Metabarcoding, Direct Stomach Observation and Stable Isotope

bioRxiv preprint doi: https://doi.org/10.1101/2020.08.13.249896; this version posted May 28, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

1 1 Title: 2 3 2 Metabarcoding, direct stomach observation and stable isotope analysis reveal a highly 4 5 6 3 diverse diet for the invasive green crab in Atlantic Patagonia 7 8 4 9 10 11 5 Authors-- 12 13 6 Georgina Cordone1, Mariana Lozada 2, Elisabet Vilacoba 3, Bettina Thalinger 4,5, Gregorio Bigatti 6,7, 14 15 3 4,5 1 16 7 Darío A. Lijtmaer , Dirk Steinke , David E. Galván * 17 18 8 19 20 9 * Corresponding author: [email protected] ; Tel: +54 (280) 488-3184 Ext. 1277 - Fax: 21 22 10 +54 (280) 488-3543 23 11 24 25 12 26 27 1 28 13 Centro para el Estudio de Sistemas Marinos (CESIMAR), Consejo Nacional de Investigaciones 29 30 14 Científicas y Técnicas (CONICET), Edificio CCT CONICET-CENPAT, Bvd. Brown 2915, 31 32 33 15 (U9120) Puerto Madryn, Chubut, Argentina. 34 35 16 2 Laboratorio de Microbiología Ambiental (IBIOMAR), Consejo Nacional de Investigaciones 36 37 38 17 Científicas y Técnicas (CONICET), Edificio CCT CONICET-CENPAT, Bvd. Brown 2915, 39 40 18 (U9120) Puerto Madryn, Chubut, Argentina. 41 42 19 3 División Ornitología, Museo Argentino de Ciencias Naturales “Bernardino Rivadavia” MACN- 43 44 45 20 CONICET. Av. Ángel Gallardo 470 (C1405) Ciudad Autónoma de Buenos Aires, Buenos Aires, 46 47 21 Argentina. 48 49 4 50 22 Centre for Biodiversity Genomics,University of Guelph, 50 Stone Rd E (N1G2W1), Guelph, 51 52 23 Ontario, Canada. 53 54 5 55 24 Department of Integrative Biology, College of Biological Science, University of Guelph, 50 Stone 56 57 25 Rd E (N1G 2W1), Guelph, Ontario, Canada. 58 59 26 6 Laboratorio de Reproducción y Biología Integrativa de Invertebrados Marinos, (LARBIM, 60 61 62 27 IBIOMAR). Consejo Nacional de Investigaciones Científicas y Técnicas (CONICET), Edificio 63 64 65 bioRxiv preprint doi: https://doi.org/10.1101/2020.08.13.249896; this version posted May 28, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

1 28 CCT CONICET-CENPAT, Bvd. Brown 2915, (U9120) Puerto Madryn, Chubut, Argentina. 2 3 29 7 Universidad Espíritu Santo, Ecuador 4 5 6 7 30 Declarations 8 9 10 31 Acknowledgements 11 12 13 14 32 We would like to thank Florencia Ríos, Manuela Funes, Juan Pablo Livore, and Maria 15 16 33 Martha Mendez for their fieldwork support. We would like to thank Agustina Ferrando, Cynthia 17 18 19 34 Ibarra, Pablo Macchi and, German Cheli for expert assessment on taxonomic identifications. We are 20 21 35 grateful to two anonymous reviewers for their constructive feedback on the original manuscript 22 23 24 36 version. 25 26 27 37 Funding 28 29 30 31 38 This research was supported by Consejo Nacional de Investigaciones Científicas y Técnicas 32 33 39 (CONICET, Argentina) and Sistema Nacional de Datos Genómicos (SNDG, Argentina). Fieldwork 34 35 36 40 campaigns and stable isotope analysis were funded by Fondo para la Investigación Científica y 37 38 41 Tecnológica (FonCyT) under the grants PICT 2017-2403, PICT 2018-0903 and PICT 2018-0969. 39 40 42 The metabarcoding analysis was funded by the Natural Sciences and Engineering Research Council 41 42 43 43 of Canada and the Canada First Research Excellence Fund. 44 45 46 44 Conflicts of interest 47 48 49 50 45 The authors declare that they have no conflict of interest. 51 52 53 46 Ethical approval 54 55 56 57 47 All applicable institutional and/or national guidelines for the care and use of animals were 58 59 48 followed. Permissions for sample collection were granted from the Secretaría de Pesca and 60 61 62 49 Secretaría de Conservación de Áreas Protegidas from Chubut Province. Exportation permissions 63 64 50 were granted in 2019 by the Dirección Nacional de Biodiversidad (Secretaría de Ambiente y 65

1 51 Desarrollo Sustentable) under ANEXO II, resolution number: IF-2019-40388897-APN- 2 3 52 DNBI#SGP. 4 5 6 7 53 Consent to participate 8 9 10 54 Not applicable 11 12 13 14 55 Consent for publication 15 16 17 56 Not applicable 18 19 20 21 57 Availability of data and material 22 23 24 58 All data generated and analysed during this study are included in Electronic Supplementary 25 26 27 59 Material: Online Resource 1, 2, 3, 4, and 5 (ESM 1, ESM 2, ESM 3, ESM 4, ESM 5, and ESM 6). 28 29 60 The sequence raw data has been uploaded to Genbank SRA (accession: SRR13517441). 30 31 32 33 61 Code availability 34 35 36 62 The code was implemented in R and it is available under request (R Core Team, 2020). 37 38 39 40 63 Author Contributions 41 42 43 64 DEG, GB, ML, and GC conceived and designed this study. DEG, GB, and GC took the 44 45 46 65 samples. GC conducted the visual analysis and prepared the samples for stable isotope analysis. GC 47 48 66 and DEG analyzed direct stomach observation and stable isotope data. ML contributed to preparing 49 50 51 67 the metabarcoding samples and analyzed metabarcoding data. EV visited as an invited researcher 52 53 68 the Centre for Biodiversity Genomics to obtain the metabarcoding data. EV, BT, DAL, and DS 54 55 69 contributed to the generation of metabarcoding data, molecular and bioinformatic analyses. All 56 57 58 70 authors contributed to manuscript writing. 59 60 61 62 63 64 65

1 71 Abstract 2 3 72 The European green crab Carcinus maenas and its sister species C. aestuarii are highly 4 5 6 73 invasive species causing damage to coastal ecosystems and severe economic losses worldwide. C. 7 8 74 maenas was first detected at the Atlantic Patagonian coast in 2001. In this work, we studied the diet 9 10 11 75 of the green crab in a recently invaded location in Golfo Nuevo, using three complementary 12 13 76 techniques: direct stomach observation, metabarcoding of the gut content and stable isotope 14 15 16 77 analysis. Direct stomach observation and metabarcoding showed that green crabs have a broad 17 18 78 omnivorous diet, ingesting most of the phyla present in the study area. Gut content metabarcoding 19 20 79 allowed a detailed description of algae diversity and revealed other taxa that went unnoticed in the 21 22 23 80 visual stomach analysis. Stable isotope analysis showed that the major contribution to crabs' diet 24 25 81 was from the phytoplankton chain (by bivalve consumption) and not directly from algae. This study 26 27 28 82 approach combining three complementary techniques also allowed us to detect some differences in 29 30 83 the diet between sexes, which suggests that male and female crabs are not as ecologically equivalent 31 32 33 84 as previously thought. Besides, we detected sequences corresponding to C. aestuarii suggesting that 34 35 85 the green crab Patagonian population is a hybrid of both sister species. These findings are key to 36 37 38 86 understanding the impacts green crabs could trigger on the local ecosystem. 39 40 87 41 42 88 Keywords: Carcinus maenas, Carcinus aestuarii, e-DNA, Rocky shore ecology, Trophic 43 44 45 89 interactions 46 47 90 48 49 50 91 1. Introduction 51 52 53 92 Invasive species, particularly invasive predators are one of the main causes of worldwide 54 55 56 93 species decline and biodiversity loss (Clavero and García-Berthou 2005; Doherty et al. 2016). Once 57 58 94 invasive species arrive at a new location, they establish new ecological interactions with resident 59 60 61 95 species and these new interactions can have positive or negative effects on native species or the 62 63 96 entire native food web (David et al. 2017). The ways biological invasions impact food webs are 64 65

1 97 complex and context-dependent as there are various direct and indirect mechanisms by which an 2 3 98 invasive species can affect other taxa (Thomsen et al. 2011a, b). Preying upon a native species is a 4 5 6 99 direct mechanism that usually ends up with the native species population decreasing in size, or even 7 8100 becoming extirpated (Bruno et al. 2005; David et al. 2017). Exploitation and apparent competition 9 10 11101 are indirect mechanisms that can also cause a reduction of native population sizes (White et al. 12 13102 2006; David et al. 2017). On the other hand invaders can trigger positive effects for resident 14 15 16103 species. In fact, for species at higher trophic positions, invaders can increase food availability and 17 18104 create new habitats (Rodriguez 2006; Thomsen et al. 2010; Thomsen et al. 2014). The invaders' 19 20105 impact depends on the food web’s invasibility which is ultimately related to network structure (Hui 21 22 23106 et al. 2016). Earlier work has shown that food webs with higher biodiversity, mainly understood as 24 25107 higher species richness, are more resistant to invasions; however, on the flip side highly diverse 26 27 28108 communities generally support more invasive species (Fridley et al. 2007; Tilman et al. 2014; David 29 30109 et al. 2017). By using simulations, Romanuk et al. (2017) have shown that trophic cascades are 31 32 33110 more frequent when invaders become secondary consumers with abundant prey and only few other 34 35111 predators. Consequently, if we want to accurately predict possible impacts of a newly arrived 36 37 38112 invasive species, we need to understand its trophic interactions in the invaded ecosystem. 39 40113 41 42114 Determining the diet of a particular species is a key requirement to understanding its trophic 43 44 45115 ecology, but it is far from being a simple achievement. There are a series of conceptual and 46 47116 methodological aspects that must be taken into account (e.g Traugott et al. 2013). It is not the same 48 49 50117 to document which organisms an animal ingests as to list which organisms are targets of a predator 51 52118 or a scavenger. Similarly, the ingestion of a certain item does not assure its digestion and 53 54 55119 assimilation. A variety of direct and indirect methods are currently used to analyze diet composition 56 57120 (Majdi et al. 2018; Nielsen et al. 2018). Methods include visual observation of feeding events and 58 59121 regurgitations, the visual inspection of faeces, pellets or stomach contents, and the molecular 60 61 62122 profiling of stomach contents or faeces using DNA sequencing or immunoassays. Each of these 63 64 65

