Stable isotopes as tracers of trophic interactions in marine mutualistic symbioses

Ecology and Evolution. 2019;9:723–740.   | 723 www.ecolevol.org 1 | INTRODUCTION The term symbiosis or “the living together of different organisms” according to the original deBary definition (De Bary, 1879) is fre‐ quently used to describe a relationship in which all partners mu‐ tually benefit from the association. Later on, symbiosis has been described as a shifting continuum from mutualism to parasitism (Thrall, Hochberg, Burdon, & Bever, 2007). In mutualistic nutritional symbioses, both partners take advantage of being together by devel‐ oping metabolic or nutritional interactions (Table 1). The exchange of nutrients allows the host and the symbionts to acquire nutrients that are limiting growth and reproduction, to expand their metabolic portfolio and, thus, the width and number of ecological niches to exploit, thereby avoiding competition with sympatric non‐symbiotic species. Mutualistic nutritional symbioses are widespread in nature. It is common to find plants that are associated with microorganisms that sequester nitrogen or phosphorus while receiving by‐products of host’s photosynthesis (Lugtenberg, 2013; van Rhijn & Vanderleyden, 1995). In marine ecosystems, nutritional symbioses exist throughout Received:14June2018 | Revised:17September2018 | Accepted:26October2018 DOI:10.1002/ece3.4712 REVIEW Stable isotopes as tracers of trophic interactions in marine mutualistic symbioses Christine Ferrier‐Pagès1 | Miguel Costa Leal2 ThisisanopenaccessarticleunderthetermsoftheCreativeCommonsAttributionLicense,whichpermitsuse,distributionandreproductioninanymedium, provided the original work is properly cited. ©2018TheAuthors.Ecology and EvolutionpublishedbyJohnWiley&SonsLtd. 1CentreScientifiquedeMonaco,Monaco, Monaco 2MARE–MarineandEnvironmental SciencesCentre,FaculdadedeCiênciasda Universidade de Lisboa, Lisbon, Portugal Correspondence MiguelCostaLeal,MARE–Marineand EnvironmentalSciencesCentre,Faculdade deCiênciasdaUniversidadedeLisboa, Campo Grande, Lisbon, Portugal. Email:miguelclea[email protected] Funding information FundaçãoparaaCiênciaeTechnologia, Grant/AwardNumber:SFRH/ BPD/99819/2014;CentreScientifiquede Monaco,Grant/AwardNumber:99819 and2014;FoundationforScienceand Technology Abstract Mutualistic nutritional symbioses are widespread in marine ecosystems. They involve the association of a host organism (algae, protists, or marine invertebrates) with sym‐ bioticmicroorganisms,suchasbacteria,cyanobacteria,ordinoflagellates.Nutritional interactions between the partners are difficult to identify in symbioses because they onlyoccurinintactassociations.Stableisotopeanalysis(SIA)hasproventobeause‐ ful tool to highlight original nutrient sources and to trace nutrients acquired by and exchangedbetweenthedifferentpartnersoftheassociation.However,althoughSIA has been extensively applied to study different marine symbiotic associations, there is no review taking into account of the different types of symbiotic associations, how theyhavebeenstudiedviaSIA,methodologicalissuescommonamongsymbioticas‐ sociations, and solutions that can be transferred from one type of association with another. The present review aims to fill such gaps in the scientific literature by sum‐ marizing the current knowledge of how isotopes have been applied to key marine symbioses to unravel nutrient exchanges between partners, and by describing the difficulties in interpreting the isotopic signal. This review also focuses on the use of compound‐specific stable isotope analysis and on statistical advances to analyze sta‐ ble isotope data. It also highlights the knowledge gaps that would benefit from future research. KEYWORDS compound‐specific stable isotope analysis, marine symbioses, mixing models, δ13C, δ15N 724 | FERRIER‐PAGÈS And LEAL the pelagic and benthic environments. They involve the association of a diverse range of algae, protists, sponges, sea squirts, corals, worms, and other marine invertebrates, with microorganisms such as bacteria (see reviews by Cavanaugh, 1994; Cavanaugh, McKiness, Newton,&Stewart,2006;Petersen&Dubilier,2009),cyanobacte‐ ria(reviewedinCarpenter&Foster,2002),anddinoflagellates(see reviews by Goodson, Whitehead, & Douglas, 2001; Venn, Loram, & Douglas, 2008). In most associations (except for algae), the hosts are heterotrophic and prey on a wide range of particles to meet their nu‐ tritional demand. They, however, complement their nutrient intake by hosting symbionts, which provide them with carbon and other mineral sources in exchange for protection and access to catabolic products(Figure1;Vennetal.,2008).Insomecases,afractionof the heterotrophically acquired nutrients is transferred from the host to the symbionts for their own needs (Tremblay, Gori, Maguer, Hoogenboom, & Ferrier‐Pagès, 2016). The amount and quality of nutrients exchanged between the partners can vary with environ‐ mentalconditions(Baker,Freeman,Wong,Fogel,&Knowlton,2018; Shantz,Lemoine,&Burkepile,2016),aswellaswiththeidentityof thehostand/orsymbionts(Lealetal.,2015).Somehost–symbiont associations can be species‐specific, with one host species being as‐ sociated with a single symbiont species, as it is observed for some scleractinian corals that associate with particular Symbiodinium spe‐ cies(Abrego,Ulstrup,Willis,&Oppen,2008;Baker,2003).However, some invertebrates, such as sponges, can host extremely complex and diverse symbiont communities that are not strictly pairwise, or even endosymbiotic. Indeed, some marine sponges harbor complex communities of generalist symbionts that live associated with the host,butnotnecessarily withinthe host’scells (Erwin &Thacker, 2007). Nevertheless, there are also marine sponges hosting spe‐ cificsymbiontspecies(Wilkinson,Nowak,Austin,&Colwell,1981). Eventhoughsomeoftheassociationsarespecies‐specific,thehost may select best performing symbionts from a population of possible partners within the same species, a process known as partner choice (Akçay, 2017; Sachs, Mueller, Wilcox, & Bull, 2004). For example, scleractinian corals can be associated with different dinoflagellate clades of Symbiodinium, depending on the prevailing environmental conditions (Little, van Oppen, & Willis, 2004; Thornhill, Howells, Wham,Steury,&Santos,2017).Overall,livingwithsymbioticpart‐ ners and having access to different nutritional pathways is funda‐ mental to many marine organisms, particularly those in nutrient‐poor environments. Nutritional interactions between the partners are particularly difficult to identify in endosymbioses because they only occur in in‐ tactassociations.Oncethe partnersareisolated, their physiology changes completely. Additionally, exchanged nutrients are usually metabolites that cannot be tracked through visual observation of feeding behavior or gut contents, as often performed for address‐ ingtrophicinteractions(Calado&Leal,2015;Leal&Ferrier‐Pagès, 2016;Nielsen,Clare,Hayden,Brett,&Kratina,2018).Nevertheless, TABLE 1 Nutritionalbenefitsinmutualisticmarinesymbioses Partner Nutrient accessibility Nutritional functions Symbionts Acquisitionofinorganicnutrients(carbon,nitrogen)and dissolvedorganicmatter(DOM) Accesstocarbon‐andnitrogen‐richhostwasteproducts Acquisitionofdiazotrophic‐derivednitrogen Acquisitionorsynthesisofessentialmetalsandvitamins Provision of entire symbiotic association by photosynthetically or chemosynthetically fixed carbon Transformationofinorganicnitrogenand/orN2intoorganic nitrogenous compounds Provision of vitamins and metals to the entire association Host Acquisitionofparticulateand/ororganicnutrientsby capture of prey Digestion and provision to the entire symbiotic association of essential carbon, nitrogen, and phosphorus compounds FIGURE 1 Nutritionalrelationshipsbetweenananimalhostanditsphototrophicandheterotrophicsymbionts | 725 FERRIER‐PAGÈS And LEAL our knowledge on nutritional exchanges in symbioses has bene‐ fited from recent methodological developments, such as proteom‐ ics, metabolomics, lipidomics, and isotopic biomarkers. Proteomics monitors change in protein expression of host and symbionts under different symbiotic states or environmental conditions, but requires genetic information on the studied models and advanced technology, suchasliquidchromatography/electrosprayionizationtandemmass