cm) samples were col-lected at selected sites, from 0 to �30 m depth, by SCUBA divers.

2.2. Chemical characterization of fluids emerging from cracks

Samples for particulate organic matter (POM) were collectedduring the two campaigns, using a Niskin bottle. Approximately2 L of water were pre-sieved through a 63-lm mesh to removelarge-sized zooplankton and large detrital particles, and thenfiltered through pre-combusted (500 �C for 4 h) Whatman GF/Ffilters (0.45 lm nominal pore size). During the final phase of thefiltration a small volume of 0.1 N HCl was added to remove carbon-ates from the filtrate. The sedimentary organic matter (SOM) wasobtained collecting surficial organic matter from small terraces.

Samples of water venting from selected cracks were obtaineddirectly by divers using syringes, from the surface to �38 m waterdepth. Samples for methane analyses were directly placed fromsyringes in evacuated septum-capped vials 12 mL vials (Exetainers,Labco Ltd.) containing HCl in order to reach a pH � 2. The vialswere filled totally avoiding bubbles. The methane analyzes wereperformed at the Stable Isotope Facilities at UC Davis. The samplesfor sulfides (RH2S) were also placed directly in evacuated Exetain-ers containing zinc acetate to preserve the sample. Sulfide wasdetermined spectrophotometrically using the methylene bluemethod (Cline, 1969). All samples were maintained in dark at4 �C until analyzes.

Redox and pH were measured on board immediately aftersamples retrieving using a Thermo-Orion 3 Star portable meter.

Fig. 1. Map of the study sites, X-Huinay (occurrence of bacterial mats associated to the venting of chemically reduced fluids) and Punta Gruesa (control) in Comau fjord,Southern, Chile (Ocean Data View map; Schlitzer, 2012).

G. Zapata-Hernández et al. / Progress in Oceanography 129 (2014) 149–158 151

2.3. Sample preparation

A total of 301 organism samples were processed live after iden-tification to the lowest possible taxon level (Table A.1). For largefauna, approximately 1 mg of tissue was dissected, washed withMilli-Q water, placed in pre-combusted vials and dried in an oven(60 �C) for 12 h. When appropriate, tissue samples were groundinto a fine powder with an agate mortar and small amounts(�0.5 mg) were placed in pre-weighed tin capsules and stored ina dessicator until stable isotope analysis. Care was taken to use car-bonate-free samples and therefore no acidification procedureswere performed. When possible, voucher specimens were pre-served in 95% ethanol, and deposited in the Biological Collectionof the Universidad Católica del Norte, Coquimbo (CBUCN) for latertaxonomic corroboration and eventual molecular analysis.

Additional stable isotope data previously published by Mayret al. (2011) of primary producers (i.e. macroalgae and cyanobacte-ria) and potential terrestrial food sources (i.e. soil peat) as well asof some consumers from both sites (XH and PG), were incorporatedinto the analyses. This complimented the existing data base at eachsite, and generated a finer resolution for the determination ofpotential trophic relationships in the benthic food web.

2.4. Stable isotope analysis

Samples were combusted at 1060 �C in a continuous helium flowin an elemental analyzer (NC2500, Carlo Erba) in the presence ofchromium oxide and silvered cobalt oxide. Among the resultinggases, nitrogen oxides and excess oxygen were reduced by passingover copper wires at 650 �C. Thereafter, water vapor was trappedwith Mg(ClO4)2 and the remaining gases (N2 and CO2) wereseparated in a gas chromatography column at 45 �C. N2 and CO2

were passed successively via a ConFloII interface into the isotope ra-tio–mass spectrometer (Delta Plus, Thermo-Finnigan) andisotopically analyzed. Carbon and nitrogen contents were deter-mined from the peak area versus sample weight ratio of each

individual sample and calibrated with the elemental standardscyclohexanone-2,4-dinitrophenylhydrazone (C12H14N4O4) andatropine (C17H23NO3) (Thermo Quest). A laboratory-internal organicstandard (Peptone) with known isotopic composition was used forfinal isotopic calibrations.

Stable isotope ratios are reported in the d notation as thedeviation relative to international standards (Vienna Pee DeeBelemnite for d13C and atmospheric N2 for d15N), so d13C ord15N = [(R sample/R standard) � 1] � 103, where R is 13C/12C or15N/14N, respectively. Typical precision of the analyses was±0.1‰ for d15N and d13C.

2.5. Trophic position of heterotrophic fauna

Calculation of the consumer trophic positions was performedusing the equation detailed by Vander Zanden and Rasmussen(1999), which has been widely used in marine ecosystems (e.g. Ikenet al., 2001, 2010). Trophic position was calculated using theequation:

TPconsumer ¼ 2þ ðd15Nconsumer � d15NbaseÞ�

3:4

where TPconsumer is the estimation of the trophic position of the con-sumer, d15Nconsumer is the measured d15N value in the consumeranalyzed. The d15Nbase value corresponds to the d15N value of thebasal consumer, which must be a primary consumer and be sessile,or have limited mobility (Sellanes et al., 2011). In this sense, due totheir wide presence and great abundance (�5 to �35 m), scarcemotility of adult individuals and the suspensivore feeding modeused by large individuals (Chaparro et al., 2002, 2013), the gastro-pod Crepipatella dilatata was selected as the reference basal con-sumer at both sites. The constant ‘2’ corresponds to the level ofprimary consumers at the food web (Iken et al., 2010). A value of3.4‰ is usually assumed as the average 15N enrichment per trophiclevel (Minagawa and Wada, 1984; Post, 2002).