1123 techniques has its advantages and disadvantages, which depend on the characteristics of the study, 2 3124 the geographic area under analysis and the taxa that are being targeted. DNA-based techniques 4 5 6125 allow the characterization of feeding interactions at high taxonomic resolution and are very useful 7 8126 when visual recognition is not possible, and in cases of omnivorous species, where a mixture of 9 10 11127 DNA from different species is present in gut content, the most effective approach is metabarcoding 12 13128 (Taberlet et al. 2012; Pompanon et al. 2012; Clare 2014; Nielsen et al. 2018; Unger et al. 2020). 14 15 16129 Metabarcoding couples DNA barcoding (Hebert et al. 2003) with high-throughput sequencing of 17 18130 community samples (i.e. gut or fecal samples in the case of diet studies) and as a result it normally 19 20131 emerges as the superior technique for prey identification (Nielsen et al. 2018). In addition, it not 21 22 23132 only provides information on ingested organisms, but it also has the advantage of providing 24 25133 complementary information, such as the gastrointestinal parasites of a predator. It should be born in 26 27 28134 mind, however, that metabarcoding success depends on various factors, including the availability of 29 30135 a relatively comprehensive database of the targeted organisms (i.e. the prey items in the case of diet 31 32 33136 studies), the right choice of markers (to avoid low taxonomic resolution) and primers, and the 34 35137 possibility to deal with putative contamination and sequencing or bioinformatics errors (Braukmann 36 37 38138 et al. 2019). Moreover, it has two relevant limitations when used as the sole technique to study diet: 39 40139 it cannot detect cannibalism and it does not provide information on the development stage of the 41 42140 prey (i.e. whether eggs, larvae or adults have been ingested; Nielsen et al. 2018; Traugott et al. 43 44 45141 2021). 46 47 48142 Direct visual observation of stomach content is a traditional technique to identify diet 49 50 51143 composition; but is not recommended to use it as the only source of information, mainly because 52 53144 the digestion of prey material can severely limit detection and introduces biases towards certain 54 55 56145 taxonomic groups, such as shell-bearing organisms or arthropods that possess exoskeletal 57 58146 structures. Moreover, identifications largely depend on the skills of the taxonomists involved (Majdi 59 60 61147 et al. 2018; Traugott et al. 2021). However, direct observation can provide indispensable first-hand 62 63148 information, allows quantifying the ingested amount, and in certain cases distinguishes between 64 65

1149 secondary predation and random ingestion (Hyslop 1980; Arnett and Whelan 2001; Zacharia et al. 2 3150 2004). In addition, it can guide the DNA-based analysis and help solving some of the 4 5 6151 metabarcoding shortcomings mentioned above (such as the identification of the development stage 7 8152 of the prey or the presence of cannibalism). As a result, the combination of visual inspection and 9 10 11153 DNA-based techniques can be considered a wise choice to describe diet composition (e.g., 12 13154 Thalinger et al. 2018). 14 15 16 17155 Aside from such direct methods, indirect tracers of diet composition are robust and 18 19156 commonly used tools. They include the characterization of biomarkers such as fatty acids, sterols or 20 21 22157 pigments present in consumer tissues and the determination of the isotopic compositions of C, N 23 24158 and S in specific tissues (e.g. Galloway et al. 2015; Stock et al. 2018). Stable isotope analysis (SIA) 25 26159 is based on predictable differences between the isotopic signature of an organism and its food 27 28 29160 resources. It allows for the quantitative estimation of fluxes in food webs, the determination of a 30 31161 species’ trophic level and insights into the patterns of resource allocation (Boecklen et al. 2011). In 32 33 34162 spite of these advantages, a potential confounding factor of SIA is the variability of trophic 35 36163 discrimination factors (TDF) and turnover rates, which may depend on environments, diet, taxon 37 38 39164 and tissue (Pinnegar and Polunin 1999; Philippsen and Benedito 2013; Hussey et al. 2014; Lefebvre 40 41165 and Dubois 2016). Similarly, isotope compositions are highly variable through space and time 42 43166 (Hyndes et al. 2013; Mackey et al. 2015). Therefore, these factors should be carefully considered 44 45 46167 when stable isotopes are used in diet studies. 47 48 49168 These direct and indirect approaches are largely complementary as the former can identify 50 51 52169 the prey which is ingested, while the latter provide information on what is assimilated. Furthermore, 53 54170 indirect methods such as SIA provide information at longer time scales while DNA or visual 55 56 57171 analyses register consumption within short time periods (typically 6-48h) (Albaina et al. 2010; 58 59172 Hayden et al. 2014). In fact, recent trophic ecology studies successfully combined SIA and 60 61 62173 metabarcoding (Compson et al. 2019; Whitaker et al. 2019) and in this way started to fill a gap that 63 64174 has been previously highlighted for the field (Majdi et al. 2018). 65

1175 2 3176 The European green crab Carcinus maenas (Linnaeus, 1758) is an invasive species native to 4 5 6177 the North Atlantic Coast. To this day it has reached a worldwide distribution (Carlton and Cohen 7 8178 2003). Not only C. maenas invaded new locations, but its sister species C. aestuarii, or hybrids of 9 10 11179 both, have also been detected at invaded sites (Darling 2011). These crabs are considered aggressive 12 13180 feeders due to their voracious predation on other invertebrates (McDonald et al. 2001). It has been 14 15 16181 shown that portunoid crabs are scavengers and predators, and that the green crab is an active 17 18182 carnivore feeding especially on mollusca and crustacea (Ebling et al. 1964; Elner 1981; Grosholz 19 20183 and Ruiz 1995; McDonald et al. 2001; Chen et al. 2004). Female and male green crabs are 21 22 23184 considered ecological equivalents as they do not show differences in their diets nor in their foraging 24 25185 behaviour (Spooner et al. 2007; Young and Elliot 2020). The green crab was first detected at the 26 27 28186 Patagonian coast in 2001, and twelve years later it was registered in the Golfo Nuevo, 250 km north 29 30187 of the point of the first observation (Torres and González-Pisani 2016). The coast of Argentina 31 32 33188 currently bears a total of 129 introduced and 72 cryptogenic marine species, and it has been 34 35189 estimated that every 178 days a new invasion is detected (Schwindt et al. 2020). The interactions 36 37 38190 between the green crab and native as well as introduced species could have unpredictable 39 40191 consequences for local trophic structure, as well as for ecosystem functioning. Very little is known 41 42192 about green crab ecology at the Argentine coasts, in particular with regards to its local trophic 43 44 45193 interactions (Hidalgo et al. 2007; Young and Elliott 2020). Moreover, the observation that feeding 46 47194 behaviour is more aggressive in invasive than in native populations (Howard et al. 2018), leads to 48 49 50195 further uncertainties on how this predator could affect the invaded ecosystem. 51 52196 53 54 55197 In this study we aim to shed a light on the diet of the green crab in Atlantic Patagonian 56 57198 coastal environments by combining three techniques that have proven useful for diet tracing: visual 58 59199 stomach content inspection, stable isotope analysis, and amplicon sequencing of a gene fragment 60 61 62200 from stomach content DNA (metabarcoding). We address the following questions to understand 63 64 65

1201 possible impacts on the trophic structure of the affected coastal communities in this coastal 2 3202 environment: What are green crabs eating in the coastal environments of Atlantic Patagonia? Are 4 5 6203 there differences in the diets of females and males in Atlantic Patagonia? Does the use of multiple 7 8204 techniques contribute to a more complete picture of the green crab’s diet? 9 10 11 12205 2. Materials and Methods 13 14 15206 2.1 Sample collection 16 17 18 19207 Green crabs were collected at Punta Este beach in the Golfo Nuevo, Patagonia, Argentina 20 21208 (42°46 ′ S, 64°57 ′ W). Sampling was done monthly from November 2018 to February 2019 (austral 22 23 24209 spring and summer) in shallow waters with depths of about 2-3 m at high tide. Crabs were captured 25 26210 manually by SCUBA diving and frozen at -20°C upon arrival to the laboratory both to kill them and 27 28 29211 correctly preserve them for subsequent analyses. A total of 223 specimens were captured. 30 31212 Individuals were separated into two groups: one for traditional visual analysis (n (vis) = 183) and the 32 33213 other for metabarcoding (n = 40). Some of the individuals of each of these two groups (n = 34 (met) (sia) 35 36214 26 in total) were used for the SIA. The samples for the three techniques correspond to the same time 37 38215 period. Crab sizes were registered by measuring carapace width (CW) in cm. Subsequently, crabs 39 40 41216 were sexed and dissected. For specimens destined for visual analysis, stomach content was 42 43217 extracted and kept in 70% ethanol. For specimens destined for metabarcoding, stomach content was 44 45 46218 manually collected from frozen samples by letting them thaw and carefully extracting contents with 47 48219 a plastic pipette tip through the mouth of the crab. Contents were placed in 100% ethanol (Merck) 49 50220 and stored at -80°C until DNA extraction. Muscle from crab chela was extracted and frozen for 51 52 53221 stable isotope analysis. In addition, five individuals of the mussel Mytilus edulis and five of the 54 55222 herbivorous snail Tegula patagonica were collected to establish isotopic baselines. This was done 56 57 58223 under the assumption that both the mussel and the herbivorous snail are primary consumers feeding 59 60224 mostly on micro- and macroalgae, respectively. We selected these two primary consumers as they 61 62 63225 integrate possible variations of environment isotopic values and their tissue usually shows shorter 64 65