spectrometry(Oakleyetal.,2016).Metabolomicsassesslowmolec‐ ular weight metabolite profiles (such as amino acids and lipids) of host and symbionts, and can detect fine‐scale changes in a rapid and quantitativemanner(Hillyer,Tumanov,Villas‐Bôas,&Davy,2016). These “omics” methods provide important information on the met‐ abolic competences of each symbiotic partner. However, they are not very accurate for tracing metabolite exchanges or at evidencing the original nutrient source (Middelburg, 2014). In contrast, stable isotopeanalysis(SIA)providesinformationontheoriginalnutrient source, which makes it one of the most helpful tools for assessing food web functioning, that is, tracing energy and nutrients from bottom producers to top predators, as well as for estimating trophic levels, resource use, and diet composition (Bouillon, Connolly, & Lee, 2008).Insymbioticassociations,SIAcantracenutrientsacquiredby the host or the symbionts both from internal and external inorganic and organic sources, including exogenous inorganic nutrient sources (Davy,Allemand,&Weis,2012;Ferrier‐Pagèsetal.,2011). SIAisbasedonbiologicallyactiveelementsexistinginmorethan one isotopic form. Generally, the lighter isotope is more abundant in the environment than the heavier isotope, but their relative abun‐ dance is altered by biological, geochemical, and anthropogenic pro‐ cesses(Rundel,Ehleringer,&Nagy,2012).Theseprocessesproduce variations in the stable isotope ratio of constitutive molecules of plant and animal tissues, which provide a good record for the exis‐ tence, and sometimes magnitude, of key processes involved with el‐ ementalcycling.Forinstance,thefractionationofnitrogenisotopes by consumers generates a gradient throughout the food web, with organisms in the bottom of the food chain displaying low δ15Nval‐ ues that gradually increase up through the food chain. This makes it possible to use this isotope to measure the trophic level of an indi‐ vidual. Information on such fractionation factor is, however, critical forusingSIAtocomparedistributionsofisotoperatiosbetweenthe animal host and its symbionts or to the consumed food. Numerousreviewshavebeenpublishedontheuseofstableiso‐ topes for coastal biogeochemistry (Bouillon et al., 2008), plant and animal ecology (Dawson, Mambelli, Plamboeck, Templer, & Tu, 2002; Martínez del Rio, Wolf, Carleton, & Gannes, 2009; Wolf, Carleton, & Martínez del Rio, 2009), and food web reconstructions (Boecklen, Yarnes, Cook, & James, 2011; Middelburg, 2014). However, an overarchingreviewfocusingontheuseofSIAtostudytrophicin‐ teractions in marine symbiotic associations is still missing. In par‐ ticular, there is still little information on how to interpret changes in isotopic signals of each symbiotic partner under different envi‐ ronmentaland/ornutritionalconditions,andhowtouseisotopesin natural abundance or in enrichment experiments tracing nutritional interactionsinsuchcomplexassociations.Aspreviouslystated,SIA has been applied to study different marine symbiotic associations, particularly chemosynthetic, photosynthetic, and nitrogen‐fixing symbioses. However, and despite sharing methodological and con‐ ceptual frameworks, there is no review that takes into account the different types of symbiotic associations, that addresses common issues,andthathighlightssimilarsolutions.Asanincreasingnumber ofresearchersareusingSIAtostudynutritionalinteractionsinma‐ rine symbiotic associations, it is important to summarize the current knowledge,highlightthedifficultiesinusingSIAinsuchsymbiotic models, and provide novel and broad insights arising from such an overarching perspective among the various marine symbiotic associ‐ ations. The present review aims to fill such gap in the scientific litera‐ tureby:i)summarizingthecurrentknowledgeofhowSIAhavebeen applied to some key marine symbioses to unravel nutrient exchanges between partners, both under natural abundance or in enrichment experiments; and ii) describing the difficulties in interpreting the isotopicsignalandusingSIAinsuchassociations.Thisreviewalso focuses on the use of compound‐specific stable isotope analysis, statistical advances to analyze stable isotope data, as well as high‐ lights the knowledge gaps that would benefit from future research. 2 | MAIN MARINE SYMBIOSES STUDIED USING STABLE ISOTOPES SIA has been widely applied to study four main types of marine endosymbioses: chemosynthetic, photosynthetic, nitrogen‐fixing symbioses, and the heterotrophic bacteria–sponge type symbiosis (Figure2).Inalltheserelationships,theanimalgenerallyfeedsonex‐ ternal particulate food sources such as phyto‐ and zooplankton or detrital organic matter and transfers a fraction of the heterotrophi‐ cally acquired nutrients to the symbionts. Such transfer has been evidenced by several studies using prey labeled with the stable iso‐ topes 13C and 15Nandshowingatransferofisotopesfromthehost to the symbionts (Hughes, Grottoli, Pease, & Matsui, 2010; Piniak, Lipschultz,&McClelland,2003;Tremblay,Maguer,Grover,&Ferrier‐ Pagès,2015).However,thetypeofnutrientsthataretransferredto the symbionts (e.g., sugars, lipids) is still an open question. The animal host also provides the symbionts with access to substrates (inorganic nutrients), which are necessary for their own generation of energy and biomass. In exchange, a portion of the inorganic nutrients fixed bythesymbiontsistransferredtothehostforitsownuse(Figure1). Firstdiscoveredathydrothermalvents,chemosynthetic symbioses between certain invertebrates (sponges, snails, mussels, clams, nematodes, tube worms, shrimps, and sea urchins) and chemoau‐ totrophic or methanotrophic bacteria are widespread in deep en‐ vironments and are responsible for the high biomass observed in such extreme environments (reviewedinCavanaughetal., 2006). Chemosynthetic symbionts are primary producers even in the ab‐ sence of light as they use a range of chemicals, such as reduced sulfur compounds, methane, and hydrogen as energy source to fix inorganic carbon or methane into organic molecules (Cavanaugh et al.,2006).SymbioticbacteriamainlyexchangewiththehostC1to 726 | FERRIER‐PAGÈS And LEAL C3 carbon compounds, which are then used for host energy and bio‐ synthesis. They are also able to take up ammonium or nitrate in their environment (Liao, Wankel, Wu, Cavanaugh, & Girguis, 2014) and even fix dinitrogen (Petersen et al., 2017). Photosynthetic symbioses are widespread associations in‐ volving cyanobacteria or microalgae symbionts. Cyanobacteria are mostly associated with marine sponges and diatoms (see re‐ viewsbyFreeman&Thacker,2011;Fosteretal.,2011;Vennet al., 2008), although they are also symbionts of haptophytes or dinoflagellates (Not et al., 2016). Prochloron, a unicellular pho‐ tosynthetic prokaryote, which is also part of the cyanobacteria phylum, is often associated with ascidians. Algae, such as dino‐ flagellates of the genus Symbiodinium, are associated with benthic animals such as reef‐building corals, sea anemones, tridacnid mol‐ luscs, jellyfish (Cassiopeasp.),andforaminifera(Freeman,Stoner, Easson,Matterson,&Baker, 2016;Sachs&Wilcox,2006; Venn et al., 2008). In the pelagic environment, photosymbiotic interac‐ tions also exist between microalgae and other protists (radiolar‐ ian, foraminifera) or metazoans (ciliates, dinoflagellates), although the exact nature of this partnership is often not formally demon‐ strated.Overall,forallthesephotosyntheticsymbioses,lightisof prime importance because it is needed by the symbionts as an en‐ ergy source to fix inorganic carbon into organic compounds called photosynthates, most of which are transferred to the host for its ownuse(Freemanetal.,2016).Inaddition,algalsymbiontsplaya major role in the acquisition of inorganic nitrogen (ammonium, ni‐ trate), phosphorus, and other macro‐ and micronutrients essential forthesymbiosis(Tanaka,Miyajima,&Koike,2006).Certaincya‐ nobacterial symbionts, called diazotrophs, can also fix dinitrogen (N2) in a symbiotic association (see below). ThethirdtypeofsymbiosisusuallystudiedwithSIAistheonede‐ veloped between animals or algae and diazotrophs. In this association, diazotrophs (nitrogen‐fixing bacteria or cyanobacteria) provide their hostwithnitrogen,whichisfixedthroughthereductionofN2toNH3 using the nitrogenase complex. Diazotrophs are known to establish sym‐ biosiswithvariousinvertebrates,suchassponges(Taylor,Radaz,Steger, &Wagner,2007),annelidworms(Stat,2016),corals(Bednarz,Grover, Maguer,Fine,&Ferrier‐Pagès,2017;Lema,Willis,&Bourne,2012),sea urchins (Guerinot & Patriquin, 1981), and protists such as dinoflagellates, diatoms,radiolarians,andtintinnids(Aminetal.,2015;Foster,Carpenter, &Bergman,2006;Foster,Subramaniam,&Zehr,2009). The last well‐studied symbiosis is the one between sponges and heterotrophic bacteria. Many marine organisms harbor dense and diverse microbial communities, but the in situ activity and func‐ tions of these microbes are still poorly known, except in sponges. Concerning nutritional functions, stable isotope experiments have, among others, demonstrated a role of bacterial symbionts in the ni‐ trogen cycle of sponges, in particular in nitrification, denitrification, and anaerobic ammonium oxidation (reviewed in Webster & Taylor, 2012). Genome sequencing also revealed that “Poribacteria” can un‐ dertake carbon fixation via the Wood–Ljungdahl pathway and pro‐ videthehostwithasourceofvitaminB12(Siegletal.,2011). 