Fig. 2. General habitat aspects at X-Huinay site: (A) Intertidal chemosynthetic bacteria attached to rocks and barnacles (scale bar: 30 cm), (B) chilean blue mussel Mytiluschilensis overgrown by chemosynthetic bacteria (scale bar: 4 cm; 5 m depth), (C) gastropod Nassarius gayi dwelling among mats (scale bar: 3 cm; 30 m depth; cotton-likestructures are mostly elemental sulfur accumulations), (D) echinoids Arbacia dufresnii (scale bar: 3 cm; 20 m depth), (E) the asteroid Poraniopsis echinaster probably grazing onthe bacterial mats (scale bar: 10 cm; 20–30 m depth), and (F) bacterial filaments attached over juveniles of the limpet Nacella deaurata (scale bar: 0.5 cm). (For interpretationof the references to colour in this figure legend, the reader is referred to the web version of this article.)

152 G. Zapata-Hernández et al. / Progress in Oceanography 129 (2014) 149–158

2.6. Comparisons of isotopic niche in trophic guilds

The following measurements of isotopic niche widths, proposedby Layman et al. (2007b), were calculated for three feeding guilds(i.e. suspension- feeders, grazers and carnivores) present at XH andPG: (1) Mean distance to centroid (CD): average Euclidean distanceof each species to the d13C and d15N centroid, where the centroid isthe mean d13C and d15N value for all species in the food web, givingadditional information about the isotopic niche amplitude andspacing between species within the guild. (2) Standard deviationof nearest neighbor distance (SDNND): a measure of the evennessof species packing in bi-plot space.

Additionally, a sample-size corrected version of standard ellipsearea (SEAc) was utilized as a measure of the mean core of the iso-topic niche occupied by all species (in each trophic guild) and thepotential primary food sources in the d13C and d15N space at eachsite (Jackson et al., 2011). This metric represents a measure ofthe total amount of niche occupied in the isotopic space. It allowsfor robust statistical comparisons between data sets with differentsample sizes and corrects bias generated when sample sizes aresmall (Jackson et al., 2011, 2012). Moreover, this metric allowedcalculating the overlapping area of the standard ellipses (and their

respective percentage) among sites, and this was used as a mea-sure of diet similarity.

A Bayesian concentration-dependence mixing model (Parnellet al., 2010), which included the elemental concentration of carbonand nitrogen from the samples, was used to estimate the potentialcontribution of chemosynthetic bacterial filaments, terrestrial or-ganic matter and marine photosynthetic production (macroalgaeand particulate organic matter) in the diet of the limpet Nacelladeaurata.

Metrics proposed by Layman et al. (2007b), SEAc and overlap-ping of standard ellipses, and Bayesian mixing model were calcu-lated using SIAR (Stable Isotope Analysis in R) package. Statisticalanalyses and SIAR calculations were performed using R 2.15.3 soft-ware (R Development Core Team, 2013).

3. Results

3.1. Chemical characterization of reducing fluids

The water venting from some cracks in the fjord wall showed aslightly lower pH (�7.4–7.6) than ambient fjord water (>8.0), as

Table 1Chemical characterization of fluids emerging from cracks at Station X-Huinay.

Date Depth range (m) pH (min–max) Redox (min–max) RH2S (mM ± SD) CH4 (mM ± SD) d13C–CH4 (‰ ± SD)

October 2012 0 8.49 152 116 ± 6 0.3 ± 0.03 �9.3 ± 0.3March 2013 0–1 7.63–8.74 �149 to 79 17 ± 10 0.7 ± 0.5 �6.6 ± 1.6

8–12 7.40–8.29 �360 to �24 67 ± 10 0.8 ± 0.6 �2.6 ± 1.736–38 7.40–8.76 �320 to �115 359 ± 163 11.4 ± 15.8a �26.6 ± 8.5

a Maximum value measured: 45.2 nM.

G. Zapata-Hernández et al. / Progress in Oceanography 129 (2014) 149–158 153

well as a very low redox potential, reaching values <�300 mV(Table 1). Chemically reduced compounds showed high spatial var-iability but normally the higher concentrations were found below30 water depth. Sulfide reached maximum values around359 ± 163 mM, and mean methane concentrations (±1SD) were11.4 ± 15.8 lM, with a maximum value of 45.2 lM (Table 1).Measured isotopic values of methane emerging from cracks werealways negative, but lowest values were observed in the deepestcracks (d13C–CH4 = �26.6 ± 8.5‰; Table 1). Although the magni-tudes of the fluxes of these fluids were not estimated they seemto be low, since the chemical composition of the bottom water isnot appreciably affected.

3.2. Stable isotope composition of food sources at XH and PG

Stable isotope composition of primary food sources at XH(including data from Mayr et al. (2011)) depicted a wide variationin their values (Fig. 3, Table A.1). The lowest d13C values wereregistered in two rhodophyte macroalgae (�34.7‰ and �33.3‰),chemosynthetic bacterial filaments (�33.3‰), terrestrial plants(d13C = �32.3‰), and soil-peat (�27.9‰, Fig. 3, Table A.1). Incontrast, relatively higher d13C values were reported in otherrhodophyte, chlorophyte, phaeophytetaxa and sedimentary organ-ic matter (SOM, d13C = �14.8‰; Fig. 3, Table A.1). The particulateorganic matter (POM) registered intermediate values of d13C(�20.1‰). At PG three rhodophyte macroalgae had low d13C values(��31.0‰) and the cyanobacteria Rivularia atra and the chloro-phyte Rhizoclonium sp. had comparatively higher d13C values(�9.3‰ and �11.9‰, respectively). Values of POM and SOM wereintermediate (�20.1‰ and �19.6‰, respectively).

Terrestrial plants, bacterial filaments and soil-peat had the low-est d15N values at XH (�6.8‰, �0.6‰, and 0.7‰, respectively) andthe macroalgae Rhodymenia howeana and Macrocystis sp. had thehighest d15N values (9.4‰ and 10.4‰, respectively; Fig. 3). Anintermediate POM value (8.2‰) was registered. On the other hand,at PG the cyanobacteria Rivularia atra and the chlorophyte Rhizoc-lonium sp. reported relatively low d15N values (�0.3‰ and 4.9‰),while Myriogramme multinervis and POM had higher d15N values(8.2‰ and 9.7‰, respectively) (Fig. 3). The SEAc values for thepotential food sources analyzed, indicate slight differences on theisotopic niches available among both sites, being relatively higherat XH (118.4‰) in contrast to PG (107‰), with an important over-lap between both (77.7‰) (Fig. B.1). However, the shape of ellipseson the isotopic space present notable differences, where the ellip-ses are strongly affected by the variability in the d15N and d13Cvalues of available sources present at XH at PG, respectively.