1226 turnover rates than secondary consumers. As a consequence, primary consumers provide the most 2 3227 appropriate isotopic signatures for distinguishing between pelagic and benthic pathways in aquatic 4 5 6228 ecosystems (Post 2002). 7 8 9229 2.2 Visual analysis of gut contents 10 11 12 13230 The visual analysis of gut contents was done using a stereomicroscope (Zeiss Stemi, model 14 15231 2000c). Prey was identified to the lowest possible taxonomic level following catalogues and keys of 16 17 18232 the local fauna (Fauchald 1977; Gosztonyi and Kuba 1996; Boschi et al. 1992; Spivak et al. 2019), 19 20233 as well as through expert consultation and by comparisons with fresh material. Presence of prey 21 22 23234 items in each stomach was recorded and the frequency of occurrence (FO (vis)%) calculated per prey 24 25235 item as the percentage of non-empty stomachs in which the item was recorded. For statistical 26 27236 analysis, we grouped prey into major groups (Annelida, Arthropoda, Mollusca, Chordata, 28 29 30237 Echinodermata and Algae). We used a bootstrap approach to estimate the FO confidence intervals 31 32238 for these groups. Bootstrap procedures are useful for non-parametric estimates and have been used 33 34 35239 in several studies of trophic ecology (Tirasin and Jorgensen 1999). This approach consists of re- 36 37240 sampling with replacement of the original data matrix and generates a new set of simulated 38 39 40241 matrices. The original matrix included green crab stomachs per row and prey items (presence or 41 42242 absences) in columns. Simulated matrices were generated by randomly choosing rows from the 43 44 45243 original matrix. This procedure was repeated until the simulated matrix reached the same number of 46 47244 rows as the original. Green crab stomachs could be chosen more than once per simulated matrix as 48 49245 resampling allowed for replacement. We simulated 1000 new matrices using the functions 50 51 52246 implemented in the R package ‘boot’ and calculated the FO for each prey item (Canty and Ripley 53 54247 2021). This is based on the assumption that the frequency distribution of the simulated FO is 55 56 57248 unbiased and fully describes the distribution function of the parameter of interest (Efron and 58 59249 Tibshirani 1993; Mooney and Duval 1993). 60 61 62 63250 2.3 Metabarcoding analysis of gut content 64 65

1251 DNA extraction and COI amplification 2 3 4252 All molecular analyses were carried out under clean-room conditions at the Centre for 5 6 7253 Biodiversity Genomics (University of Guelph, Canada). First, the ethanol of the gut content sample 8 9254 was evaporated in a fume hood overnight. Remaining pellets were lysed using 180µl ATL buffer + 10 11 12255 20µl Proteinase K (20 mg/ml) per sample and incubated on a rocking platform at 56°C overnight. 13 14256 Afterwards, the DNA was extracted using the DNeasy Blood and Tissue Kit (Qiagen, Hilden, 15 16 17257 Germany), following manufacturers’ instructions except for an additional centrifugation step 18 19258 (14000 rpm) for removing the buffer AW2 from the silica membrane and for elution in two steps 20 21 22259 using 100 µl of Buffer AE per each. A fume hood control and extraction control were processed 23 24260 along the stomach content samples to check for cross-contamination. DNA extracts were quantified 25 26261 using the Qubit dsDNA HS Assay Kit and a Qubit 3.0 Fluorometer (Thermo Fisher Scientific, MA, 27 28 29262 USA). For the amplification of gut content DNA, we used a two-step PCR protocol following 30 31263 Elbrecht et al. (2019). PCRs were carried out using the Multiplex PCR Master Mix Plus (Qiagen, 32 33 34264 Hilden, Germany), 0.5 mM of each primer (IDT, Skokie, Illinois), and molecular grade water 35 36265 (HyPure, GE, Utha, USA) for a total volume of 25 ml. For the first PCR step, amplification of 37 38 39266 cytochrome oxidase I gene was performed using primers BF3/BR2 (5’ 40 41267 CCHGAYATRGCHTTYCCHCG 3’ / 5’ TCDGGRTGNCCRAARAAYCA 3’) and 4 µl of DNA 42 43268 extract (BR2: Elbrecht and Leese 2017, BF3: Elbrecht et al. 2019; see Electronic supplementary 44 45 46269 material: Online Resource ESM 1). For the second PCR step, 1 µl of the product obtained from the 47 48270 first PCR was used for amplification using fusion primers with individual tags per sample (see 49 50 51271 Online Resource ESM 2, section 2.3) (Elbrecht and Leese 2017; Elbrecht et al. 2019; Elbrecht and 52 53272 Steinke 2019). Thermocycling conditions for both PCR steps included an initial denaturation at 54 55 56273 95°C for 5 minutes, 24 cycles of denaturation at 95°C for 30 seconds, annealing at 50°C for 30 57 58274 seconds, and extension at 72°C for 50 seconds; and a final extension at 68°C for 10 minutes. After 59 60 61275 the second PCR, amplification success and band strength were checked on 1.5% agarose gels. For 62 63276 several samples which showed no or only very weak amplification, both PCRs were repeated, but 64 65

1277 with 35 cycles instead of 24 in the first PCR. These replicates received individual tag combinations 2 3278 and were prepared for sequencing on an Illumina MiSeq platform along with the extraction controls 4 5 6279 and seven PCR negative controls. 7 8 9280 Normalization of the second PCR products was performed using the SequalPrep 10 11 12281 Normalization Plate Kit (Invitrogen, CA, USA). Samples were subsequently pooled into one tube 13 14282 and cleaned using SPRIselect following the left side size protocol with a sample-to-volume ratio of 15 16 17283 0.76X (Beckman Coulter,CA, USA). Library quantification was performed using the Qubit dsDNA 18 19284 HS Assay Kit and a Qubit 3.0 Fluorometer (Thermo Fisher Scientific, MA, USA). Sequencing was 20 21 22285 done at the University of Guelph (Advanced Analysis Centre) on an Illumina MiSeq using the 600 23 24286 cycle Illumina MiSeq Reagent Kit v3 (2 × 300) and 5% PhiX spike in. Sequencing results were 25 26287 uploaded to the NCBI Sequence Read Archive (SRA, Genbank, accession: SRR13517441). 27 28 29 30288 Sequence processing 31 32 33289 Sequences were preprocessed using JAMP (https://github.com/VascoElbrecht/JAMP). 34 35 36290 Briefly, sequences were demultiplexed, and forward and reverse paired-end reads were merged. 37 38291 Primer sequences were removed, and sequences between 380bp and 440 bp (+/- 30bp of expected 39 40 41292 fragment length) were kept in the dataset, in order to preserve the most of the possible diversity 42 43293 present in the sample. 44 45 46 47294 The resulting files were processed in two ways: i) the reads were quality filtered and 48 49295 taxonomy was assigned using the Multiplex Barcode Research And Visualization Environment in 50 51 52296 combination with the Barcode of Life platform (mBRAVE/BOLD; Ratnasingham 2019; 53 54297 Ratnasingham and Hebert 2007), and ii) quality filtering was carried out in JAMP and the resulting 55 56298 haplotypes were grouped into OTUs and blasted against the NCBI nucleotide database (for details 57 58 59299 see Online Resource ESM 2). 60 61 62300 Description of crab diet based on molecular data 63 64 65

1301 Taxonomic assignments obtained by the two pipelines (mBRAVE/BOLD-based dataset and 2 3302 JAMP/NCBI-based dataset) were manually curated. First, genus level taxonomic tables were 4 5 6303 screened for locally occurring species and, in the case of the mBRAVE dataset, the quality of the 7 8304 BOLD references (BINs) was determined to define the most reliable taxonomic level. For instance, 9 10 11305 genera which are not described for Argentina, not known from the marine environment, and/or 12 13306 whose BIN included multiple genera or could be considered unreliable (e.g. composed by only one 14 15 16307 sequence) were considered as “unidentified” members of that family/class. The NCBI dataset was 17 18308 curated at the OTU level, based on % similarity of the first hit to OTU representative sequences. 19 20309 Only hits with pairwise identity >85% and alignment coverage >75% were kept for analysis. Of 21 22 23310 these, representative sequences were assigned to: species (at least 98% similarity with the first hit), 24 25311 genus (95%), family (90%), and order (85%) (Elbrecht et al. 2017). For analysis of taxa occurrence, 26 27 28312 both analyses were combined into one that contained BOLD-based assignments plus NCBI 29 30313 assignments of taxa not detected in mBRAVE. The results were used to calculate the frequency of 31 32 33314 occurrence (FO (met)) at different taxonomic levels in the same way as FO (vis) (see section 2.2). We 34 35315 decided to use occurrence data and not quantitative data such as read count because gut content 36 37 38316 material is usually at various states of digestion and the sequence counts are less likely to contain a 39 40317 meaningful quantitative signal compared to e.g. fecal matter (Nakahara et al. 2015; Deagle et al. 41 42318 2019). In this way, we avoid incorporating possible biases due to the stochasticity of the 43 44 45319 amplification success and the variability in digestion times. Complete tables of taxonomic 46 47320 assignments are provided in Online Resources (ESM 3: Table S2 and ESM 4: Table S3). 48 49 50 51321 2.4 Statistical analysis of diet description 52 53 54322 To test if sample size was sufficient to obtain a representative description of the diet, we 55 56 57323 constructed cumulative trophic diversity curves using the quadrats methods following Koen Alonso 58 59324 et al. (2002) for both visual and metabarcoding datasets. We calculated trophic diversity as the 60 61 62325 Brillouin index (Hz) and considered diversity curves as asymptotic if the two previous values (i.e. 63 64326 n-1 and n-2) were in the range Hz ± 0.05 Hz for phylum and genus classification (Pielou 1966, 65