3 | STABLE ISOTOPES IN NATURAL ABUNDANCE FOR STUDYING NUTRITIONAL INTERACTIONS Studies on natural stable isotope abundance are based on the small differences in isotopic ratios as found in nature (Hayes, 2001; Peterson & Fry, 1987). These changes in stable isotopic ratios are caused by the preferential use of the light isotopes compared to the heavy ones in many biological and chemical processes, which is called isotopic fractionation. SIA provides critical information on carbon FIGURE 2 Main types of nutritional symbioses in the marine environment. (a)CO2,O2,SO3,SO4, H2S,DIN,CH2O: carbon dioxide, oxygen, sulfite, sulfate, hydrogen sulfide, dissolved inorganic nitrogen, saccharides; (b) CH4,NH4: methaneandammonium;(c)H2CO3,DIP, DFAA:bicarbonate,dissolvedinorganic phosphorus, dissolved free amino acids, (d)N2: dinitrogen | 727 FERRIER‐PAGÈS And LEAL and nitrogen origin, on the consumer’s trophic ecology and trophic level, as well as on the ecology and evolution of predators and their trophicrelationships(Bearhop,Adams,Waldron,Fuller,&MacLeod, 2004;Duarte,Flores,Vinagre,&Leal,2017).Forexample,aspecial‐ ized feeding strategy is evidenced by isotopic data that present low variability (or are homogeneous) among individuals, assuming that they are presented an isotopically consistent food source or mix of food sources. Conversely, large isotopic variability among individuals indicates that they specialize on different food sources or feed on isotopically distinct microhabitats in a heterogeneous “landscape”. δ13C (13C:12C) and δ15N(15N:14N)arethetwomostcommonstable isotopes commonly used for assessing nutritional interactions. In the fol‐ lowing sections, we describe these isotopes and their application to study nutritionalinteractionsinmarinemutualisticendosymbiosis.Other,less used stable isotopes, such as δ18Oandδ34S,arealsodescribed. 3.1 | δ13C isotopes δ13C is a very useful tool to help distinguish autotrophic metabolisms, because enzymes involved in carbon fixation pathways discrimi‐ nate differently against the use of the heavier carbon isotope (13C) (Cavanaugh et al., 2006). Consequently, different primary producer symbionts exhibit distinct δ13C values due to diverse carbon fixa‐ tionpathways(Figure3).Forexample,algaeandphytoplanktonpref‐ erentially assimilate the lighter isotope 12C and thus usually display δ13Cvaluesof−20‰to−18‰(Fry&Sherr,1989;Gearing,Gearing, Rudnick, Requejo, & Hutchins, 1984; Goericke, 1994). δ13C derived from chemosynthesis occurring at vents is either considerably lighter than phytoplankton δ13C (enriched in 12Cfrom−9‰to−16‰)orheav‐ ier (depleted in 12C,from−27‰to−35‰)(Fisher&Childress,1992; Levin & Michener, 2002; Robinson et al., 2003). These two extreme chemosynthetic groups show contrasting δ13C values due to the use of twodifferentformsofribulose‐1,5‐bisphosphatecarboxylase/oxyge‐ nase (Rubisco I and II), which catalyses the carbon fixation step of the Calvin‐Benson cycle. Rubisco form I discriminates more than Rubisco form II against 13C leading to lower, that is, more depleted, δ13C values (Robinson&Cavanaugh,1995).Finally,symbiosesinvolvingmethano‐ trophic bacteria can be even more depleted in 13C, with δ13C ranging from−37‰to−55‰forthermogenicmethane,andfrom−60‰to −80‰forbiogenicmethane(Barryetal.,2002;Cavanaugh,1993). The δ13C result can be used as an indicator of the main food sub‐ strate(s)becauseitbecomesenrichedbyonly0.4‰to1‰between each trophic level (Conway, 1994; McCutchan, Lewis, Kendall, & McGrath, 2003). Therefore, in symbiotic associations, the δ13C values of host and symbionts should be similar. Comparisons between δ13C values of symbionts and host tissue, or between symbiont‐containing and symbiont‐free host tissues, can be theoretically used to trace the exchange of carbon between the symbionts and their animal host, and to estimate the importance of symbiont‐derived carbon supplied to the host.Forexample,inthecoral–dinoflagellatesymbiosis,inwhichcar‐ bon is mainly acquired through the dinoflagellates photosynthesis, the host tissue will display δ13C values in the same range as the symbionts (algal)values,thatis,−11‰to−16‰(Alamaru,Loya,Brokovich,Yam, &Shemesh,2009).Inchemosyntheticsymbioses,andsimilartosym‐ biotic corals, some animals can totally rely on the carbon transferred by the symbionts. This is the case of the solemyid protobranch Solemya velum that shows a δ13Cvaluerangingfrom−38.4to−45.3‰(Conway & Capuzzo, 1991), the seep vestimentiferan tube worms and some methanotrophic mussels that show a δ13Cvaluerangingfrom−40‰ to‐65‰(KennicuttIIetal.,1992),andthyasiridclamsthatshowonav‐ erage a δ13Cvalueof−38‰(Fiala‐Médioni,Boulegue,Ohta,Felbeck, & Mariotti, 1993). In some situations, however, the δ13C values of the host can be difficult to interpret, because host tissues rarely have δ13C values solely influenced by the symbionts. Most hosts can indeed use alternative (external) sources of nutrition, such as predation on free‐ living zooplankton and phytoplankton, or uptake of dissolved organic material. Consequently, the δ13C value of the host tissue will be a mix ofalltheanimal’sdiet.Suchmixed dietisobserved,forinstance, in the western Pacific vent mussel Bathymodiolus brevior. In this spe‐ cies, the symbiont‐containing gill tissue showed δ13Cvalues(−30.8‰ to −35.8‰) significantly lower than those from symbiont‐free foot tissue, demonstrating that this species supplemented its diet via fil‐ ter feeding on external particles (Dubilier, Windoffer, & Giere, 1998). Someorganismsalsohostseveralsymbionts,suchasthehydrother‐ mal mussels Bathymodiolus azoricus that live in association with both thio‐ and methanotrophic bacteria, which ultimately confounds the δ13C value of the host (Cavanaugh, Wirsen, & Jannasch, 1992; Trask &VanDover,1999).Forinstance,small‐sizedindividualscandisplay δ13Cvaluesrangingfrom−27‰to−34‰,suggestingthatthiotrophy is the dominant nutritional pathway, with methanotrophy and filter feeding emerging as secondary strategies. However, higher δ13C val‐ ues were measured in larger mussels, suggesting that they rely more heavily on carbon from methanotrophic endosymbionts as they grow (De Busserolles et al., 2009). Several other factors can confound the identification of the food sources in symbiotic associations, such as unknown frac‐ tionation factors, the impossibility to isolate the symbionts and FIGURE 3 Different ranges of δ13C according to the C source 728 | FERRIER‐PAGÈS And LEAL determine their δ13C value, or the small carbon contribution of the symbionts to the nutrition of the symbiosis. This latter situation applies, for example, to the temperate gorgonian (Eunicella sin‐ gularis) that lives in symbiosis with Symbiodinium dinoflagellates, whose photosynthesis (and inorganic nutrient acquisition) is low all year round due to low irradiance experienced in temperate environments (Cocito et al., 2013). This gorgonian mostly relies on the host feeding and therefore shows δ13C (and δ15N)values closetothezooplanktonvaluesallyearround(−23‰and8‰for δ13C and δ15N, respectively), despite nutrient acquisition by the symbionts(Figure4).Althoughautotrophicallyacquirednutrients from inorganic sources are expected to be transferred between the symbionts and the host, this nutrient source may represent only a small fraction of the total ingested nutrients and cannot