3.3. Stable isotope composition of consumers at XH and PG

Among the heterotrophic species analyzed at XH, the ascidianAplidium variable, a juvenile of the limpet Nacella deaurata andthe echinoid Pseudoechinus sp. reported low d15N values (5.7‰,8.0‰ and 8.3‰, respectively) (Fig. 3, Table A.1). In contrast, thePatagonian redfish Sebastes oculatus and the Chilean sandperch

Pinguipes chilensis registered the highest d15N values (16.4‰, bothspecies) (Fig. 3, Table A.1).

At PG, the hermit crab Pagurus villosus had the lowest d15N val-ues (5.3‰), and the fish Helicolenus lengerichi, the opisthobranchs(Diaulula puntuolata and Diaulula sp.) and the crustacean Harpacti-coid Copepoda sp. 2 reported the highest (17.0‰, 17.1‰, 18.5‰

and 18.9‰, respectively) (Fig. 3, Table A.1).At XH, juveniles of N. deaurata depicted lowest d13C values

(�26.5‰), followed by the polychaete Nereis callaona and thesmall, semi-sessile Amphipoda sp. 1 (�21‰) (Fig. 3). Estimationsof the potential contribution of microbial bacterial filaments tothe diet of N. deaurata, indicate similar contributions of this sourcein comparison with terrestrial and marine photosynthetic foodsources (�15% to �36%, each source) (Fig. B.2, Table A.2). On theother hand, the opisthobranch Cadlina sparsa and the limpetSiphonaria lessoni registered relatively high d13C values (�14.3‰

and �13.8‰, respectively) (Fig. 3, Table A.1).At PG the amphipod Paracaprella sp. and the polyplacophoran

Chiton cumingsii reported the lowest d13C values (�22.7‰ and�20.3‰, respectively), while two opisthobranch taxa (Opistho-branch indet. and Diaulula sp.) registered relatively high d13C val-ues (�11.4‰ and �12.8‰, respectively) (Fig. 3, Table A.1).

3.4. Trophic structure of benthic consumers

From a total of 130 animal taxa included in this study, 45 aresuspension-feeders (28 and 27 taxa for XH and PG, respectively,with 14 shared taxa) (Table A.1). Most species were positioned atthe second trophic level. Despite this, some cnidarians from XH(i.e. Acontiaria sp., Acontiaria sp. 2, Actinostola chilensis and thecold-water Desmophyllum dianthus) and the alcyonid Primnoellasp. were positioned at the third trophic level. At PG four spongesand the holothurian Psolidium disciformis were included at thethird trophic level (Table A.1). On the other hand, some taxaincluding the ascidian Aplidium variable (XH) and the hermit crabPagurus villosus (PG) reported extremely low trophic positions(TP = 0.5 and 0.8, respectively) (Fig. 3, Table A.1). Also some notabledifferences in the trophic positions of taxa present at both areaswere detected in D. dianthus and P. villosus (Table A.1).

In the case of grazers, 18 taxa were reported (15 and 10 for XHand PG, respectively, with 7 shared taxa) (Table A.1). Most taxawere principally positioned at trophic level 2, however, some taxafrom XH, including the limpet N. deaurata (juvenile and adult) andthe echinoid Pseudoechinus sp. reported slightly lower trophic posi-tions (TP = 1.5, 1.8 and 1.6, respectively) (Fig. 3, Table A.1).

On the other hand, a total of 12 detritivore taxa were reported(6 and 10 for XH and PG, respectively, with 4 shared taxa)(Table A.1). In general taxa were positioned at trophic level 2 andonly the polychaete Oenodidae sp. 2 from PG was classified as asecondary consumer (TP = 3.2) (Table A.1). Some slight differencesin the trophic position between sites were reported for the gastro-pod Nassarius gayi (TP = 2.4 vs. 2.8 at XH and PG, respectively), andfor the crab Halicarcinus planatus (TP = 2.8 vs. 1.8 at XH and PG,respectively) (Table A.1).

Fig. 3. Summary plot of d13C and d15N values and trophic levels estimated in benthic consumers (using Crepipatella dilatata as base consumers) and potential food sources atX-Huinay (upper box) and Punta Gruesa (lower box). IFM: Indeterminate feeding mode, POM: Particulate organic matter, TOM: Terrestrial organic matter, SOM: Sedimentaryorganic matter.

154 G. Zapata-Hernández et al. / Progress in Oceanography 129 (2014) 149–158

Finally, a total of 40 carnivore taxa were reported (21 and 31for XH and PG, respectively, with 11 shared taxa) (Table A.1). Awide spectrum of trophic positions (including three trophic lev-els) were estimated for the carnivore taxa present at both sites.In this sense, it is important to note that several taxa (e.g.polychaetes, nemertean, gastropods, opistobranchs, crustaceans,pycnogonids and asteroids) reported low d15N values and werepositioned at the trophic level two (Table A.1); excluding thesetaxa, in general the opisthobranch taxa present at both sitesreported high trophic positions (TP = 3.5–4.4), similar to gastro-pods and asteroids (TP = 3.1–3.8) and fish species (TP = 3.7–4.0)(Table A.1).

3.5. Isotopic niche width within trophic guilds

Layman’s metrics for carnivore guild showed similar values atboth sites (CD = 2.5 and SDNND = �1.9). On the other side, the SEAcvalues were slightly higher for the carnivore guild from PG than forthose of XH (Fig. 4, Table 2), with an overlapping area between theellipses <50% and a shift of carnivores from XH toward lowertrophic levels (Fig. 4A, Table 2).