1327 Koen Alonso et al. 2002). Since this condition was met the number of stomachs in our study, n (vis) 2 3328 and n was considered sufficient to describe the green crab diet. 4 (met) 5 6 7329 To test differences between male and female green crab diet, we performed a distance-based 8 9330 multivariate Welch t-test (Wd ∗ -test) both for Class-level assigned metabarcoding and visual 10 11 12331 analyses. The Wd∗ -test is a robust alternative to PERMANOVA when heteroscedasticity or 13 14332 imbalanced data are present (Alekseyenko et al. 2016; Hamidi et al. 2019). The tests were 15 16 17333 performed using a Bray-Curtis similarity matrix implemented in vegan and the MicEco R package 18 19334 (Oksanen et al. 2019; Russel 2020). 20 21 22 23335 Alpha and beta diversity metrics, ordinations, and other statistical analyses were calculated 24 25336 using OTUs resulting from the JAMP routine. Analyses were performed using the R packages 26 27337 vegan and phyloseq (McMurdie and Holmes 2013; Oksanen et al. 2019). Briefly, OTU richness was 28 29 30338 calculated using the rarefy function in vegan, which gives the expected species richness in random 31 32339 community subsamples of a given size. A Kruskal-Wallis non-parametric test was used to analyze 33 34 35340 whether there were differences in prey richness between sexes. For the beta diversity analyses, a 36 37341 similarity matrix based on OTUs was built using Jaccard similarity index, from which NMDS 38 39 40342 ordinations were constructed. 41 42 43343 2.5 Stable isotope analysis 44 45 46 47344 Muscle samples for SIA were extracted from the chela of the frozen specimens and placed 48 49345 back in the freezer in individual tubes. We made sure that samples remained frozen during the 50 51 52346 entire procedure by using ice and sampling the crabs in small groups (4 individuals), as repeated 53 54347 freeze/thaw cycles can alter sample isotopic values (De Lecea et al. 2011). M. edulis and T. 55 56348 patagonica samples were obtained in the same way. After obtaining all the samples, they were dried 57 58 59349 for 48 h at 60 °C, ground into a fine powder and ~1.25 μg per sample was sent to the Stable 60 61350 Isotopes Center at the University of New Mexico for C and N stable isotope determination. Internal 62 63 64351 standards used in the laboratory were blue grama, buckeye, cocoa powder, casein, tuna muscle, 65

1352 graphite, whey, elemental protein, serine and IAEA 1N. Standard deviations (SDs) of reference 2 3353 materials were 0.05‰ for δ 13C and 0.07‰ δ 15N. A total of 36 muscle samples were sent, of which 4 5 6354 26 corresponded to green crabs (10 males, 11 females and 5 duplicates), five to M. edulis and five 7 8355 to T. patagonica. The error estimated through duplicate samples for green crabs muscle was 0.14‰ 9 10 13 15 11356 for both δ C and δ N values. 12 13 14357 We estimated the trophic position (TP) of the green crabs using the two baselines (i.e δ 13C 15 16 15 17358 and δ N values of mussels and snails) and the Bayesian approach proposed by Quezada-Romegialli 18 19359 et al. (2018) (See equations in Online Resource ESM 2: eq. 1, eq. 2 and eq. 3). The TDF used for 20 21 13 15 22360 TP estimations were 1.80 ± 0.40 ‰ (δ C) and 2.35 ± 1.26‰ (δ N) following published values for 23 24361 omnivorous crustaceans (Parker et al. 1989; Vanderklist and Ponsard 2003; Rudnick and Resh 25 26362 2005; Yokoyama et al. 2005). For TP estimations sex was taken into account (i.e. independent 27 28 29363 estimates were undertaken for female and male crabs). We performed pairwise comparisons of 30 31 364 posterior TP as the probability of TP A to be less than or equal to TP B (TPA <= TPB) for comparisons 32 33 34365 between sexes. We similarly performed pairwise comparisons of the proportional contribution of 35 36366 each baseline (α). The Bayesian probability was considered to be significant when p > 0.95 as an 37 38 39367 analogous of the frequentist p-value. Estimates of TP and α were done using the 40 41368 ‘TwoBaselinesFull’ model in the R package ‘tRophic Position’ (Quezada-Romegialli et. al. 2018). 42 43 44 45369 The trophic niche of green crabs was described in terms of isotopic niche characteristics 46 47370 following the standard ellipse approach (Jackson et al. 2011). We calculated the standard ellipse 48 49371 area for males and females separately considering 40% of the data and corrected by sample size 50 51 52372 (SEAc) and the Bayesian standard ellipses area (SEAb) using the ‘SIBER’ and “rjags” packages on 53 54373 R (Jackson et al. 2011). Subsequently, we calculated the percentage of overlap between females and 55 56 57374 males SEA as the area at the interception divided by the SEA of male or female respectively. To 58 59375 test differences between SEAb we calculated 100,000 posterior iterations of SEAb based on our 60 61 62376 data set and estimated the probability that the bigger SEAb was higher than the smaller SEAb (i.e. 63 64377 we estimated the probability of male SEAb > female SEAb) (Jackson et al. 2011). 65

1378 3. Results 2 3 4 5379 What are green crabs eating in the coastal environments of Atlantic Patagonia? 6 7 8380 We considered the number of stomachs analyzed (n (vis) =183 and n (met) = 40) sufficient to 9 10381 provide a good description of the diet, because the trophic diversity curves were asymptotic and all 11 12 13382 (n-1) and (n-2) iterations were within (Hz – 0.05) range (Fig. S1 in Online Resource ESM 2). 14 15383 Overall, 15.32% of the visually analyzed stomachs were empty and 35.48% contained material that 16 17 18384 was impossible to identify due to its high degree of digestion. 19 20 21385 Metabarcoding generated a total of 17.47 mio reads. After mBRAVE denoising and filtering 22 23 24386 3,163,504 reads (259,167 ± 164,274 SD per sample) could be assigned to taxa, 2,390,785 (75,6%) 25 26387 of which matched Carcinus. After JAMP processing 1,984,045 quality filtered sequences (110,498 27 28 29388 ± 80,004 SD per sample) were used for clustering and blasted against Genbank’s NR database. 30 31389 1,125,343 of these were assigned to Carcinus (56,7%) (Table S4 in Online Resource ESM 6). One 32 33390 sample did not provide any high-quality reads and four samples did not contain any DNA that could 34 35 36391 be assigned to prey species (i.e. they only contained host/green crab DNA). The first sample was 37 38392 excluded from the analysis and the other four were considered as empty stomachs resulting in 39 40 41393 10.2% of empty guts for metabarcoding. The visual analysis of the gut content showed that green 42 43394 crabs at Golfo Nuevo waters had a broad diet in terms of phylogenetic diversity, with 23 different 44 45 46395 taxa identified (Table 1). Identifications were possible due to prey remains (typically hard parts) 47 48396 found in gut contents. Polychaetes were identified by jaws and mollusks by opercle and radula. We 49 50397 also found some well preserved organisms such as tanaidaceans and amphipods, but finding an 51 52 53398 almost fully intact organism was very unusual (see Fig. S2 in Online Resource ESM 1). Overall, we 54 55399 detected the presence of different prey taxa such as mollusks, algae, arthropoda, annelida, fishes and 56 57 58400 echinodermata. The main prey items were Mytilidae (bivalves) with a FO (vis) of 35.6% followed by 59 60401 the gastropod T. patagonica (12.7%). 61 62 63 64402 Metabarcoding analysis also showed a very wide omnivorous diet, detecting a higher 65

1403 number of phyla and providing a better resolution at the genus level than the visual analysis. In 2 3404 general, FO were higher than FO , indicating that the level of detection with metabarcoding 4 (met) (vis) 5 6405 was higher than through traditional analysis. Here, Arthropoda was the dominant phylum in the 7 8 406 crab's gut, with a FO (met) of 94.3%, followed by Ochrophyta (i.e. Phaeophyta) and Mollusca. The 9 10 11407 main prey items were amphipods of the Gammaridae family (82.9%), a dipteran of the genus 12 13408 Simulium (82.9%), an unidentified species of Heterodonta (45.7%), and algae (Dictyota and an 14 15 16409 unidentified member of Ralfsiales, both 40%) (Table 1). The prey items with high FOs (met) also had 17 18410 the highest read counts (see ESM 3 and ESM 4). 19 20 21 22411 23 24 25412 Are there differences in diet between females and males in Atlantic Patagonia? 26 27 28 29413 We observed small differences of prey items between males and females, evidenced by the 30 31414 overlapping of CIs of the simulated FO (vis). The main difference was a higher frequency of 32 33415 arthropods and a lower frequency of molluscs in the female gut compared to male gut contents (Fig. 34 35 36416 1a), but this difference was not significant (Wd*(vis) = 3.14 and p-value(vis) = 0.058). The FO(met) 37 38417 showed fewer differences between male and female main prey groups (Fig. 1b); and we did not find 39 40 41418 significant differences between their diets either (Wd* (met) = 0.77 and p-value (met) = 0.847). 42 43419 Similarly, OTU ordinations showed a high degree of overlap between males and females (Fig. 2a). 44 45 46 47420 In spite of this small differentiation in prey items, we did detect differences between the gut 48 49421 contents of males and females in their trophic diversity. Metabarcoding OTU richness was 50 51 52422 significantly higher in males than in females (6.6 ± 3.1 and 4.3 ± 1.9, respectively; Kruskal-Wallis 53 54423 chi-squared = 5.5048, p-value = 0.018) (Fig. 2b). In fact, out of 70 OTUs considered prey, only 19 55 56424 were shared between males and females (27%). These shared prey items corresponded roughly with 57 58 59425 the most frequently occurring OTUs, explaining the lack of clear grouping in ordinations (Fig. 2a). 60 61426 A great proportion of items (53%, 37 OTUs) were male-only OTUs while only 20% (14 OTUs) 62 63 64427 were female-only, indicating a higher richness in male diet. 65