be traced using the natural isotopic values. The reverse is ob‐ tainedwithshallowwatertropicalscleractiniancorals(Figure4). These organisms have few planktonic prey in the water and have to derive most of their carbon from the symbiotic dinoflagellates (Muscatine, McCloskey, & Marian, 1981). The δ13C values of the coral tissue are similar to those of the symbionts, and any hetero‐ trophic input from the host will be masked by the large autotrophic inputfromthesymbionts(Bakeretal.,2015;Leal,Rocha,Anaya‐ Rojas,Cruz,&Ferrier‐Pagès,2017;Nahonetal.,2013;Reynaud etal.,2002).Finally,inautotrophicsymbioses,theδ13C signature of the symbiotic association also largely varies according to the photosynthetic rates of the symbionts, which is primarily asso‐ ciated with light irradiance: Carbon fractionation increases with decreasing irradiance, which leads to a lower δ13C signature under low light (Heikoop et al., 2000). In this latter condition, δ13C of the symbiotic partners will reach the values of living or detrital particulateorganicmatter(POM)suspendedinthewatercolumn, which makes it difficult to decipher if the predominant carbon source is autotrophic or heterotrophic. This is the case of scler‐ actinian corals, which can thrive from 5 m, where light reaches a daily mean of 500 µmol photons m−2 s−1, down to 150 m depth, with an irradiance of 20 µmol photons m−2 s−1. In shallow waters, the δ13Cofthecoraltissueissignificantlymorepositive(−10‰ to−14‰)thanthatofPOM(ca.−20‰)(Muscatine,Goiran,Land, & Jaubert, 2005; Swart, Saied, & Lam, 2005), which clearly in‐ dicates that the symbiont photosynthates are the main carbon source of the coral tissue. With increasing depth, the δ13C of coral tissues becomes more negative (−23‰) and approaches that of thezooplankton(Swartetal.,2005).Inthiscase,δ13C depletion can be due to a higher heterotrophic input following zooplankton ingestion (Muscatine, Porter, & Kaplan, 1989) or to a higher car‐ bonfractionationbythesymbiontswithdecreasinglight(Swart, 1983;Williams,Röttger,Schmaljohann,&Keigwin,1981),aswell as higher internal carbon cycling between the host and the symbi‐ onts(Einbinderetal.,2009).Theδ13C of a symbiotic association fi‐ nally varies with the δ13C values of the seawater inorganic carbon FIGURE 4 Isotopic variability in a temperate and tropical autotrophic symbiosis (dinoflagellate–gorgonian association). The isotopic signature of temperate organisms that live in a particle‐rich environment, and in which symbionts are not very active, will mirror the isotopic signature of the organic nutrients. In contrast, the isotopic signature of tropical gorgonians that live in oligotrophic and particle‐poor environments will be influenced by the activity of the symbionts and by their uptake of dissolved inorganic nitrogen. Isotopic data for the gorgoniantissueandorganicfoodarefromCocitoetal.(2013),forthetemperatesystemandfromWard‐Paige,Risk,andSherwood(2005), for the tropical system. Isotopic data for the inorganic carbon are from Gillikin and Bouillon (2007) and those for inorganic nitrogen are from York, Tomasky, Valiela, and Repeta (2007) | 729 FERRIER‐PAGÈS And LEAL sources.Forexample,whenseawaterhasahighpartialpressure incarbondioxide(pCO2),suchasinpCO2 vents, the δ13C values of the symbionts and the host can be significantly lighter than the signatures obtained under normal pCO2 levels (Horwitz, Borell, Yam,Shemesh,&Fine,2015).Thus,inthesymbioticseaanemone Anemonia viridis, the δ13Cvalueofsymbiontsshiftedfrom−15‰ incontrolsitesto−18‰inthepCO2 vent, wrongly suggesting that the sea anemones were more heterotrophic in the vents. Onewaytoestimatethecontributionofsymbiontstothediet of their host is to calculate the difference between the δ13C of the host (δ13CH) and of the symbionts (δ13CS), also named ∆δ13Chost‐sym‐ biont. The contribution of the symbionts to the diet of their host is inversely proportional to the ∆δ13Chost‐symbiont, or, in other words, the contribution will decrease from 100% to lower values when ∆δ13Chost‐symbiontincreasesfromzerotohighervalues.Forexample, low ∆δ13Chost‐symbiont was recorded in coral species from surface waters of Moorea Island (from −1.44±0.23‰ to 2.98±0.58‰), which are corals that entirely rely on their symbionts for their ener‐ geticneeds(Nahonetal.,2013).Inopposite,the∆δ13Chost‐symbiont of deepcoralsrelyingonheterotrophicfoodsourcescouldreach8‰ (Muscatine et al., 1989). Asonlyonestableisotopedoesnotalwaysallowdeterminingthe main food source in a mixed diet, most studies often combine several stable isotopes (Phillips, 2012). Particularly, the δ15Nvalueisoften combined with δ13C to obtain a better identification of the main food sources or to better address the nutrient exchanges between the symbionts and their host. 3.2 | δ15N isotopes The δ15N value of plants and animals varies according to two processes: assimilative and metabolic fractionation (Zanden & Rasmussen,2001).Assimilative fractionationresultsfromisotopic discrimination during nitrogen assimilation or isotopic differences between nitrogen pools. Metabolic discrimination is due to frac‐ tionation during amino acid transamination and deamination. During these processes in non‐symbiotic organisms, 14Naminegroupsare preferentially removed to produce isotopically light metabolites (Figure 5a), leaving the remaining nitrogen pool enriched in 15N (Gannes, O’Brien, & Del Rio, 1997). This metabolic fractionation induces an increase in the δ15Nvaluebetweenpreyandpredators (Figure5a),whichultimatelyallowsestimatingthetrophicposition of consumers (Zanden & Rasmussen, 1999, 2001 ). It is often as‐ sumed that the δ15Nvalueofaconsumerisenrichedby2.3‰–3.4‰ over that of its diet (Minagawa & Wada, 1984), but the fractionation factor is species‐specific, and can differ significantly from this value (discussedinZanden&Rasmussen,2001).δ15Nfractionationincar‐ nivores is indeed relatively stable and varies within a narrow range (meanof3.2‰±0.4‰),whereasδ15Nfractionationbetweenplants and herbivores is highly variable (mean 2.5‰±2.5‰; Zanden & Rasmussen,2001).Nostudyhas,however,determinedfractionation factors for symbiotic associations such as corals, probably because of the difficulties described below. In symbiotic organisms, the δ15Nvalueofthehost(consumer) does not follow the same rule as for asymbiotic organisms. Due to the assimilative fractionation, the δ15Nofthe“entire”symbioticas‐ sociation (i.e., host and symbionts) varies depending on the main nitrogen source(s) (atmospheric, dissolved, or particulate) used by each partner. The δ15Nvaluealsovarieswiththeimportanceofthe internalrecyclingbetweenthesymbiontsandthehost.Figure5bis an example of the differences between symbiotic and asymbiotic organisms due to the internal recycling. Compared to asymbiotic organisms, the 14Noftheparticulatefoodingestedbythehostis not excreted as waste product, but recycled by the symbiont, and FIGURE 5 Difference in the isotopic signature of non‐symbiotic and symbiotic organisms. In non‐symbiotic organisms, the light isotopes are removed first during the processes of respiration and excretion, leaving the heavy isotopes in the tissue. The tissue will thus be enrichedbyca.3‰in15Nand1‰in 13C at each trophic step. In symbiotic associations, nutrients are continuously recycled within the association, and the isotopic signature of the symbiosis will be a mix between the different food sources 730 | FERRIER‐PAGÈS And LEAL re‐transferred to the host, thereby “diluting” the heterotrophic 15N enrichment. This host‐mediated signal is also combined with the symbiont‐mediatedsignal.Symbiontstendtotakeupthelightiso‐ tope (14N)fromtheinorganicnutrientpool,leavingtheheavyiso‐ tope (15N)inseawater.Thislightisotopeisagaintranslocatedtothe host for its own needs. The final δ15Nvalueofthesymbioticassocia‐ tion will thus be a mix of all these nitrogen sources. In addition, sym‐ bionts and hosts can assimilate nitrogen sources with different δ15N values (summarized in Sigman, Karsh, & Casciotti, 2009). In sum‐ mary,atmosphericnitrogen(N2), which has a very low δ15Nvalue closeto0‰(Petersenetal.,2017),canonlybefixedbydiazotroph symbionts, which lowers the δ15Nvalueofthesymbioticassociation (France, Holmquist, Chandler, & Cattaneo, 1998). Dissolved inor‐ ganicnitrogensources(DIN),suchasammonium,nitrate,andurea, which are assimilated by the symbionts, have a higher