In the case of the suspension-feeder guild, there were no signif-icant differences in CD and SDNND metrics among sites (Table 2).However, the ellipses areas (SEAc) presented small differences intheir isotopic niche width, where the suspension-feeders from

Fig. 4. Convex hull areas (dotted lines) and corrected standard ellipses areas (SEAc; solid lines) in the isotopic space for trophic guilds: (A) Carnivores, (B) suspension-feeders,and (C) grazers, present at X-Huinay (black ellipses) and Punta Gruesa (red ellipses). (For interpretation of the references to color in this figure legend, the reader is referred tothe web version of this article.)

Table 2Summary of niche community metrics of benthic communities sampled at X-Huinay (XH) and Punta Gruesa (PG), in Comau fjord. Values are reported as the mean and its 95%confidence interval. CD: mean distance to centroid; SDNND, standard deviation of nearest neighbor distance; SEAc, standard ellipse area corrected.

Metrics Suspension-feeders Grazers Carnivores

XH PG XH PG XH PG

CD (‰) 2.5 (1.5–4.9) 2.5 (1.6–4.6) 3.4 (2.2–6.0) 2.9 (1.6–6.4) 2.5 (1.5–4.8) 2.5 (1.6–4.5)SDNND (‰) 2.0 (0.4–8.5) 1.9 (0.4–7.7) 2.4 (0.8–7.1) 1.9 (0.4–7.9) 1.8 (0.4–7.4) 1.9 (0.4–7.4)SEAc (‰) 5.5 8.2 14.5 7.0 5.4 7.5Overlapping SEAc (‰) 4.8 4.6 2.7Total overlapping (%) 87 59 32 66 50 36

G. Zapata-Hernández et al. / Progress in Oceanography 129 (2014) 149–158 155

XH had a slightly lower trophic niche than those at PG (Fig. 4B, Ta-ble 2). The estimation of the overlapping area and the position ofthe ellipses on the biplot space suggest a high similarity of the iso-topic niches of species present at both sites (Fig. 4B, Table 2).

Finally, grazers from XH reported higher CD and SDNND valuesthan those of PG (Table 2). However, SEAc measures were abouttwo times larger at XH in contrast to PG (SEAc = 14.5‰ and7.0‰, respectively; Table 2) and the ellipse overlapping areas werelower (�32%) for grazers present at XH (Table 2) and with a shift ofthe ellipse toward low trophic levels (Fig. 4C).

4. Discussion

4.1. Carbon sources supporting benthic consumers

The importance of numerous and heterogenic energy flowpathways is considered a primary mechanism for the stabilizationof food webs (Rooney et al., 2006). In this sense, available carbon infjord ecosystems mainly originates from phytoplankton, macroalgae,benthic microalgae, allochthonous plant material (Nilsen et al.,2008) and chemosynthetic bacteria fueled by litter decomposition

156 G. Zapata-Hernández et al. / Progress in Oceanography 129 (2014) 149–158

(McLeod and Wing, 2009). In the Comau fjord, the emanations ofreducing fluids (e.g. sulfide and methane) associated to geologicprocesses could additionally serve as energy and carbon sourcesfor chemosynthetic microorganisms (e.g. sulfide-oxidizing andmethanotrophic bacteria). However, the microbial communitiesassociated to bacterial mats principally use carbon via CO2 fixationthrough the reverse tricarboxylic acid (rTCA) pathway and to alesser extent the RuBisCO form I of the Calvin Benson-Bassham(CBB) cycle (Ugalde et al., 2013). The rTCA cycle is actually recog-nized as an important carbon fixation pathway at the base ofdeep-sea hydrothermal food webs (Hügler and Sievert, 2011),producing distinctive d13C values (>�13‰) in contrast to CBB cycle(<�22‰ to �30‰) (Reid et al., 2013). However, at XH the benthicfauna reported low d13C values (<�14‰) pointing to autochtho-nous photosynthetic food sources (i.e. POM and macroalgae),suggesting a main direct support from these marine sources.

Additionally, many factors could affect the availability and typeof carbon available to benthic consumers. In this context, seasonal-ity in the photosynthetic productivity of Comau fjord, associated toa dynamic low-salinity layer (LSL), may generate changes in thepelagic structure and hence carbon fluxes into the benthic ecosys-tem (Sánchez et al., 2011). In addition, although benthic macroal-gae growing within the LSL in the Comau fjord exhibited theinfluence of land-derived DIN pools affecting their isotopic compo-sition, benthic suspension-feeders were not influenced noticeably(Mayr et al., 2011). Furthermore, naturally and anthropogenicallyinduced changes in the nutrient ratios (N:P and Si:N) and loadsmay promote algal blooms which may reduce water quality, gener-ate hypoxia (Iriarte et al., 2010), and form hydrogen sulfide alteringboth the structure and function of benthic communities (Levinet al., 2009). On the other hand, temporal variations in the flux ofchemically reducing fluids (e.g. sulfide, methane), could alter thecommunity structure through the mass mortality of sessile organ-isms sensitive to these chemical compounds (i.e. cold-water cor-als), and also increase the biomass of these chemosyntheticmicroorganisms which can flourish rapidly in short periods (e.g.some days) (Taylor et al., 1999). All these processes would impactcarbon transfer in the benthic ecosystem through bottom-up andtop-down mechanisms.