1428 Stable isotope analysis (SIA) results were in accordance with metabarcoding analysis. Males 2 3429 showed a broader trophic niche than females, with SEAc larger for males (SEAc = 1.03 and 4 males 5 6430 SEAcfemales = 0.51, Fig. 4a). The Bayesian reconstruction of the SEA showed that the probability of 7 8431 this difference was p = 0.05 (Fig. 4b). However, it should be noted that some overlap at standard 9 10 11432 ellipses between males and females was present (11.89% of male SEA and 24.59% of female SEA) 12 13433 (Fig. 4a and Fig. S3 in Online Resource ESM 2). With respect to the trophic position (TP), female 14 15 16434 crabs showed a posterior estimated TP that was slightly higher than that for males (TP males = 3.02, 17 18435 TPfemales = 3.16 and TP overall = 3.10, Fig 3a) but pairwise comparisons did not result in significant 19 20436 differences for TP or for alpha (α) posterior estimates (Table S5 in Online Resource ESM 2). The 21 22 23437 posterior estimates of the proportional contribution of each baseline showed the phytoplankton 24 25438 energy pathway to be more important than the macroalgae pathway for the nutrition of green crabs 26 27 28439 (α males = 0.74, α females = 0.88 and α overall = 0.86, Fig 3b). This result is in accordance with bivalves 29 30440 (filter feeders) representing the main prey item. 31 32 33 34441 Findings not linked to the green crab’s diet 35 36 37442 Metabarcoding also showed other complementary results that are not diet-related but also 38 39 40443 relevant. In the first place, sequences from both C. maenas and C. aestuarii were detected in most 41 42444 samples, highlighting the possibility of the presence of a hybrid crab population in Patagonian 43 44 45445 coastal waters. Cannibalism could be an alternative explanation and, in fact, it has been previously 46 47446 observed in some green crab populations (Gehrels et al. 2017). However, cannibalism was ruled out 48 49447 because no parts of C. maenas were found in the visual inspections of the stomach contents. 50 51 52448 Besides, the high sequence abundance of the two species in each individual sample suggests 53 54449 hybridization (see Fig. S4 in Online Resource ESM 2). In addition, we detected sequences assigned 55 56 57450 to the genus Carcinonemertes, a nemertean parasite known to infect populations of this crab 58 59451 (Torchín et al. 1996). 60 61 62 63452 64 65

1453 4. Discussion 2 3454 What are green crabs feeding on in the coastal environments of Atlantic Patagonia? 4 5 6 7455 Green crabs in the study site at the Patagonian Atlantic coast display a broad omnivorous 8 9456 diet, consuming the majority of phyla present at the study area. Phyla mostly reported for the 10 11 12457 sampling site (Mollusca, Arthropoda and Phaeophyceae, Rechimont et al. 2013) were also the prey 13 14458 items with the highest occurrences in our dietary samples. Rhodophyta and Annelida were also 15 16 17459 found frequently but at a lower occurrence rate. These results are consistent with previous studies in 18 19460 other areas, where green crabs have been classified as opportunist omnivorous, feeding on a wide 20 21 22461 range of prey species (Elner 1981; Singh 1991; Pardal et al. 2006). Other green crab populations, 23 24462 both native and invasive, have shown a strong preference for mollusks (bivalves and gastropods) 25 26463 (Young and Elliot 2020). In this study, mollusks were among the most common prey items based 27 28 29464 on visual examinations, whilst metabarcoding indicated species of amphipods of the Gammaridae 30 31465 family, which were present in more than 80% of the samples, as the most common prey item. 32 33 34466 Metabarcoding also revealed the presence of other prey items such as Nemertea, Porifera and 35 36467 Ascidiacea, which were not detected in visual analysis of stomach content. It is worth noting that 37 38 39468 the metabarcoding analysis detected several sequences of insects some of which are species that 40 41469 inhabit the intertidal and its surroundings, such as chironomids and other flies. Other species that 42 43470 were detected despite employing strict filtering steps during bioinformatic processing were of 44 45 46471 terrestrial or freshwater origin and do not reside near the study area, such as Ephemeroptera and 47 48472 Trichoptera. These species were likely incorporated into the crab's diet as detritus. Our sampling 49 50 51473 site is located in Northern Patagonia, an arid ecosystem with a precipitation regime below 300 52 53474 mm/year and strong western winds. Western winds have the capacity to transport particles several 54 55 56475 kilometers away from Cordillera de Los Andes to the Southern Atlantic Ocean (Crespi-Abril et al. 57 58476 2018). Insect remains could be carried by these western winds and reach the intertidal zone where 59 60 61477 green crabs ingest them. To our knowledge, this is the first time that evidence for terrestrial insects 62 63478 is found in green crabs guts (Cohen et al. 1995). 64 65

1479 2 3480 Metabarcoding also provided a detailed description of algal diversity in the green crab diet. 4 5 6481 Although algae have been previously detected, identification achieved by visual analysis has been 7 8482 usually very poor (Cohen et al. 1995; Pardal et al. 2006; Wilcox and Rochette 2015). This is due to 9 10 11483 the little amount of intact algae that is usually found and the difficulties associated with taxonomic 12 13484 recognition. DNA analysis clearly represents an improvement for the identification of taxa that are 14 15 16485 notoriously difficult to recognize in degraded samples such as gut contents (Smith et al. 2005; 17 18486 García-Robledo et al. 2013; Taberlet et al. 2018). Although algae were frequently detected in our 19 20487 crabs' diet, stable isotope analysis revealed that the major contribution to green crab diet stems from 21 22 23488 the phytoplankton food chain (i.e. the pelagic pathway of matter and energy). Green crabs in 24 25489 Atlantic Patagonia feed in an opportunistic manner, with bivalves and amphipods being their 26 27 28490 preferred prey; with those they also ingest some algae, but this does not significantly contribute to 29 30491 their nutrition. 31 32 33 34492 Are there differences in diet between female and male green crabs? 35 36493 37 38 39494 In this study, we found some differences between male and female green crab diets 40 41495 suggesting that the two sexes are not as ecologically equivalent as previously thought (Spooner et 42 43496 al. 2007; Young and Elliot 2020). Male and female crabs were found to have similar diets and 44 45 46497 trophic positions (TP), but males had a broader diet than females (i.e. higher trophic diversity). This 47 48498 difference was evidenced by both the isotopic niche and metabarcoding OTU richness. One possible 49 50 51499 explanation for males having a higher trophic diversity is related to differences in behavior between 52 53500 the sexes. Female crabs are usually found in deeper waters and spend a greater part of their life 54 55 56501 subtidally (Reid et al. 1997; Klassen and Locke 2007). It is thought that green crabs migrate 57 58502 towards the intertidal zone to feed; then, females spending more time in the subtidal zone would 59 60 61503 have less opportunities to encounter rare prey than males and therefore they have a lower trophic 62 63504 diversity (Dare and Edwards 1981). Another possible explanation for the sex differences in trophic 64 65

1505 diversity is the sexual dimorphism of the chela. Male crabs have bigger and stronger master chela 2 3506 than females and chela size has been shown to be a good predictor of prey handling time and thus 4 5 6507 can influence feeding rate and energy intake (Elner 1980; Howard et al. 2018). In this context, if 7 8508 males can manipulate their food faster and thus spend less time on each prey item and more time 9 10 11509 searching for food, they could not only eat more, but also increase their chances of encountering 12 13510 less common prey items. 14 15 16511 Finally, it should be noted that dietary overlap between the sexes remains high and that the 17 18512 differences are rather subtle. New dietary studies focusing on seasonality would be very helpful in 19 20513 exploring these differences in Atlantic Patagonia. 21 22 23514 24 25515 What changes can green crabs trigger in local food webs? 26 27 28516 29 30517 Green crabs from newly established populations have shown to be superior foragers 31 32 33518 compared to long-established invaders (Rossong et al. 2012). Therefore, we expect that the most 34 35519 dramatic changes in local food webs will occur in the early stages of the invasion until a new 36 37 38520 equilibrium state with new relative abundances has been reached. However, such a new food web 39 40521 state can bring serious consequences for native populations and even lead to extirpation of some 41 42522 species. The potential impact of green crabs as predators at the study site can be illustrated by 43 44 45523 comparing results to an earlier survey that was part of the Census of Marine Life, NAGISA project. 46 47524 In this study, a total of 35 species of benthic invertebrates and 29 macroalgae were recorded 48 49 50525 (Rechimont et al. 2013), while in the present study we found 31 benthic invertebrates and 23 51 52526 macroalgae in the stomachs of green crabs. Hidalgo et al. (2007) observed that green crabs are 53 54 55527 bigger and feed with greater voracity than native predators on Patagonia's rocky shores. They have 56 57528 provoked negative impacts worldwide such as damaging populations of M. edulis, American oysters 58 59529 (Crassostrea virginica), and rock crabs (Cancer irroratus), among others. They have even caused 60 61 62530 major economic losses for different fisheries (DeGraaf and Tyrrel 2004; Miron et al. 2005; 63 64 65