δ15N,closerto 4‰–6‰.Thisisthesameforthedissolvedorganicsources,taken upbybothsymbiontsandhosts((Sigmanetal.,2009).Finally,par‐ ticulatedetritalorlivingnitrogensources(PON)takenupbyanimal hosts display an even higher δ15N value (6‰‐14‰), except when it originates from mangrove forests, where the values are lower (Corbisieretal.,2006;Kao,Tsai,Shih,Tsai,&Handley,2002;Riera, Stal,&Nieuwenhuize,2004).Insymbiotic associations,thesedif‐ ferent sources are often used in combination and continuously re‐ cycledwithintheassociation(Figure5b),whichmakeitdifficultto interpret the δ15Nvalueandtheoriginalnitrogensource.Inautotro‐ phic symbioses, for example, the δ15Nvalueisoftenamixbetween DINtakenupbythesymbionts(Grover,Maguer,Reynaud‐Vaganay, &Ferrier‐Pages,2002)andPONcapturedbythehost(Houlbrèque &Ferrier‐Pagès,2009),andisusuallyintherangeofthatofsurface organic material (4‰–10‰). It is, therefore, difficult to estimate which nitrogen source is predominant. In addition, photosynthetic processes affect the final δ15Nvalueofthesymbioticassociation. Asforcarbon,decreasinglightlevelsignificantlydecreasestheδ15N value of the symbiotic association (Baker, Kim, Andras, & Sparks, 2011; Heikoop et al., 1998). This is due to an important assimilation ofDINbythesymbiontsunderhighlight,whichstronglydepletes thehostDINpoolandleadstoareducedfractionationrelativeto external DIN. In symbiotic associations involving more than one symbiont type, such as those containing both autotrophic and di‐ azotrophic symbionts (Bednarz et al., 2017; Mohamed, Colman, Tal, & Hill, 2008), the δ15Nvaluewillbetheresultofbothsymbiont contributions. If the nitrogen supply by diazotrophs largely exceeds thatbythealgae/animal,theδ15Nvaluesoftheanimaltissuewillbe very low. This has been observed in autotrophic–symbiotic sponges that presented δ15N values ranging from 0‰ to 4‰ (Freeman & Thacker, 2011; Mohamed et al., 2008). A similar depletion in 15N (low to negative δ15Nvalues)hasbeenobservedinchemoautotro‐ phic and methanotrophic symbioses, because of an important con‐ tributionofN2 (Lee & Childress, 1994; Petersen et al., 2017) and an important biomass of methane‐oxidizing bacteria highly depleted in 15N(Macko,Fogel,Hare,&Hoering,1987).Finally,theδ15Nvalue of the symbiosiswill alsovarywith anthropogenic DIN pollution, with a consistent δ15Nenrichment insymbiontsofpollutedareas compared to oligotrophic ones (Baker, Murdoch, Conti‐Jerpe, & Fogel,2017;Wong,Duprey,&Baker,2017). 3.3 | Other stable isotopes δ34Sisnotcommonlyusedtostudynutritionalsymbioses,although it can be useful to trace the production of sulfur compounds such asDMS(P)bysymbionts.Thesecompoundsareinvolvedinmultiple physiological functions in algae and bacteria, such as osmoprotection (Motard‐Côté & Kiene, 2015), antioxidant defense (Sunda, Kieber, Kiene, & Huntsman, 2002), dissipation of excess energy (Stefels, 2000),amongothers.DMSPproductionhasbeenmonitoredinsym‐ biotic protists (Gutierrez‐Rodriguez et al., 2017) and anthozoans such ascorals(Gardner,Raina,Ralph,&Petrou,2017;Jones,Curran,Swan, & Deschaseaux, 2017). The interpretation of isotopic changes in δ34S duringDMSPcleavageintoDMSis,however,complex,becauseofdif‐ ferent fractionation factors between taxonomically different groups. Forexample,δ34S‐DMSPandδ34S‐DMSvaluesweresimilarinmicro‐ bialassemblagesofthe RedSea, whiletheδ34Svalueof DMSpro‐ ducedbysymbioticacantharianswas1.5%lowerthanthatofDMSP (Amrani,Said‐Ahmad,Shaked,&Kiene,2013).Inthesameway,there is usually a slight 34S‐depletioninDMSPinsymbioticorganismscom‐ pared to seawater SO4 2‐, even though a high enrichment was also observed in a protist–microalgae symbiosis (Gutierrez‐Rodriguez et al., 2017). δ34Scanalsobeusedtohelpdiscriminateorganicmatter sources, particularly between terrestrial and marine sources (Granek, Compton, & Phillips, 2009), as well as in hydrothermal systems where sulfur is critical for biogeochemical cycles (Kennicutt II et al., 1992). Overall,thesefewstudieswith,sometimes,oppositeresultsclearly show that more investigations are needed to understand the factors and processes affecting the δ34Svalueofsymbioticassociations. Future research involving stable isotopes in natural concen‐ trations should start focusing on combining the isotopes reviewed above with others that have been ignored in marine nutritional stud‐ ies.Forexample,changesinthestableisotopevalueofmetals,such as δ66Znandδ65Cu, have been widely used in mice and humans, to trace pathological conditions (Balter et al., 2015) or different di‐ etary conditions (Costas‐Rodríguez, Van Heghe, & Vanhaecke, 2014; Jaouen,Pons,&Balter,2013;Jaouen,Szpak,&Richards,2016).In corals, δ66Znandδ65Cu can also be used as tracers to record bleach‐ ing(Ferrier‐Pagès,Sauzéat,&Balter,2018).Theircombinationwith δ13C and δ15N,whosechangeswithenvironmentalconditionsortro‐ phic status are better explored, is likely to bring novel insights into the nutritional interactions of marine symbioses. 3.4 | Mixing models The estimation of the contribution of the different food sources to a consumer’s diet has largely benefited from Bayesian mixing models (Parnell et al., 2013; Phillips, 2012). These models take into account of the isotopic ratios of the consumers’ tissues and food sources, as well as the isotopic fractionation, or trophic enrichment factor, from the prey to the predator. Mixing models are, however, not always the | 731 FERRIER‐PAGÈS And LEAL best tools for tracing food sources, especially when the signature of the food sources largely overlaps in isotope space or when the con‐ sumers feed on a large diversity of prey. There are also other com‐ mon problems that often lead to an incorrect use of mixing models. Forinstance,thetypeoffoodconsumedbytheanimalhostisoften unknown, which can lead to the analysis of food sources that are ecologically irrelevant but that might show a stable isotope signature that fits within the stable isotopic niche potentially consumed by the target species. The application of mixing models to marine mutualistic sym‐ bioses is relatively scarce, with most studies targeting food web dynamics of deep‐sea benthic communities. Levin and Michener (2002) were among the firsts using mixing models to show that the combination of low δ15N and δ13C values evidenced chemoauto‐ trophic symbioses in bivalve and pogonophoran taxa. Using similar statistical methods, McLeod, Wing, and Skilton (2010) as well as Riekenberg,Carney,andFry(2016)assessedthecarboncontribu‐ tion from chemoautotrophic and methanotrophic symbiotic bacteria to bivalve nutrition. In tropical symbioses, a mixing model was used byFreemanandThacker(2011)toquantifythepercentageofsymbi‐ ont‐derivedversusPOM‐derivedcarbonassimilatedbyreefsponges with microbial symbionts. While stable isotope mixing models are becoming increasingly used in ecology and evolution studies, as well as becoming more complex and accurate (Phillips et al., 2014), their application to pro‐ vide new insights on the nutritional ecology of marine mutualistic symbioses is still poorly explored. This is likely associated with the poor number of studies using combined stable isotope values of the symbiotic host and symbionts, as well as the relatively low sample sizethatisusedinsuchstudies.Someofthesestudies(Cocitoet al.,2013;Ferrier‐Pagèsetal.,2011;Lealetal.,2014)calculatedthe theoretical food source based on subtracting a trophic enrichment factor from the δ13C and δ15Nvaluesofthehostorganism(1‰for δ13C, and 3.5‰ for δ15N). While this simple approach is scientifi‐ cally valid, it fails to consider several factors that are examined in mixing models, particularly the isotopic variability of the different food sources, the variability of trophic enrichment factors between consumers and the different food sources, and concentration de‐ pendence means, that is, the estimated proportion of carbon and nitrogen in each food source (Parnell et al., 2013). Moreover, the spa‐ tial distance between the theoretical food source and the isotopic values of the different food sources cannot be statistically analyzed usingthestableisotopevalueofthetheoreticalfoodsource.Amix‐ ing model approach can provide statistically robust estimates and confidence intervals for the contribution of each food source to the nutritionofthecoralhost.Futurestudiesthataimtoapplysuchmix‐ ing models should use a robust number of replicate samples to max‐ imizethestatisticalaccuracyofthemixingmodels.Second,theTEF used to fuel the model should be obtained empirically and not based onpreviousestimates.NotonlymoreaccurateTEFareneeded,but variability estimates are also important to improve the quality of the mixing model output. Third, and although it is not mandatory, the mixing model requires the proportion of carbon and nitrogen of each food source. This primary and fundamental information is not always availableasmoststudiesfailtoperformsuchsimpleanalyses.Forin‐ stance, while the carbon and nitrogen content of live planktonic or‐ ganisms is usually known, their proportion in an individual organism isusuallyunknown.Suchgapofknowledgeisimportanttoaddress in order to improve the estimates of stable isotope mixing models. 