Despite the above, low d13C values of the limpet Nacella deaura-ta (juvenile) and the polychaetes Nereis callaona could indicate anincorporation of chemosynthetic biomass. In addition, the broadrange in d13C values of the polychaetes Neanthes kerguelensis, thepolyplacophoran Chiton magnificus, the limpet Fissurella picta andthe tanaid Zeuxoides sp. could reflect an expansion in their trophicniches, including also chemosynthetic production in their diets. Inrelation to this, evidence of chemosynthetic carbon incorporationhas been reported by Sellanes et al. (2011) for similar invertebrategroups (i.e. gastropods, polyplacophoran, polychaetes, peracarids)in a shallow methane seep from central Chile. These authors alsoobserved direct ingestion of microbial filaments in some tanaida-cea taxa (e.g. Nototanais dimorphus and Zeuxo marmoratus). Inaddition, although C and N stable isotope data indicate a photosyn-thetic-based diet for all other consumers, new field observationsopen the possibility that the gastropod Nassarius gayi (Fig. 2C),the echinoids Pseudoechinus sp. and Arbacia dufresnii (Fig. 2D),and the asteroid Poraniopsis echinaster (Fig. 2E) can directly con-sume chemosynthetic bacteria attached to the hard substrate(e.g. rocks, dead corals, shells). This supports previous observationsfrom some deep-sea cold seeps, where mobile megafaunal taxa arerecognized as important consumers of chemosynthetic biomass(e.g. MacAvoy et al., 2002, 2005; Carney, 2010; Zapata-Hernándezet al., 2013).

Finally, potential use of terrestrial organic matter was notclearly detected in benthic consumers, given that this organic mat-ter could be easily traced due to its low d13C and d15N values. In

contrast, Vargas et al. (2011), suggests that POC from terrestrialsources may contribute to ca. 20–50% of body carbon in copepodsinhabiting Chilean fjord ecosystems further south of Comau fjord.Terrestrial organic matter (TOM) may enter indirectly into the ben-thic food web throughout pelagic pathways (e.g. copepod grazing;Vargas et al., 2011) or as detritus assimilated via heterotrophicbacteria (McLeod and Wing, 2009). Further analysis of the foodweb using other techniques like sulfur stables isotopes ratios(d34S) could help to better corroborate this assertion, given thatthey undergo low fractionation through the food web and gener-ally d34S values exhibit significant differences in primary producersaccording to C-fixation pathways (Connolly et al., 2004 and refer-ences therein).

4.2. Trophic position of benthic consumers

Low trophic positions were estimated in some carnivorous (i.e.nemertean, gastropods, polychaetes and asteroids). A similar pat-tern has been observed for some taxa (i.e. asteroids, nemerteanand polychaetes) at Bouvet Island (Jacob et al., 2006) and the ArcticKongsfjorden fjord (Renaud et al., 2011). These authors suggestthat these species have a wide alimentary flexibility or may befeeding on prey with low d15N values. However, it is accepted thatthe trophic enrichment factor varies with some characteristics ofthe food (e.g. protein quantity and quality) or of the organism(e.g. specific tissue, level of intake, growth rate or metabolic rate)(Robbins et al., 2010 and references therein). Therefore, some ofthese taxa could have a trophic enrichment factor less than theaverage 3.4‰ proposed by Post (2002). For instance, the biochem-ical form of nitrogen assimilation and excretion in different taxo-nomic groups may have an influence on the degree of 15Nfractionation (Vanderklift and Ponsard, 2003), but clearly morestudies under controlled conditions are needed to fix the varying15N fractionations between these groups. Despite this, trophic posi-tions of consumers are generally supported by functional groupsassigned from literature and the trophic enrichment used fits wellfor numerous taxa. In our study, several consumers also reportedsimilar trophic position in both sites, further confirming Crepipatel-la dilatata as a good reference consumer.

Some suspension-feeder and grazer taxa at both sites, had lowd15N values, probably associated to the consumption of sourceswith low d15N values. In this sense, some of the variability ind15N values of suspension-feeders could be related to the incre-ment of size of particulate organic matter (seston) since d15Nvalues of seston increase with particle size in Comau fjord (Mayret al., 2011). Therefore, we suggest that low d15N values could beassociated to the consumption of organic particles of small size-fractions (nano and pico-plankton). These particles would form anutritional component of low trophic levels for selected species(Rau et al., 1990). On the other hand, the low d15N values in grazerscould be associated to grazing on bacterial biofilms, microalgae ormacroalgae propagules attached to substrate (Grall et al., 2006).However, juveniles of the limpet N. deaurata clearly could havebeen grazing on chemosynthetic bacteria filaments, indicated bytheir low d13C values with respect to other benthic consumersand a partial contribution in their diet (Fig. B.2, Table A.2). In fact,the patchy distribution of bacterial mats and the presence of thesefilaments over their shells (Fig. 2F), as observed for other mollusks(Fig. 2B), could be indicating that this species can easily adapt topermanently inhabit inside these microbial mats, where bacteriamight reduce the animal exposure to hydrogen sulfide, eventuallyacting as an environmental detoxifier (Gallardo et al., 1994), allow-ing thus the survival of fauna in such conditions, and also servingas a refuge for other invertebrate taxa (Bailey et al., 2011).

Finally, similar carbon sources and trophic positions in the mostabundant fishes at XH, the Chilean sandperch Pinguipes chilensis

G. Zapata-Hernández et al. / Progress in Oceanography 129 (2014) 149–158 157

and the Patagonian redfish Sebastes oculatus, suggest a similarpredatory role at a high trophic level. Both species can consumeinvertebrate taxa (crustacean principally) and small fishes (Ojedaand Fariña, 1996; Pérez-Matus et al., 2012; Haussermann and Förs-terra, 2009) and have been characterized as specialists in theirdiets (both species; Barrientos et al., 2006; Pérez-Matus et al.,2012) or generalist (only P. chilensis; González and Oyarzún,2003). Nevertheless, traditional gut content analysis and a repre-sentative isotopic data of each specie could be useful for corrobo-rate their trophic behaviors in fjord ecosystems and to determineadditional population trophic attributes (e.g. trophic overlapping,trophic niche width, ontogenic and sexual variation in their diet,among others), which are necessary for a better understanding oftheir respective roles in the benthic communities at Comau fjord.