1531 Sigurdsson and Rochette 2013; Leignel et al. 2014; Young and Elliot 2020). In addition, bird mass 2 3532 mortality has been attributed to the transmission of parasites by green crabs (Camphuysen et al., 4 5 6533 2002). Coastal birds in Patagonia are already incorporating green crabs into their diet, according to 7 8534 a recent kelp gull (Larus dominicanus) study that showed green crab as one of their prey items 9 10 11535 (Yorio et al. 2020). Carcinophagus fish (e.g. Mustelus schmitti) are likely to incorporate green crabs 12 13536 into their diet as well, although this has not been confirmed yet (Alessandra Pasti pers. comm. July 14 15 16537 2020). The fact that the nemertean parasite Carcinonemertes was observed in one of our samples 17 18538 highlights the possible risk of dispersion of these parasites, which had not been reported before 19 20539 among green crabs or native crabs in Argentina (Bigatti and Signorelli 2018). 21 22 23 24540 Community structure at intertidal Patagonian rocky shores is driven by harsh physical 25 26541 conditions, rather than by consumer pressure (Bertness et al. 2006; Hidalgo et al. 2007). Invasive 27 28 29542 green crabs could have a negative impact on the community structure not only by the increase in 30 31543 predation pressure, but also by interfering with processes that maintain biodiversity such as 32 33 34544 facilitation (Silliman et al. 2011). That means green crabs could trigger both direct and indirect 35 36545 effects on Patagonian food webs. We predict that changes will begin with their preferred prey: 37 38 39546 mollusks, especially bivalves, as well as crustaceans, and then scale up to the entire food web 40 41547 through a cascade effect (White et al. 2006; Davies et al. 2017) . We also predict that high predation 42 43548 on foundation species such as mitylids will constitute a loss of refuge for other intertidal species, 44 45 46549 which could result in a decline in biodiversity (Hidalgo et al. 2007). Changes at Golfo Nuevo coasts 47 48550 have been recently observed such as a dramatic decrease of Perumytilus purpuratus and 49 50 51551 Brachidontes rodriguezii populations (Mendez et al. 2019); however, the causes for the mass 52 53552 mortality are unknown and it is currently not possible to establish whether it relates to the green 54 55 56553 crab invasion or to environmental factors. In addition, other species of common native crabs are 57 58554 much more difficult to find at the intertidal since the arrival of green crabs, suggesting that they 59 60 61555 could have a negative interaction with other species of local crabs. 62 63556 64 65

1557 Does the use of multiple techniques contribute to a more complete picture of the green crab’s 2 3558 diet? 4 5 6 7559 The inclusion of complementary techniques to trophic ecology such as stable isotope 8 9560 analyses and metabarcoding can add new results that would otherwise go unnoticed when only 10 11 12561 relying on traditional techniques such as visual inspection of gut contents. These techniques 13 14562 complement each other and our study adds to previous work that demonstrates the usefulness of 15 16 17563 combining SIA and metabarcoding (e.g. Compson et al. 2019; Whitaker et al. 2019). Metabarcoding 18 19564 analysis contributes to the knowledge of diet richness and provides quantitative information about 20 21 22565 it. Stable isotope analysis allows for estimates of species trophic levels and trophic pathways (i.e. 23 24566 which sources contribute the most to species nutrition). The combination of these two techniques in 25 26567 our study provided information on different aspects of trophic ecology, such as interspecific prey- 27 28 29568 predator interactions and intrapopulation differences. In particular, we observed differences 30 31569 between male and female diets related to niche breadth and we achieved a more detailed taxonomic 32 33 34570 resolution for certain taxa due to the use of metabarcoding. 35 36 37571 It should be mentioned at this point, that visual inspections were not redundant in our 38 39 40572 analysis. In addition to the fact that they helped to rule out the possibility of cannibalism 41 42573 (suggesting that this is a hybrid population of C. maenas and C. aestuarii), some prey items found 43 44 45574 in the visual inspections were not detected by metabarcoding (for example: Siphonaria lessoni or 46 47575 Pseudechinus magellanicus). We think that these represent real prey items (not false positives 48 49576 generated by visual inspection) as they (Mollusca, Annelida, Arthropoda, Echinodermata, Fishes) 50 51 52577 have been previously observed as prey items of other green crab populations along the world and 53 54578 they are conspicuous in the intertidal. We have also observed this in laboratory tests, e.g. for 55 56 57579 Siphonaria lessoni. In fact, the discrepancy between metabarcoding and visual inspection is 58 59580 showing some limitations of metabarcoding in our case study: First, the primers that are used for 60 61 62581 metabarcoding are not perfectly universal, which means that some taxa are less likely to be 63 64582 amplified (mainly fish and algae, but also some mollusks are affected). In addition, some of the taxa 65

1583 that were present in the gut content are not part of the reference library yet. Finally, distinct hard 2 3584 parts of some taxa can withstand digestion and are easy to identify morphologically, even after all 4 5 6585 the DNA has been broken down during digestion. Therefore, visual inspection is still recommended 7 8586 and can yield results that complement those of molecular methods such as metabarcoding, 9 10 11587 particularly in geographic areas or particular environments that have not been heavily barcoded yet, 12 13588 such as the Atlantic Patagonian coast. However, it is important to note that the visual and 14 15 16589 metabarcoding techniques have different sensitivity to prey detection, so a strict comparison 17 18590 between the two techniques is not representative. In fact, the visual and metabarcoding data 19 20591 collected in this study were used in a complementary, non-comparative way. 21 22 23 24592 In conclusion, the complementary techniques used in this study allowed us to better 25 26593 understand the trophic biology of the green crab in a recently invaded habitat, showing that this 27 28 29594 species has a broad diet in Atlantic Patagonia, informing about its trophic placement and also 30 31595 suggesting that there are previously unnoticed differences between male and female crab diets. This 32 33 34596 highlights the importance of monitoring coastal ecosystems, in particular in relation to their 35 36597 invasive species, to be able to detect as early as possible and better understand the negative changes 37 38 39598 that they can introduce. 40 41 42599 43 44 45600 REFERENCES 46 47601 Albaina A, Fox C J, Taylor N, Hunter E, Maillard M, and Taylor MI (2010) A TaqMan real- 48 49602 time PCR based assay targeting plaice (Pleuronectes platessa L.) DNA to detect predation by the 50 51 52603 brown shrimp (Crangon crangon L.) and the shore crab (Carcinus maenas L.)—assay development 53 54604 and validation. Journal of Experimental Marine Biology and Ecology, 391(1-2), 178-189. 55 56 57605 Alekseyenko AV (2016) Multivariate Welch t-test on distances. Bioinformatics, 32(23), 58 59606 3552-3558. 60 61 62607 Arnett RT, and Whelan J (2001) Comparing the diet of cod (Gadus morhua) and grey seals 63 64608 (Halichoerus grypus): an investigation of secondary ingestion. Marine Biological Association of the 65

1609 United Kingdom. Journal of the Marine Biological Association of the United Kingdom, 81(2), 365. 2 3610 Bertness MD, Crain CM, Silliman BR, Bazterrica MC, Reyna MV, Hildago F, and Farina 4 5 6611 JK (2006) The community structure of western Atlantic Patagonian rocky shores. Ecological 7 8612 Monographs, 76(3), 439-460. 9 10 11613 Bigatti G, and Signorelli J (2018) Marine invertebrate biodiversity from the Argentine Sea, 12 13614 south western Atlantic. ZooKeys, (791), 47. 14 15 16615 Boecklen WJ, Yarnes CT, Cook BA, and James AC (2011) On the use of stable isotopes in 17 18616 trophic ecology. Annual review of ecology, evolution, and systematics, 42, 411-440. 19 20617 Boschi EE, Fischbach CE, and Iorio MI (1992) Catálogo ilustrado de los crustáceos 21 22 23618 estomatópodos y decápodos marinos de Argentina. 24 25619 Braukmann TW, Ivanova NV, Prosser SW, Elbrecht V, Steinke D, Ratnasingham S, de 26 27 28620 Waard JR, Sones JE, Zakharov EV, and Hebert PD (2019) Metabarcoding a diverse arthropod mock 29 30621 community. Molecular ecology resources, 19(3), 711-727. 31 32 33622 Bruno JF, Fridley JD, Bromberg KD, and Bertness MD (2005) Insights into biotic 34 35623 interactions from studies of species invasions. Species invasions: Insights into ecology, evolution, 36 37 38624 and biogeography, 13-40. 39 40625 Carlton JT, and Cohen AN (2003) Episodic global dispersal in shallow water marine 41 42626 organisms: the case history of the European shore crabs Carcinus maenas and C. aestuarii. Journal 43 44 45627 of Biogeography, 30(12), 1809-1820. 46 47628 Camphuysen CJ, Berrevoets CM, Cremers HJWM, Dekinga A, Dekker R, Ens BJ, van der 48 49 50629 Have TM, Kats RKH, Kuiken T, Leopold MF, van der Meer J, and Piersma, T (2002) Mass 51 52630 mortality of common eiders (Somateria mollissima) in the Dutch Wadden Sea, winter 1999/2000: 53 54 55631 starvation in a commercially exploited wetland of international importance. Biological 56 57632 conservation, 106(3), 303-317. 58 59633 Canty A, and Ripley B (2021) boot: bootstrap R (S-Plus) Functions. Package version 1.3-27. 60 61 62634 Chen RB, Watanabe S, and Yokota M (2004) Feeding habits of an exotic species, the 63 64 65

1635 Mediterranean green crab Carcinus aestuarii, in Tokyo Bay. Fisheries science, 70(3), 430-435. 2 3636 Clare EL (2014) Molecular detection of trophic interactions: emerging trends, distinct 4 5 6637 advantages, significant considerations and conservation applications. Evolutionary applications, 7 8638 7(9), 1144-1157. 9 10 11639 Clavero M, and García-Berthou E (2005) Invasive species are a leading cause of animal 12 13640 extinctions. Trends in ecology and evolution, 20(3), 110. 14 15 16641 Cohen AN, Carlton JT, and Fountain MC. (1995) Introduction, dispersal and potential 17 18642 impacts of the green crab Carcinus maenas in San Francisco Bay, California. Marine Biology, 19 20643 122(2), 225-237. 21 22 23644 Compson ZG, Monk WA, Hayden B, Bush A, O’Malley Z, Hajibabaei M, Porter TM, 24 25645 Wright MTG, Baker CJO, Sadnan Al Manir M, Curry RA, and Baird DJ (2019) Network-based 26 27 28646 biomonitoring: exploring freshwater food webs with stable isotope analysis and DNA 29 30647 metabarcoding. Front. Ecol. Evol. 7: 395. doi: 10.3389/fevo. 31 32 33648 Crespi-Abril AC, Soria G, De Cian A, and López-Moreno C (2018) Roaring forties: An 34 35649 analysis of a decadal series of data of dust in Northern Patagonia. Atmospheric Environment, 177, 36 37 38650 111-119. 39 40651 Dare PJ, and Edwards DB (1981) Underwater television observations on the intertidal 41 42652 movements of shore crabs, Carcinus maenas, across a mudflat. Journal of the Marine biological 43 44 45653 Association of the United Kingdom, 61(1), 107-116. 46 47654 Darling JA (2011) Interspecific hybridization and mitochondrial introgression in invasive 48 49 50655 Carcinus shore crabs. Plos One, 6(3), e17828. 51 52656 David P, Thebault E, Anneville O, Duyck PF, Chapuis E, and Loeuille N (2017) Impacts of 53 54 55657 invasive species on food webs: a review of empirical data. In Advances in ecological research (Vol. 56 57658 56, pp. 1-60). Academic Press. 58 59659 De Lecea AM, Smit AJ, and Fennessy ST (2011) The effects of freeze/thaw periods and 60 61 62660 drying methods on isotopic and elemental carbon and nitrogen in marine organisms, raising 63 64 65