3.5 | Compound‐specific isotope analyses (CSIA) Compound‐specific isotope analysis (CSIA) has been increasingly used in ecological studies as a new tool for analyzing natural food webs, especially coupled with bulk stable isotope analysis. Compared toSIA,whichisbasedonthestableisotopevalueofthetotaltis‐ suesorthatoftotalplanktoncells,CSIAcorrespondstotheisotopic signatureoftheorganicmattercompounds,thatis,fattyacids(FA) andaminoacid(AA).CSIAisatthebiochemicalbuilding‐blocklevel and thus allows tracing the exchange of precise molecules in a food web(Evershedetal.,2007).Inaddition,theinterpretationofCSIA requiresfewerassumptionsthanbulkisotopicvalues.Specifically, the physiological forces affecting isotopic values of a single group of compounds are less numerous and often better understood than thediversityofforcesthatareknowntoaffectbulktissues.CSIA is, therefore, particularly more successful than the other trophic markers when (a) organisms cannot be physically isolated from each other, such as in symbiotic associations; (b) when there is a need to trace quantitatively minor but qualitatively important components; or (c) when different food sources have similar bulk δ13C signatures (Gladyshev,Sushchik,Kalachova,&Makhutova,2012). Fatty acid compound‐specific isotope analysis (CSIA‐FA) is based on the fact that δ13C values of a specific FA compound reflect its synthetic pathway and hence its source (Hayes, 1993; Teece,Fogel,Dollhopf,&Nealson,1999).Thesevalueswillthus be different when compounds are derived from direct biosynthe‐ sisorfromanindirectdietarysource(AbrajanoJr,Murphy,Fang, Comet,& Brooks, 1994;Fanget al., 1993).Essential fattyacids (EFA),suchasthe“omega‐6”and“omega‐3”FA(linoleicacid,ara‐ chidonic acid, eicosapentaenoic acid, and docosahexaenoic acid), cannot be directly synthesized by animals and have to be acquired through predation or symbiont transfer; therefore, the δ13C‐EFA values of the animal will be comparable to that of the symbionts or the external prey, since little or no isotopic fractionation oc‐ curs during this process (Treignier, Tolosa, Grover, Reynaud, & Ferrier‐Pagès,2009).Inthecaseofnon‐essentialFAs,whichcan be synthesized de novo by the animal, the δ13C values will reflect the competing processes of assimilation from external food and de novo synthesis (Gladyshev et al., 2012; Villinski, Hayes, Villinski, Brassell,&Raff,2004).Forthedenovosynthesisoffattyacids, transferase and desaturase induce a δ13C isotope depletion in the synthesized fatty acid, while elongase provide a δ13C enrichment inthesynthesizedfattyacid(Figure6).CSIA‐FAhasbeenusedin symbiotic associations for the first time to evaluate the dietary strategies of marine mytilids from a normal coastal ecosystem and fromacoldhydrocarbonseepecosystem(AbrajanoJretal.,1994). 738 | FERRIER‐PAGÈS And LEAL the regulation of seawater temperatures and climate in the Great Barrier Reef. American Journal of Climate Change, 6,328.https://doi. org/10.4236/ajcc.2017.62017 Kao, W.‐Y., Tsai, H.‐C., Shih, C.‐N., Tsai, T.‐T., & Handley, L. L. (2002). Nutrientcontents,δ13C and δ15Nduringleafsenescenceintheman‐ grove, Kandelia candel (L.) Druce. Botanical Bulletin of Academia Sinica, 43. Kennicutt II, M., Burke, R. Jr, MacDonald, I., Brooks, J., Denoux, G., & Macko,S.(1992).Stableisotopepartitioninginseepandventorgan‐ isms: Chemical and ecological significance. Chemical Geology: Isotope Geoscience Section, 101, 293–310. Kopp, C., Pernice, M., Domart‐Coulon, I., Djediat, C., Spangenberg, J., Alexander,D.,…Meibom,A.(2013).Highlydynamiccellular‐levelre‐ sponse of symbiotic coral to a sudden increase in environmental nitro‐ gen. Mbio, 4,e00052–e13.https://doi.org/10.1128/mBio.00052‐13 Kopp, C., Wisztorski, M., Revel, J., Mehiri, M., Dani, V., Capron, L., … Mouajjah,D.(2015).MALDI‐MSandNanoSIMSimagingtechniques to study cnidarian–dinoflagellate symbioses. Zoology, 118, 125–131. https://doi.org/10.1016/j.zool.2014.06.006 Leal, M. C., & Ferrier‐Pagès, C. (2016). Molecular trophic markers in marine food webs and their potential use for coral ecology. Marine Genomics, 29,1–7.https://doi.org/10.1016/j.margen.2016.02.003 Leal,M.C.,Ferrier‐Pagès,C.,Calado,R.,Brandes,J.A.,Frischer,M.E.,& Nejstgaard,J.C.(2014).Trophicecologyofthefacultativesymbiotic coral Oculina arbuscula. Marine Ecology Progress Series, 504, 171–179. https://doi.org/10.3354/meps10750 Leal,M.C.,Hoadley,K.,Pettay,D.T.,Grajales,A.,Calado,R.,&Warner, M.E.(2015).Symbionttypeinfluencestrophicplasticityofamodel cnidarian‐dinoflagellate symbiosis. Journal of Experimental Biology, 218,858–863.https://doi.org/10.1242/jeb.115519 Leal, M. C., Rocha, R. J., Anaya‐Rojas, J. M., Cruz, I. C., & Ferrier‐ Pagès, C. (2017). Trophic and stoichiometric consequences of nu‐ trification for the intertidal tropical zoanthid Zoanthus sociatus. Marine Pollution Bulletin, 119, 169–175. https://doi.org/10.1016/j. marpolbul.2017.03.054 Lee,R.W., &Childress, J. J.(1994). Assimilationof inorganicnitrogen by marine invertebrates and their chemoautotrophic and metha‐ notrophic symbionts. Applied and Environmental Microbiology, 60, 1852–1858. Lema,K.A.,Willis,B.L.,&Bourne,D.G.(2012).Coralsformcharacter‐ istic associations with symbiotic nitrogen‐fixing bacteria. Applied and Environmental Microbiology, 78,3136–3144.https://doi.org/10.1128/ AEM.07800‐11 Levin,L.A.,&Michener,R.H.(2002).Isotopicevidenceforchemosyn‐ thesis‐based nutrition of macrobenthos: The lightness of being at Pacific methane seeps. Limnology and Oceanography, 47,1336–1345. https://doi.org/10.4319/lo.2002.47.5.1336 Liao,L.,Wankel,S.D.,Wu,M.,Cavanaugh,C.M.,&Girguis,P.R.(2014). Characterizing the plasticity of nitrogen metabolism by the host and symbionts of the hydrothermal vent chemoautotrophic symbioses Ridgeia piscesae. Molecular Ecology, 23, 1544–1557. Little,A.F.,vanOppen,M.J.H.,&Willis,B.L.(2004).Flexibilityinalgal endosymbiosis shapes growth in reef corals. Science, 304, 1492–1494. Lugtenberg, B. (2013). Recognition in microbe‐plant symbiotic and patho‐ genic interactions.Berlin,Germany:SpringerScience&BusinessMedia. MacAvoy,S.,Macko,S.,&Joye,S.(2002).Fattyacidcarbonisotopesig‐ natures in chemosynthetic mussels and tube worms from Gulf of Mexico hydrocarbon seep communities. Chemical Geology, 185, 1–8. https://doi.org/10.1016/S0009‐2541(01)00394‐1 Macko,S.A.,Fogel,M.L.,Hare,P.E.,&Hoering,T.(1987).Isotopicfrac‐ tionation of nitrogen and carbon in the synthesis of amino acids by microorganisms. Chemical Geology: Isotope Geoscience Section, 65, 79–92.https://doi.org/10.1016/0168‐9622(87)90064‐9 Martínez del Rio, C., Wolf, N., Carleton, S. A., & Gannes, L. Z. (2009). Isotopic ecology ten years after a call for more laboratory experiments. Biological Reviews, 84, 91–111. https://doi. org/10.1111/j.1469‐185X.2008.00064.x McClelland, J. W., & Montoya, J. P. (2002). Trophic relationships and the nitrogen isotopic composition of amino acids in plankton. Ecology, 83, 2173–2180. https://doi.org/10.1890/0012‐9658(2002)083[217 3:TRATNI]2.0.CO;2 McCutchan, J. H., Lewis, W. M., Kendall, C., & McGrath, C. C. (2003). Variation in trophic shift for stable isotope ratios of car‐ bon, nitrogen, and sulfur. Oikos, 102, 378–390. https://doi. org/10.1034/j.1600‐0706.2003.12098.x McLeod,R.J.,Wing,S.R.,&Skilton,J.E.(2010).Highincidenceofinver‐ tebrate‐chemoautotroph symbioses in benthic communities of the New Zealand fjords. Limnology and Oceanography, 55, 2097–2106. https://doi.org/10.4319/lo.2010.55.5.2097 McMahon,K.W.,Hamady,L.L.,Thorrold,S.R.J.O.,&Review,M.B. A.A.