4.3. Isotopic niche width in trophic guilds

Similar CD and SDNND values in the suspension-feeder and car-nivore trophic guilds at both sites indicate similar characteristics ofspacing and evenness of packaging between species (distributionof trophic niches), respectively (Layman et al., 2007b), which couldindicate a similar heterogeneity on the exploitation of food re-sources at both sites. On the other hand, grazers from XH probablyexploit available food sources in a more heterogeneous way thanspecies present at PG. The latter could be directly related to theincorporation of a wide range of carbon sources (probably fromphotosynthetic to chemosynthetic origin), which could be reflectedin their wider isotopic niche (SEAc measure).

Small differences in SEAc measures in suspension-feeders andcarnivores suggest slight differences in their respective trophicniche, indicating also a similar magnitude of exploitation of foodsources at both sites. Furthermore, the overlapping area of theellipses indicates that niches of species present a high level ofsimilarity. In contrast, the grazer guild from XH has a broaderexpansion of its trophic niche, pronounced by the presence ofspecies with lower d13C and higher d15N values, possibly associatedto the consumption of a mix of food sources (e.g. photosynthetic,chemosynthetic and terrestrial carbon).

5. Conclusions

Nitrogen isotope analysis has demonstrated that the benthicecosystem of the Comau fjord comprises four trophic levels.Whether the food web is maintained mainly by autochthonousprimary production, or if terrestrial sources and cold vents alsoplay a role as energy source, was resolved using carbon isotopeanalysis. Carbon isotope values of terrestrial organic matter andbacterial filaments from cold vents were markedly lower thanautochthonous particulate organic matter. Thus, these isotopic dif-ferences can be traced in the food web. However, results suggestthat only a few benthic grazers used chemosynthetic food sources.For example, low d13C values of juveniles of the limpet N. deauratatogether with field observations, indicated that their food shouldconsist partially of bacterial filaments. This leaves open thepossibility that other species (primarily grazers) are also able toconsume chemosynthetic production. Finally, in contrast toprevious studies in other Chilean fjords, the influence of terrestrialcarbon sources for benthic communities in the Comau fjord isconsidered relatively low.

Acknowledgments

We thank the staff at Huinay Scientific Field Station (HSFS) fortheir help and hospitality, in particular for support duringfield work. Thanks to Daniel Genter, Yonatan González, Kaitlin

McConnell, Mauricio Melipillán and Ulrich ‘Ulo’ Pörschmann.Special thanks go also to Javier Naretto for his help during divingoperations and laboratory work. We acknowledge Camila Henríquezfor her help in the preparation of isotope samples and LaurentiusSauer is thanked for assistance in the isotope laboratory. This isContribution No. 90 of HSFS. This work was funded by FONDECYTProject No. 1120469 to JS.

Appendix A. Supplementary material

Supplementary data associated with this article can be found, inthe online version, at http://dx.doi.org/10.1016/j.pocean.2014.03.005.

References

Bailey, J.V., Salman, V., Rouse, G.W., Schulz-Vogt, H.N., Levin, L.A., Orphan, V.J., 2011.Dimorphism in methane seep-dwelling ecotypes of the largest known bacteria.The ISME Journal 5, 1926–1935.

Barrientos, C.A., González, M.T., Moreno, C.A., 2006. Geographical differences in thefeeding patterns of red rockfish (Sebastes capensis) along South American coasts.Fishery Bulletin 104, 489–497.

Belmar, C., Ferrada-Fuentes, S., Ruiz-Tagle, N., Gallardo, M.J., Ugalde, J., Fonseca, A.,Alonso, A., Gómez-Uchida, D., Musleh, S., Espinoza, C., Gallardo, V.A., 2012.Caracterización filogenética y morfológica de megabacterias hidrotermalesmarina del fiordo Comau, X región, Chile. XXXII Congreso de Ciencias del Mar,Punta Arenas, Chile.

Cabana, G., Rasmussen, J.B., 1996. Comparison of aquatic food chains using nitrogenisotopes. Proceedings of the National Academy of Sciences 93, 10844–10847.

Carlier, A., Riera, P., Amouroux, J.M., Bodiou, J.Y., Gremare, A., 2007. Benthic trophicnetwork in the Bay of Banyuls-sur-Mer (northwest Mediterranean, France): anassessment based on stable carbon and nitrogen isotopes analysis. Estuarine,Coastal and Shelf Science 72, 1–15.

Carlier, A., Ritt, B., Rodrigues, C.F., Sarrazin, J., Olu, K., Grall, J., Clavier, J., 2010.Heterogeneous energetic pathways and carbon sources on deep easternMediterranean cold seep communities. Marine Biology 157, 2545–2556.

Carney, R.S., 2010. Stable isotope trophic patterns in echinoderm megafauna inclose proximity to and remote from Gulf of Mexico lower slope hydrocarbonseeps. Deep-Sea Research II 57, 1965–1971.

Chaparro, O.R., Thompson, R.J., Pereda, S.V., 2002. Feeding mechanisms in thegastropod Crepidula fecunda. Marine Ecology Progress Series 234, 171–181.

Chaparro, O.R., Montory, J.A., Pereda, S.V., Thompson, R.J., Rivera, G., Osores, S.J.A.,Segura, C.J., 2013. Particle selection and regulation of particle uptake by theslipper limpet Crepipatella fecunda. Marine Ecology Progress Series 474, 167–177.

Cline, J.D., 1969. Spectrophotometric determination of hydrogen sulfide in naturalwaters. Limnology and Oceanography 14, 454–458.

Connolly, R.M., Guest, M.A., Melville, A.J., Oakes, J.M., 2004. Sulfur stable isotopesseparate producers in marine food-web analysis. Oecologia 138, 161–167.

DeNiro, M., Epstein, S., 1978. Influence of diet on the distribution of carbon isotopesin animals. Geochimica et Cosmochimica Acta 42, 495–506.