1661 questions on sample preparation. Rapid Communications in Mass Spectrometry, 25(23), 3640-3649 2 3662 Deagle BE, Thomas AC, McInnes JC, Clarke LJ, Vesterinen EJ, Clare EL, Kartzinel TR, 4 5 6663 and Eveson JP (2019) Counting with DNA in metabarcoding studies: How should we convert 7 8664 sequence reads to dietary data?. Molecular ecology, 28(2), 391-406. 9 10 11665 DeGraaf JD, and Tyrrell MC (2004) Comparison of the feeding rates of two introduced crab 12 13666 species, Carcinus maenas and Hemigrapsus sanguineus, on the blue mussel, Mytilus edulis. 14 15 16667 Northeastern Naturalist, 11(2), 163-167. 17 18668 Doherty TS, Glen AS, Nimmo DG, Ritchie EG, and Dickman CR (2016) Invasive predators 19 20669 and global biodiversity loss. Proceedings of the National Academy of Sciences, 113(40), 11261- 21 22 23670 11265. 24 25671 Ebling FJ, Kitching JA, Muntz L, and Taylor CM (1964) Experimental observations of the 26 27 28672 destruction of Mytilus edulis and Nucella lapillus by crabs. Journal of Animal Ecology, 33, 73-82. 29 30673 Efron B, and Tibshirani RJ (1993) An introduction to the bootstrap Chapman & Hall. New 31 32 33674 York, 436. 34 35675 Elbrecht V, Braukmann TW, Ivanova NV, Prosser SW, Hajibabaei M, Wright M, Zakharov 36 37 38676 EV, Hebert PDN, and Steinke D (2019) Validation of COI metabarcoding primers for terrestrial 39 40677 arthropods. PeerJ, 7, e7745. 41 42678 Elbrecht V, and Steinke D (2019) Scaling up DNA metabarcoding for freshwater 43 44 45679 macrozoobenthos monitoring. Freshwater Biology, 64(2), 380-387. 46 47680 Elbrecht V, Vamos EE, Meissner K, Aroviita J, and Leese F (2017) Assessing strengths and 48 49 50681 weaknesses of DNA metabarcoding‐ based macroinvertebrate identification for routine stream 51 52682 monitoring. Methods in Ecology and Evolution, 8(10), 1265-1275. 53 54 55683 Elbrecht V, and Leese F (2017) Validation and development of COI metabarcoding primers 56 57684 for freshwater macroinvertebrate bioassessment. Frontiers in Environmental Science, 5, 11. 58 59685 Elner R.W (1981) Diet of green crab Carcinus maenas (L.) from Port Hebert, southwestern 60 61 62686 Nova Scotia. J. Shellfish Res. 1, 89–94. 63 64 65

1687 Elner RW (1980) The influence of temperature, sex and chela size in the foraging strategy of 2 3688 the shore crab, Carcinus maenas (L.). Marine and Freshwater Behaviour and Phy, 7(1), 15-24. 4 5 6689 Fauchald K (1977) The polychaete worms. Definitions and keys to the orders, families and 7 8690 genera. Natural History Museum of Los Angeles County, Science Series. 9 10 11691 Fridley JD, Stachowicz JJ, Naeem S, Sax DF, Seabloom EW, Smith MD, Stohlgren TJ, 12 13692 Tilman D, and Holle BV (2007) The invasion paradox: reconciling pattern and process in species 14 15 16693 invasions. Ecology, 88(1), 3-17. 17 18694 Galloway AW, Brett MT, Holtgrieve GW, Ward EJ, Ballantyne AP, Burns CW, Kainz MJ, 19 20695 Müller-Navarra DC, Persson J, Ravet JL, Strandberg U, Taipale SJ, and Alhgren G (2015) A fatty 21 22 23696 acid based Bayesian approach for inferring diet in aquatic consumers. PloS one, 10(6), e0129723. 24 25697 García-Robledo C, Erickson DL, Staines CL, Erwin TL, and Kress WJ (2013) Tropical 26 27 28698 plant–herbivore networks: reconstructing species interactions using DNA barcodes. PLoS One, 29 30699 8(1), e52967. 31 32 33700 Gehrels H, Tummon Flynn P, Cox R, and Quijón PA (2017) Effects of habitat complexity 34 35701 on cannibalism rates in European green crabs (Carcinus maenas Linnaeus, 1758). Marine Ecology, 36 37 38702 38(5), e12448 39 40703 Gosztonyi AE, and Kuba L (1996) Atlas de huesos craneales y de la cintura escapular de 41 42704 peces costeros patagónicos. Informe Técnico: Plan de Manejo integrado de la zona costera 43 44 45705 patagónica. Fundación Patagonia Natural. 46 47706 Grosholz ED, and Ruiz GM (1995) Spread and potential impact of the recently introduced 48 49 50707 European green crab, Carcinus maenas, in central California. Marine Biology, 122(2), 239-247. 51 52708 Hamidi B, Wallace K, Vasu C, and Alekseyenko AV (2019) W d ∗ $ W_ {d}^{*} $-test: 53 54 55709 robust distance-based multivariate analysis of variance. Microbiome, 7(1), 1-9. 56 57710 Hayden B, Harrod C, and Kahilainen KK (2014) Dual fuels: Intra ‐ annual variation in the 58 59711 relative importance of benthic and pelagic resources to maintenance, growth and reproduction in a 60 61 62712 generalist salmonid fish. Journal of Animal Ecology, 83(6), 1501-1512. 63 64 65

1713 Hebert PD, Cywinska A, Ball SL, and Dewaard JR (2003) Biological identifications through 2 3714 DNA barcodes. Proceedings of the Royal Society of London. Series B: Biological Sciences, 4 5 6715 270(1512), 313-321. 7 8716 Hidalgo FJ, Silliman BR, Bazterrica MC, and Bertness MD (2007) Predation on the rocky 9 10 11717 shores of Patagonia, Argentina. Estuaries and Coasts, 30(5), 886-894. 12 13718 Howard BR, Barrios-O’Neill D, Alexander ME, Dick JT, Therriault TW, Robinson TB, and 14 15 16719 Côté IM (2018) Functional responses of a cosmopolitan invader demonstrate intraspecific 17 18720 variability in consumer-resource dynamics. PeerJ, 6, e5634. 19 20721 Hui C, Richardson DM, Landi P, Minoarivelo HO, Garnas J, and Roy HE (2016) Defining 21 22 23722 invasiveness and invasibility in ecological networks. Biological Invasions, 18(4), 971-983. 24 25723 Hussey NE, MacNeil MA, McMeans BC, Olin JA, Dudley SF, Cliff G, Wintner SP, 26 27 28724 Fennessy ST, and Fisk AT (2014) Rescaling the trophic structure of marine food webs. Ecology 29 30725 letters, 17(2), 239-250. 31 32 33726 Hyndes G, Hanson C, Vanderklift M (2013) The magnitude of spatial and temporal 34 35727 variation in δ15N and δ13C differs between taxonomic groups: implications for food web studies. 36 37 38728 Estuar. Coast. Shelf Sci. 119, 176–187. 39 40729 Hyslop EJ (1980) Stomach contents analysis—a review of methods and their application. 41 42730 Journal of fish biology, 17(4), 411-429. 43 44 45731 Jackson AL, Inger R, Parnell AC, and Bearhop S (2011) Comparing isotopic niche widths 46 47732 among and within communities: SIBER–Stable Isotope Bayesian Ellipses in R. Journal of Animal 48 49 50733 Ecology, 80(3), 595-602. 51 52734 Klassen GJ, and Locke A (2007) A biological synopsis of the European green crab, 53 54 55735 Carcinus maenas. Moncton, New Brunswick, Canada: Fisheries and Oceans Canada. 56 57736 Koen Alonso MK, Crespo EA, García NA, Pedraza SN, Mariotti PA, and Mora NJ (2002) 58 59737 Fishery and ontogenetic driven changes in the diet of the spiny dogfish, Squalus acanthias, in 60 61 62738 Patagonian waters, Argentina. Environmental Biology of Fishes, 63(2), 193-202. 63 64 65