(2013).Oceanecogeochemistry:Areview.Oceanography and Marine Biology: an Annual Review, 51, 327–374. McMahon, K. W., Thorrold, S. R., Elsdon, T. S., & McCarthy, M. D. J. L. (2015). Trophic discrimination of nitrogen stable isotopes in amino acids varies with diet quality in a marine fish. Limnology and Oceanography, 60,1076–1087.https://doi.org/10.1002/lno.10081 Middelburg,J.(2014).Stableisotopesdissectaquaticfoodwebsfromthe top to the bottom. Biogeosciences, 11,2357.https://doi.org/10.5194/ bg‐11‐2357‐2014 Minagawa, M., & Wada, E. (1984). Stepwise enrichment of 15N along foodchains:Furtherevidenceandtherelationbetweenδ15Nandan‐ imal age. Geochimica (Beijing)Et Cosmochimica Acta, 48, 1135–1140. https://doi.org/10.1016/0016‐7037(84)90204‐7 Mohamed,N.M.,Colman,A.S.,Tal,Y.,&Hill,R.T.(2008).Diversityand expression of nitrogen fixation genes in bacterial symbionts of ma‐ rine sponges. Environmental Microbiology, 10, 2910–2921. https:// doi.org/10.1111/j.1462‐2920.2008.01704.x Morganti, T., Coma, R., Yahel, G., & Ribes, M. (2017). Trophic niche sep‐ aration that facilitates co‐existence of high and low microbial abun‐ dance sponges is revealed by in situ study of carbon and nitrogen fluxes. Limnology and Oceanography, 62, 1963–1983. https://doi. org/10.1002/lno.10546 Motard‐Côté,J.,&Kiene,R.P.(2015).Osmoprotectiveroleofdimethyl‐ sulfoniopropionate(DMSP)forestuarinebacterioplankton.Aquatic Microbial Ecology, 76,133–147.https://doi.org/10.3354/ame01772 Mueller, C.,Larsson,A.,Veuger,B., Middelburg,J.,&VanOevelen,D. (2014).Opportunisticfeedingonvariousorganicfoodsourcesbythe cold‐water coral Lophelia pertusa. Biogeosciences, 11, 123–133. Musat, N., Foster, R., Vagner, T., Adam, B., & Kuypers, M. M. (2012). Detecting metabolic activities in single cells, with emphasis on nanoSIMS. FEMS Microbiology Reviews, 36, 486–511. https://doi. org/10.1111/j.1574‐6976.2011.00303.x Muscatine,L.,Goiran,C.,Land,L.,&Jaubert,J.(2005).Stableisotopes (13Cand15N)oforganicmatrixfromcoralskeleton.Proceedings of the National Academy of Sciences of the United States of America, 102, 1525–1530. Muscatine,L.,McCloskey,L.,&Marian,R.(1981).Estimatingthedaily contribution of carbon from zooxanthellae to coral animal respira‐ tion. Limnology and Oceanography, 26,601–611. Muscatine, L., Porter, J. W., & Kaplan, I. R. (1989). Resource partition‐ ing by reef corals as determined from stable isotope composition. I. δ13C of zooxanthellae and animal tissue vs depth. Marine Biology, 100, 185–193.https://doi.org/10.1007/BF00391957 Nahon, S., Richoux, N. B., Kolasinski, J., Desmalades, M., Pages, C. F., Lecellier,G.,…Lecellier,V.B.(2013).Spatialandtemporalvariations in stable carbon (δ13C) and nitrogen (δ15N)isotopiccompositionof symbiotic scleractinian corals. PLoS ONE, 8, e81247. https://doi. org/10.1371/journal.pone.0081247 Nielsen, J. M., Clare, E. L., Hayden, B., Brett, M. T., & Kratina, P. (2018). Diet tracing in ecology: Method comparison and | 739 FERRIER‐PAGÈS And LEAL selection. Methods in Ecology and Evolution, 9,278–291.https://doi. org/10.1111/2041‐210X.12869 Not,F.,Probert,I.,Ribeiro,C.G.,Crenn,K.,Guillou,L.,Jeanthon,C.,& Vaulot,D.(2016).Photosymbiosisinmarinepelagicenvironments.In The Marine Microbiome(pp.305–332).NewYork,NY:Springer. Oakley,C.A.,Ameismeier,M.F.,Peng,L.,Weis,V.M.,Grossman,A.R., &Davy,S.K.(2016).Symbiosisinduceswidespreadchangesinthe proteome of the model cnidarian Aiptasia. Cellular Microbiology, 18, 1009–1023. Parnell, A.C.,Phillips,D. L.,Bearhop, S.,Semmens,B. X.,Ward,E. J., Moore,J.W.,…Inger,R.(2013).Bayesianstableisotopemixingmod‐ els. Environmetrics, 24,387–399.https://doi.org/10.1002/env.2221 Pernice,M.,Meibom,A.,VanDenHeuvel,A.,Kopp,C.,Domart‐Coulon, I.,Hoegh‐Guldberg,O.,&Dove,S.(2012).Asingle‐cellviewofam‐ monium assimilation in coral–dinoflagellate symbiosis. The ISME Journal, 6,1314.https://doi.org/10.1038/ismej.2011.196 Petersen,J.M.,&Dubilier,N.(2009).Methanotrophicsymbiosesinma‐ rine invertebrates. Environmental Microbiology Reports, 1, 319–335. https://doi.org/10.1111/j.1758‐2229.2009.00081.x Petersen, J. M., Kemper, A., Gruber‐Vodicka, H., Cardini, U., van der Geest,M.,Kleiner,M.,…Seah,B.K.(2017).Chemosyntheticsym‐ bionts of marine invertebrate animals are capable of nitrogen fixation. NatureMicrobiology, 2, 16195. https://doi.org/10.1038/ nmicrobiol.2016.195 Peterson,B.J.,&Fry,B.(1987).Stable isotopes in ecosystem studies. Annual Review of Ecology and Systematics, 18,293–320.https://doi. org/10.1146/annurev.es.18.110187.001453 Phillips, D. L. (2012). Converting isotope values to diet composition: The use of mixing models. Journal of Mammalogy, 93,342–352.https:// doi.org/10.1644/11‐MAMM‐S‐158.1 Phillips,D.L.,Inger,R.,Bearhop,S.,Jackson,A.L.,Moore,J.W.,Parnell, A.C.,…Ward,E.J.(2014).Bestpracticesforuseofstableisotope mixing models in food‐web studies. Canadian Journal of Zoology, 92, 823–835.https://doi.org/10.1139/cjz‐2014‐0127 Piniak,G.,Lipschultz,F.,&McClelland,J.(2003).Assimilationandparti‐ tioning of prey nitrogen within two anthozoans and their endosym‐ biotic zooxanthellae. Marine Ecology Progress Series, 262, 125–136. https://doi.org/10.3354/meps262125 Popp,B.N.,Graham,B.S.,Olson,R.J.,Hannides,C.C.,Lott,M.J.,López‐ Ibarra,G.A.,…Fry,B.(2007).Insightintothetrophicecologyofyel‐ lowfin tuna, Thunnus albacares, from compound‐specific nitrogen isotope analysis of proteinaceous amino acids. Terrestrial Ecology, 1, 173–190. Pruski,A.,Decker,C.,Stetten,E.,Vétion,G.,Martinez,P.,Charlier,K.,… Olu,K.(2017).EnergytransferintheCongodeep‐seafan:Fromter‐ restrially‐derived organic matter to chemosynthetic food webs. Deep Sea Research Part II: Topical Studies in Oceanography, 142, 197–218. https://doi.org/10.1016/j.dsr2.2017.05.011 Rädecker,N.,Pogoreutz,C.,Voolstra,C.R.,Wiedenmann,J.,&Wild,C. (2015). Nitrogencycling incorals: Thekeytounderstanding holo‐ biont functioning? Trends in Microbiology, 23, 490–497. https://doi. org/10.1016/j.tim.2015.03.008 Reynaud,S.,Ferrier‐Pagés,C.,Sambrotto,R.,Juillet‐Lecrec,A.,Jaubert, J.,&Gattuso,J.P.(2002).Effectoffeedingonthecarbonandoxy‐ gen isotopic composition in the tissues and skeleton of the zooxan‐ thellate coral Stylophora pistillata. Marine Ecology Progress Series, 238, 81–89.https://doi.org/10.3354/meps238081 Ribes, M., Dziallas, C., Coma, R., & Riemann, L. (2015). Microbial diver‐ sity and putative diazotrophy in high‐and low‐microbial‐abundance Mediterranean sponges. Applied and Environmental Microbiology, 81, 5683–5693.https://doi.org/10.1128/AEM.01320‐15 Riekenberg, P. M., Carney, R., & Fry, B. (2016). Trophic plasticity of the methanotrophic mussel Bathymodiolus childressi in the Gulf of Mexico. Marine Ecology Progress Series, 547, 91–106. https://doi. org/10.3354/meps11645 Riera,P.,Stal,L.,&Nieuwenhuize,J.(2004).Utilizationoffoodsourcesby invertebrates in a man‐made intertidal ecosystem (Westerschelde, the Netherlands): A δ13C and δ15N study. Journal of the Marine Biological Association of the United Kingdom, 84,323–326. Riou,V.,Halary,S.,Duperron,S.,Bouillon,S.,Elskens,M.,Bettencourt,R., …Colaço,A.(2008).InfluenceofCH4andH2Savailabilityonsymbi‐ ont distribution, carbon assimilation and transfer in the dual symbiotic vent mussel Bathymodiolus azoricus. Biogeosciences, 5,1681–1691. Rix,L.,deGoeij,J.M.,vanOevelen,D.,Struck,U.,Al‐Horani,F.A.,Wild, C.,&Naumann,M.S.J.F.E.(2017).Differentialrecyclingofcoral and algal dissolved organic matter via the sponge loop. Functional Ecology, 31,778–789.https://doi.org/10.1111/1365‐2435.12758 Rix,L.,deGoeij,J.M.,vanOevelen,D.,Struck,U.,Al‐Horani,F.A.,Wild, C.,&Naumann,M.S.J.M.E.P.S.(2018).Reefspongesfacilitatethe transfer of coral‐derived organic matter to their associated fauna via the sponge loop. Marine Ecology Progress Series, 589,85–96.https:// doi.org/10.3354/meps12443 Rix,L.,DeGoeij,J.M.,Mueller,C.E.,Struck,U.,Middelburg,J.J.,Van Duyl,F.C.,…VanOevelen,D.J.S.R.(2016).Coralmucusfuelsthe sponge loop in warm‐and cold‐water coral reef ecosystems. Scientific Reports, 6,18715.https://doi.org/10.1038/srep18715 Robinson,J.J.,&Cavanaugh,C.M.(1995).ExpressionofformIandformII Rubisco in chemoautotrophic symbioses: Implications for the interpre‐ tation of stable carbon isotope values. Limnology and Oceanography, 40,1496–1502.https://doi.org/10.4319/lo.1995.40.8.1496 Robinson,J.J.,Scott,K.M.,Swanson,S.T.,O'Leary,M.H.,Horken,K., Tabita,F.R.,&Cavanaugh,C.M.