Gallardo, V., Canete, J.I., Roa, R., Enriquez-Briones, S., Baltazar, M., 1994. Recruitmentof the squat lobster Pleuroncodes monodon on the continental shelf off CentralChile. Journal of Crustacean Biology 14, 665–669.

González, P., Oyarzún, C., 2003. Diet of the Chilean sandperch, Pinguipes chilensis(Perciformes, Pinguipedidae) in southern Chile. Journal of Applied Ichthyology19, 371–375.

Grall, J., Le Loc’h, F., Guyonnet, B., Riera, P., 2006. Community structure and foodweb based on stable isotopes (d15N and d13C) analysis of a North EasternAtlantic maerl bed. Journal of Experimental Marine Biology and Ecology 338, 1–15.

Haussermann, V., Försterra, G., 2009. Marine Benthic Fauna of Chilean Patagonia.Nature in Focus, Santiago, Chile.

Häussermann, V., Försterra, G., Plotnek, E., 2012. Sightings of marine mammals andbirds in the Comau fjord, Northern Patagonia between 2003 and mid 2012.SPIXIANA 35 (2), 247–262.

Hügler, M., Sievert, S.F., 2011. Beyond the Calvin Cycle: autotrophic carbon fixationin the ocean. The Annual Review of Marine Science 3, 261–289.

Iken, K., Brey, T., Wand, U., Voigt, J., Junghans, P., 2001. Food web structure of thebenthic community at the Porcupine Abyssal Plain (NE Atlantic): a stableisotope analysis. Progress in Oceanography 50, 383–405.

Iken, K., Bluhm, B., Dunton, K., 2010. Benthic food-web structure under differingwater mass properties in the southern Chukchi Sea. Deep-Sea Research II 57,71–85.

Iriarte, J.L., González, H.E., Nahuelhual, L., 2010. Patagonian fjord ecosystems insouthern Chile as a highly vulnerable region: problems and needs. Ambio 39,463–466.

Jackson, A.L., Parnell, A.C., Inger, R., Bearhop, S., 2011. Comparing isotopic nichewidths among and within communities: SIBER – Stable Isotope BayesianEllipses in R. Journal of Animal Ecology 80, 595–602.

158 G. Zapata-Hernández et al. / Progress in Oceanography 129 (2014) 149–158

Jackson, M.C., Donohue, I., Jackson, A.L., Britton, J., Harper, D.M., Grey, J., 2012.Population-level metrics of trophic structure based on stable isotopes and theirapplication to invasion ecology. PLoS One 7 (2), e31757. http://dx.doi.org/10.1371/journal.pone.0031757.

Jacob, U., Brey, T., Fetzer, I., Kaehler, S., Mintenbeck, K., Dunton, K., Beyer, K., Struck,U., Pakhomov, E.A., Arntz, W.E., 2006. Towards the trophic structure of theBouvet Island marine ecosystem. Polar Biology 29, 106–113.

Layman, C.A., Quattrochi, J.P., Peyer, C.M., Allgeier, J.E., 2007a. Niche width collapsein a resilient top predator following ecosystem fragmentation. Ecology Letters10, 937–944.

Layman, C., Arrington, D.A., Montana, C.G., Post, D.M., 2007b. Can stable isotoperatios provide for community-wide measures of trophic structure? Ecology 88,42–48.

Lenk, S., Arnds, J., Zerjatke, K., Musat, N., Amann, R., Mußmann, M., 2011. Novelgroups of Gammaproteobacteria catalyse sulfur oxidation and carbon fixation ina coastal, intertidal sediments. Environmental Microbiology 13 (3), 758–774.http://dx.doi.org/10.1111/j.1462-2920.2010.02380.x.

Levin, L.A., Mendoza, G.F., 2007. Community structure and nutrition of deepmethane-seep macrobenthos from the North Pacific (Aleutian) Margin and theGulf of Mexico (Florida Escarpment). Marine Ecology 28, 131–151.

Levin, L.A., Michener, R., 2002. Isotopic evidence of chemosynthesis-based nutritionof macrobenthos: the lightness of being at Pacific methane seeps. Limnologyand Oceanography 47, 1336–1345.

Levin, L.A., Ekau, W., Gooday, A.J., Jorissen, F., Middelburg, J.J., Naqvi, S.W.A., Neira,C., Rabalais, N.N., Zhang, J., 2009. Effects of natural and human-induced hypoxiaon coastal benthos. Biogeosciences 6, 2063–2098.

MacAvoy, S.E., Carney, R.S., Fisher, C.R., Macko, S.A., 2002. Use of chemosyntheticbiomass by large mobile benthic predators in the Gulf of Mexico. MarineEcology Progress Series 225, 65–78.

MacAvoy, S., Fisher, C., Carney, R., Macko, S., 2005. Nutritional associations amongfauna at hydrocarbon seep communities in the Gulf of Mexico. Marine EcologyProgress Series 292, 51–60.

Mayr, C., Försterra, G., Häussermann, V., Wunderlich, A., Grau, J., Zieringer, M.,Altenbach, A., 2011. Stable isotope variability in a Chilean fjord food web:implications for N- and C-cycles. Marine Ecology Progress Series 428, 89–104.

McLeod, R.J., Wing, S.R., 2009. Strong pathways for incorporation of terrestriallyderived organic matter into benthic communities. Estuarine, Coastal and ShelfScience 82, 645–653.

Melzer, R., Schrödl, M., Häussermann, V., Försterra, G., Montoya, M.F., 2006.Pycnogonids on cnidarians at fjord Comau, Southern Chile: a report on 2005SCUBA collections. SPIXIANA 29, 237–242.

Minagawa, M., Wada, E., 1984. Stepwise enrichment of 15N along food chains:further evidence and the relation between d15N and animal age. Geochimica etCosmochimica Acta 48, 1135–1140.