(2003).Kineticisotopeeffectand characterizationofformIIRubisCOfromthechemoautotrophicen‐ dosymbionts of the hydrothermal vent tubeworm Riftia pachyptila. Limnology and Oceanography, 48, 48–54. Rundel,P.W.,Ehleringer,J.R.,&Nagy,K.A.(2012).Stable isotopes in eco‐ logical research.Berlin,Germany:SpringerScience&BusinessMedia. Sachs,J.L.,Mueller,U.G.,Wilcox,T.P.,&Bull,J.J.(2004).Theevolution of cooperation. The Quarterly Review of Biology, 79,135–160.https:// doi.org/10.1086/383541 Sachs,J.L.,&Wilcox,T.P.(2006).Ashifttoparasitisminthejellyfish symbiont Symbiodinium microadriaticum. Proceedings of the Royal Society B: Biological Sciences, 273,425–429.https://doi.org/10.1098/ rspb.2005.3346 Shantz,A.A.,Lemoine,N.P.,&Burkepile,D.E.(2016).Nutrientloading alters the performance of key nutrient exchange mutualisms. Ecology Letters, 19,20–28.https://doi.org/10.1111/ele.12538 Siegl, A., Kamke, J., Hochmuth, T., Piel, J., Richter, M., Liang, C., … Hentschel,U.J.T.I.j.,(2011).Single‐cellgenomicsrevealsthelifestyle of Poribacteria, a candidate phylum symbiotically associated with marine sponges. The ISME Journal, 5, 61. https://doi.org/10.1038/ ismej.2010.95 Sigman,D.,Karsh,K.,&Casciotti,K.(2009)Oceanprocesstracers:nitro‐ gen isotopes in the ocean. Stat,L.J.(2016).Theeuphoticrealm.InL.Stal,&M.S.Cretoiu(Eds.),The Marine Microbiome(pp.209–225).Berlin,Germany:Springer. Stefels,J.(2000).Physiologicalaspectsoftheproductionandconversion ofDMSPinmarinealgaeandhigherplants.Journal of Sea Research, 43,183–197.https://doi.org/10.1016/S1385‐1101(00)00030‐7 Streit,K.,Bennett,S.A.,VanDover,C.L.,&Coleman,M.(2015).Sources of organic carbon for Rimicaris hybisae: Tracing individual fatty acids at two hydrothermal vent fields in the Mid‐Cayman rise. Deep Sea Research Part I: Oceanographic Research Papers, 100,13–20.https:// doi.org/10.1016/j.dsr.2015.02.003 Sunda, W., Kieber, D., Kiene, R., & Huntsman, S. (2002). An antioxi‐ dantfunctionforDMSPandDMSinmarinealgae.Nature, 418, 317. https://doi.org/10.1038/nature00851 Swart,P.K.(1983).Carbonandoxygenisotopefractionationinsclerac‐ tiniancorals:Areview.Earth Science Reviews, 19,51–80.https://doi. org/10.1016/0012‐8252(83)90076‐4 740 | FERRIER‐PAGÈS And LEAL Swart, P. K., Saied, A., & Lam, K. (2005). Temporal and spatial vari‐ ation in the δ15N and δ13C of coral tissue and zooxanthellae in Montastraea faveolata collectedfromtheFloridareeftract.Limnology Oceanography, 50, 1049–1058. Tanaka,Y.,Miyajima,T.,&Koike,I.(2006).Translocationandconserva‐ tion of organic nitrogen within the coral‐zooxanthella symbiotic sys‐ tem of Acropora pulchra, as demonstrated by dual isotope‐labeling techniques. Journal of Experimental Marine Biology and Ecology, 110– 119.https://doi.org/10.1016/j.jembe.2006.04.011 Taylor, M. W., Radaz, R., Steger, D., & Wagner, M. (2007). Sponge‐as‐ sociated microorganisms: Evolution, ecology, and biotechnological potential. Microbiology and Molecular Biology Reviews, 71, 295–347. https://doi.org/10.1128/MMBR.00040‐06 Teece,M.A.,Estes,B.,Gelsleichter,E.,&Lirman,D.(2011).Heterotrophic and autotrophic assimilation of fatty acids by two Scleractinian corals, Montastraea faveolata and Porites astreoides. Limnology and Oceanography, 56,1285–1296. Teece,M. A.,Fogel, M.L.,Dollhopf,M. E., &Nealson, K.H.(1999). Isotopic fractionation associated with biosynthesis of fatty acids by a marine bacterium under oxic and anoxic conditions. Organic Geochemistry, 30, 1571–1579. https://doi.org/10.1016/ S0146‐6380(99)00108‐4 Thornhill, D., Howells, E., Wham, D., Steury, T., & Santos, S. (2017). Population genetics of reef coral endosymbionts (Symbiodinium, Dinophyceae). Molecular Ecology, 26,2640–2659. Thrall, P. H., Hochberg, M. E., Burdon, J. J., & Bever, J. D. (2007). Coevolution of symbiotic mutualists and parasites in a community context. Trends in Ecology & Evolution, 22, 120–126. https://doi. org/10.1016/j.tree.2006.11.007 Tolosa,I.,Treignier,C.,Grover,R.,&Ferrier‐Pages,C.(2011).Impactof feeding and short‐term temperature stress on the content and iso‐ topic signature of fatty acids, sterols, and alcohols in the scleractin‐ ian coral Turbinaria reniformis. Coral Reefs, 30, 763–774. https://doi. org/10.1007/s00338‐011‐0753‐3 Trask,J.L.,&VanDover,C.L.(1999).Site‐specificandontogeneticvari‐ ations in nutrition of mussels (Bathymodiolus sp.) from the Lucky Strike hydrothermal vent field. Mid‐Atlantic Ridge. Limnology and Oceanography, 44, 334–343. Treignier, C., Tolosa, I., Grover, R., Reynaud, S., & Ferrier‐Pagès, C. (2009). Carbon isotope composition of fatty acids and sterols in the scleractinian coral Turbinaria reniformis: Effect of light and feeding. Limnology and Oceanography, 54, 1933–1940. https://doi. org/10.4319/lo.2009.54.6.1933 Tremblay,P.,Gori,A.,Maguer,J.F.,Hoogenboom,M.,&Ferrier‐Pagès,C. (2016). Heterotrophypromotesthere‐establishmentofphotosyn‐ thate translocation in a symbiotic coral after heat stress. Scientific Reports, 6,38112.https://doi.org/10.1038/srep38112 Tremblay,P.,Grover,R.,Maguer,J.F.,Hoogenboom,M.,&Ferrier‐Pagès, C. (2014). Carbon translocation from symbiont to host depends on ir‐ radiance and food availability in the tropical coral Stylophora pistillata. Coral Reefs, 33,1–13.https://doi.org/10.1007/s00338‐013‐1100‐7 Tremblay,P.,Grover,R.,Maguer,J.F.,Legendre,L.,&Ferrier‐Pages,C. (2012).Autotrophiccarbonbudgetincoraltissue:Anewδ13C‐based model of photosynthate translocation. The Journal of Experimental Biology, 215, 1384–1393. Tremblay,P.,Maguer,J.F.,Grover,R.,&Ferrier‐Pagès,C.(2015).Trophic dynamics of scleractinian corals: Stable isotope evidence. Journal of Experimental Biology, 218, 1223–1234. https://doi.org/10.1242/ jeb.115303 Tremblay, P.,Naumann, M. S., Sikorski, S.,Grover,R.,& Ferrier‐Pagès, C.(2012).Experimentalassessmentoforganiccarbonfluxesinthe scleractinian coral Stylophora pistillata during a thermal and photo stress event. Marine Ecology Progress Series, 453,63–77. van Rhijn, P., & Vanderleyden, J. (1995). The Rhizobium‐plant symbiosis. Microbiological Reviews, 59, 124–142. Venn,A.,Loram,J.,&Douglas,A.(2008).Photosyntheticsymbiosesin animals. Journal of Experimental Botany, 59,1069–1080.https://doi. org/10.1093/jxb/erm328 Villinski,J. C.,Hayes, J. M., Villinski,J.T.,Brassell,S.C.,& Raff,R. A. (2004). Carbon‐isotopic shifts associated with heterotrophy and biosynthetic pathways in direct‐and indirect‐developing sea ur‐ chins. Marine Ecology Progress Series, 275, 139–151. https://doi. org/10.3354/meps275139 Ward‐Paige,C.A.,Risk,M.J.,&Sherwood,O.A.(2005).Reconstruction of nitrogen sources on coral reefs: δ15Nandδ13C in gorgonians from FloridaReefTract.Marine Ecology Progress Series, 296,155–163. Webster,N.S.,&Taylor,M.W.(2012).Marinespongesandtheirmicrobial symbionts: Love and other relationships. Environmental Microbiology, 14,335–346.https://doi.org/10.1111/j.1462‐2920.2011.02460.x Wilkinson,C.R.,Nowak,M.,Austin,B.,&Colwell,R.R.J.M.E.(1981 SpecificityofbacterialsymbiontsinMediterraneanandGreatBarrier Reef sponges. Microbial Ecology, 7, 13–21.https://doi.org/10.1007/ BF02010474 Williams,D.F.,Röttger,R.,Schmaljohann,R.,&Keigwin,L.J.P.(1981). Oxygenandcarbonisotopicfractionationandalgalsymbiosisinthe benthic foraminiferan heterostegina depressa. Palaeogeography, Palaeoclimatology, Palaeoecology, 33, 231–251. https://doi. org/10.1016/0031‐0182(81)90041‐9 Wolf,N.,Carleton,S.A.,&MartínezdelRio,C.(2009).Tenyearsofex‐ perimental animal isotopic ecology. Functional Ecology, 23, 17–26. https://doi.org/10.1111/j.1365‐2435.2009.01529.x Wong, C. M., Duprey, N. N., & Baker, D. M. (2017). New insights on the nitrogen footprint of a coastal megalopolis from coral‐hosted Symbiodinium δ15N. Environmental Science & Technology, 51, 1981–1987. York,J.K.,Tomasky,G.,Valiela,I.,&Repeta,D.J.(2007).Stableisotopic detection of ammonium and nitrate assimilation by phytoplankton in the Waquoit Bay estuarine system. Limnology and Oceanography, 52, 144–155.https://doi.org/10.4319/lo.2007.52.1.0144 Zanden, M., & Rasmussen, J. B. (1999). Primary consumer δ13C and δ15N and the trophic position of aquatic consumers. Ecology, 80, 1395–1404. Zanden,M.J.V.,&Rasmussen,J.B.(2001).Variationinδ15Nandδ13C trophic fractionation: Implications for aquatic food web studies. Limnology and Oceanography, 46,2061–2066. How to cite this article:Ferrier‐PagèsC,LealMC.Stable isotopes as tracers of trophic interactions in marine mutualistic symbioses. Ecol Evol. 2019;9:723–740. https://doi. org/10.1002/ece3.4712