Nilsen, M., Pedersen, T., Magnus, E., Fredriksen, S., 2008. Trophic studies in a high-latitude fjord ecosystem – a comparison of stable isotope analyses (d13C andd15N) and trophic-level estimates from a mass-balance model. Canadian Journalof Fisheries and Aquatic Science 65, 2791–2806.

Ojeda, F.P., Fariña, J.M., 1996. Temporal variations in the abundance, activity, andtrophic patterns of the rockfish, Sebastes capensis, off the central Chilean coast.Revista Chilena de Historia Natural 69, 205–211.

Parnell, A., Inger, R., Bearhop, S., Jackson, A., 2010. Source partitioning using stableisotopes: coping with too much variation. PLoS One 5, e9672.

Pérez-Matus, A., Pledger, S., Díaz, F.J., Ferry, L.A., Vásquez, J.A., 2012. Plasticity infeeding selectivity and trophic structure of kelp forest associated fishes fromnorthern Chile. Revista Chilena de Historia Natural 85, 29–48.

Post, D.M., 2002. Using stable isotopes to estimate trophic position: models,methods, and assumptions. Ecology 83, 703–718.

Quiroga, E., Ortiz, P., Gerdes, D., Reid, B., Villagran, S., Quiñones, R., 2012. Organicenrichment and structure of macrobenthic communities in the glacial BakerFjord, Northern Patagonia, Chile. Journal of Marine Biological Association of theUnited Kingdom 92 (1), 73–83.

R Development Core Team, http://www.R-project.org/, 2013. R: A Language andEnvironment for Statistical Computing. R Foundation for Statistical Computing,Vienna, Austria. ISBN 3-900051-07–0.

Rau, G.H., Teyssie, J.-L., Rassoulzadegan, F., Fowler, S.W., 1990. C13/C12 and N15/N14

variations among size fractionated marine particles: implications for theirorigin and trophic relationships. Marine Ecology Progress Series 59, 33–38.

Reid, W.D.K., Sweeting, C.J., Wigham, B.D., Zwirglmaier, K., Hawkes, J.A., McGill,R.A.R., Linse, K., Polunin, N.V.C., 2013. Spatial differences in East Scotia ridgehydrothermal vent food webs: influences of chemistry, microbiology andpredation on trophodynamics. PLoS One 8 (6), e65553. http://dx.doi.org/10.1371/journal.pone.0065553.

Renaud, P.E., Tessmann, M., Evenset, A., Christensen, G.N., 2011. Benthic food webstructure of an Arctic fjord (Kongsfjorden, Svalbard). Marine Biology Research 7(1), 13–26.

Robbins, C.T., Felicetti, L.A., Florin, S.T., 2010. The impact of protein quality on stablenitrogen isotope ratio discrimination and assimilated diet estimation. Oecologia162, 571–579.

Rooney, N., McCann, K., Gellner, G., Moore, J.C., 2006. Structural asymmetry and thestability of diverse food webs. Nature 442, 265–269.

Sánchez, N., Gonzales, H.E., Iriarte, J.L., 2011. Trophic interactions of pelagiccrustaceans in Comau fjord (Chile): their role in the food web structure. Journalof Plankton Research 33 (8), 1212–1229.

Schlitzer, R., http://odv.awi.de.2012, 2012. Ocean Data View.Schwabe, E., Försterra, G., Häussermann, V., Melzer, R., Schrödl, M., 2006. Chitons

(Mollusca: Polyplacophora) from the southern Chilean Comau fjord, withreinstatement of Tonicia calbucensis Plate, 1897. Zootaxa 1341, 1–27.

Sellanes, J., Zapata-Hernández, G., Pantoja, S., Jessen, G.L., 2011. Chemosynthetictrophic support for the benthic community at an intertidal cold seep site atMocha Island off central Chile. Estuarine, Coastal and Shelf Science 95, 431–439.

Taylor, C.D., Wirsen, C.O., Gaill, F., 1999. Rapid Microbial Production of FilamentousSufur Mats at Hydrothermal Vents. Applied and Environmental Microbiology 65(5), 2253–2255.

Ugalde, J.A., Gallardo, M.J., Belmar, C., Muñoz, P., Ruiz-Tagle, N., Ferrada-Fuentes, S.,Espinoza, C., Allen, E.E., Gallardo, V.A., 2013. Microbial life in a fjord:metagenomic analysis of a microbial mat in Chilean Patagonia. PLoS One 8(8), e71952. http://dx.doi.org/10.1371/journal.pone.0071952.

Van Dover, C.L., 2008. Stable isotope studies in marine chemoautotrophically basedecosystems: an update. In: Michener, R., Lajtha, K. (Eds.), Stable Isotopes inEcology and Environmental Science, second ed. Blackwell Publishing Ltd.,Oxford, UK, http://dx.doi.org/10.1002/9780470691854.ch8.

Vander Zanden, M.J., Rasmussen, J.B., 1999. Primary consumer d13C, d15N and thetrophic position of aquatic consumers. Ecology 80, 1395–1404.

Vanderklift, M.A., Ponsard, S., 2003. Sources of variation in consumer-diet d15Nenrichment: a meta-analysis. Oecologia 136, 169–182.

Vargas, C.A., Martinez, R.A., San Martin, V., Aguayo, M., Silva, N., Torres, R., 2011.Allochthonous subsidies of organic matter across a lake–river–fjord landscapein the Chilean Patagonia: implications for marine zooplankton in inner fjordareas. Continental Shelf Research 31, 187–201.

Yamamoto, M., Takai, K., 2011. Sulfur metabolism in epsilon- and gamma-Proteobacteria in deep-sea hydrothermal fields. Frontiers in Microbiology 2.http://dx.doi.org/10.3389/fmicb.2011.00192.

Zapata-Hernández, G., Sellanes, J., Thurber, A., Levin, L., Chazalon, F., Linke, P., 2013.New insights on the trophic ecology of bathyal communities from the methaneseep area off Concepción, Chile (�36 S). Marine Ecology. http://dx.doi.org/10.1111/maec.12051.