1739 Lefebvre S, and Dubois S (2016) The stony road to understand isotopic enrichment and 2 3740 turnover rates: insight into the metabolic part. Vie Et Milieu-life And Environment, 66(3-4), 305- 4 5 6741 314. 7 8742 Leignel VSJH, Stillman JH, Baringou S, Thabet R, and Metais I (2014) Overview on the 9 10 11743 European green crab Carcinus spp. (Portunidae, Decapoda), one of the most famous marine 12 13744 invaders and ecotoxicological models. Environmental Science and Pollution Research, 21(15), 14 15 16745 9129-9144. 17 18746 Mackey AP, Hyndes GA, Carvalho MC, and Eyre BD (2015) Physical and biogeochemical 19 20747 correlates of spatio-temporal variation in the 13δ C of marine macroalgae. Estuarine, Coastal and 21 22 23748 Shelf Science, 157, 7-18. 24 25749 McDonald PS, Jensen GC, and Armstrong DA (2001) The competitive and predatory 26 27 28750 impacts of the nonindigenous crab Carcinus maenas (L.) on early benthic phase Dungeness crab 29 30751 Cancer magister Dana. Journal of Experimental Marine Biology and Ecology, 258(1), 39-54. 31 32 33752 Majdi, N, Hette-Tronquart N, Auclair E, Bec A, Chouvelon T, Cognie B, Danger M, 34 35753 Decottignies P, Dessier A, Desvilettes C, Dubois S, Dupuy C, Fritsch C, Gaucherel C, Hedde M, 36 37 38754 Jabot F, Lefebvre S, Marzloff M, Pey B, Peyrard N, Powolny T, Sabbadin R, Thebault E, & Perga 39 40755 ME. (2018) There's no harm in having too much: A comprehensive toolbox of methods in trophic 41 42756 ecology. Food webs, 17, e00100. 43 44 45757 McMurdie PJ, and Holmes S (2013) Phyloseq: An R package for reproducible interactive 46 47758 analysis and graphics of microbiome census data. PLoS ONE 8(4):e61217. 48 49 50759 Mendez MM, Livore JP, Márquez F, Gotlieb E, and Bigatti G (2019) Drástica pérdida de 51 52760 cobertura en mejillinares del golfo nuevo.3° Congreso Argentino de Malacología. 4 al 6 de 53 54 55761 diciembre de 2019. Bahía Blanca, Argentina. 56 57762 Miron G, Audet D, Landry T, and Moriyasu M (2005) Predation potential of the invasive 58 59763 green crab (Carcinus maenas) and other common predators on commercial bivalve species found 60 61 62764 on Prince Edward Island. Journal of Shellfish Research, 24(2), 579-586. 63 64 65

1765 Mooney CZ, and Duvall R (1993) Bootstrapping: A nonparametric approach to statistical 2 3766 inference (No. 95). 4 5 6767 Nakahara F, Ando H, Ito H, Murakami A, Morimoto N, Yamasaki M, Takayanagi A, and 7 8768 Isagi Y (2015) The applicability of DNA barcoding for dietary analysis of sika deer. DNA 9 10 11769 Barcodes, 3(1). 12 13770 Nielsen JM, Clare EL, Hayden B, Brett MT, and Kratina P (2018) Diet tracing in ecology: 14 15 16771 Method comparison and selection. Methods in Ecology and Evolution, 9(2), 278-291. 17 18772 Oksanen J, Blanchet, FG, Kindt R, Legendre P, O’hara RB, Simpson GL, Solymos P, Henry 19 20773 M, Szoecs S, Szoecs E, and Wagner H (2019) Vegan: community ecology package. R package 21 22 23774 version 2.5-6. URL http://CRAN. R-project. org/package= vegan. 24 25775 Pardal M, Baeta A, Marques J, and Cabral H (2006) Feeding ecology of the green crab, 26 27 28776 Carcinus maenas (L., 1758) in a temperate estuary, Portugal. Crustaceana, 79(10), 1181-1193. 29 30777 Parker PL, Anderson RK, and Lawrence A (1989) A δ 13C and δ15N tracer study of nutrition 31 32 33778 in aquaculture: Penaeus vannamei in a pond growout system. In Stable isotopes in ecological 34 35779 research (pp. 288-303). Springer, New York, NY. 36 37 38780 Philippsen JS, and Benedito E (2013) Discrimination factor in the trophic ecology of fishes: 39 40781 a review about sources of variation and methods to obtain it. Oecologia Australis, 17(2), 205-2016. 41 42782 Pielou EC (1966) The measurement of diversity in different types of biological collections. 43 44 45783 Journal of theoretical biology, 13, 131-144. 46 47784 Pinnegar JK, and Polunin NVC (1999) Differential fractionation of δ 13C and δ 15N among 48 49 50785 fish tissues: implications for the study of trophic interactions. Functional ecology, 13(2), 225-231. 51 52786 Pompanon F, Deagle BE, Symondson WO, Brown DS, Jarman SN, and Taberlet P (2012) 53 54 55787 Who is eating what: diet assessment using next generation sequencing. Molecular ecology, 21(8), 56 57788 1931-1950. 58 59789 Post, D M (2002) Using stable isotopes to estimate trophic position: models, methods, and 60 61 62790 assumptions. Ecology, 83(3), 703-718. 63 64 65

1791 Quezada-Romegialli C, Jackson AL, Hayden B, Kahilainen KK, Lopes C, and Harrod C 2 3792 (2018) tRophicPosition, an R package for the Bayesian estimation of trophic position from 4 5 6793 consumer stable isotope ratios. Methods in Ecology and Evolution, 9(6), 1592-1599. 7 8794 R Core Team (2020) R: A language and environment for statistical computing. R 9 10 11795 Foundation for Statistical Computing, Vienna, Austria. URL https://www.R-project.org/. 12 13796 Ratnasingham S (2019) mBRAVE: The Multiplex Barcode Research And Visualization 14 15 16797 Environment. Biodiversity Information Science and Standards, 3, e37986. 17 18798 Ratnasingham S, and Hebert PD (2007) BOLD: The Barcode of Life Data System 19 20799 (http://www. barcodinglife. org). Molecular ecology notes, 7(3), 355-364. 21 22 23800 Rechimont ME, Galvan DE, Sueiro MC, Casas G, Piriz ML, Diez ME, Primost M, Zabala 24 25801 MS, Márquez F, Brogger M, Alfaya JF, and Bigatti G (2013) Benthic diversity and assemblage 26 27 28802 structure of a north Patagonian rocky shore: a monitoring legacy of the NaGISA project. Marine 29 30803 Biological Association of the United Kingdom. Journal of the Marine Biological Association of the 31 32 33804 United Kingdom, 93(8), 2049. 34 35805 Reid, DG, Abelló P, Kaiser, MJ, and Warman CG (1997) Carapace colour, inter-moult 36 37 38806 duration and the behavioural and physiological ecology of the shore crabCarcinus maenas. 39 40807 Estuarine, Coastal and Shelf Science, 44(2), 203-211. 41 42808 Rodriguez LF (2006) Can invasive species facilitate native species? Evidence of how, when, 43 44 45809 and why these impacts occur. Biol Invasions 8: 927−939 46 47810 Romanuk TN, Zhou Y, Valdovinos FS, and Martinez ND (2017) Robustness trade-offs in 48 49 50811 model food webs: invasion probability decreases while invasion consequences increase with 51 52812 connectance. In Advances in Ecological Research (Vol. 56, pp. 263-291). Academic Press. 53 54 55813 Rossong MA, Quijón PA, Snelgrove PV, Barrett TJ, McKenzie CH, and Locke A (2012) 56 57814 Regional differences in foraging behaviour of invasive green crab (Carcinus maenas) populations in 58 59815 Atlantic Canada. Biological Invasions, 14(3), 659-669. 60 61 62816 63 64 65

1817 Rudnick D, and Resh V (2005) Stable isotopes, mesocosms and gut content analysis 2 3818 demonstrate trophic differences in two invasive decapod crustacea. Freshwater Biology, 50(8), 4 5 6819 1323-1336 7 8820 Russel J (2020) MicEco: Various functions for microbial community data. R package 9 10 11821 version 0.9.9. 12 13822 Schwindt E, Carlton JT, Orensanz JM, Scarabino F, and Bortolus A (2020) Past and future 14 15 16823 of the marine bioinvasions along the Southwestern Atlantic. Aquatic Invasions, 15(1). 17 18824 Sigurdsson GM, and Rochette R (2013) Predation by green crab and sand shrimp on settling 19 20825 and recently settled American lobster postlarvae. J. Crustac. Biol. 33, 10–14 21 22 23826 Silliman BR, Bertness MD, Altieri AH, Griffin JN, Bazterrica MC, Hidalgo FJ, Crain CM, 24 25827 and Reyna MV (2011) Whole-community facilitation regulates biodiversity on Patagonian rocky 26 27 28828 shores. PLoS One, 6(10), e24502. 29 30829 Singh R (1991) Natural Diet of, and Shelter-Use by, the Green Crab, Carcinus maenas (L.) 31 32 33830 in Nova Scotia. Master’s Thesis, University of New Brunswick, Fredericton, NB, Canada, 34 35831 Smith PJ, McVeagh SM, Allain V, and Sanchez C (2005) DNA identification of gut 36 37 38832 contents of large pelagic fishes. Journal of Fish Biology, 67(4), 1178-1183. 39 40833 Spivak ED, Farías NE, Ocampo EH, Lovrich GA, and Luppi TA (2019) Annotated 41 42834 catalogue and bibliography of marine and estuarine shrimps, lobsters, crabs and their allies 43 44 45835 (Crustacea: Decapoda) of Argentina and Uruguay (Southwestern Atlantic Ocean). Frente Marítimo, 46 47836 26, 1-164. 48 49 50837 Spooner EH, Coleman RA, and Attrill MJ (2007) Sex differences in body morphology and 51 52838 multitrophic interactions involving the foraging behaviour of the crab Carcinus maenas. Marine 53 54 55839 Ecology, 28(3), 394-403. 56 57840 Stock BC, Jackson AL, Ward EJ, Parnell AC, Phillips DL, and Semmens BX (2018) 58 59841 Analyzing mixing systems using a new generation of Bayesian tracer mixing models. PeerJ, 6, 60 61 62842 e5096. 63 64 65

1843 Taberlet P, Bonin A, Coissac E, and Zinger L (2018) Environmental DNA: For biodiversity 2 3844 research and monitoring. Oxford University Press. 4 5 6845 Taberlet P, Coissac E, Pompanon F, Brochmann C, and Willerslev E (2012) Towards 7 8846 next‐ generation biodiversity assessment using DNA metabarcoding. Molecular ecology, 21(8), 9 10 11847 2045-2050. 12 13848 Thomsen MS, Byers JE, Schiel DR, Bruno JF, Olden JD, Wernberg T, and Silliman BR 14 15 16849 (2014) Impacts of marine invaders on biodiversity depend on trophic position and functional 17 18850 similarity. Marine Ecology Progress Series, 495, 39-47. 19 20851 Thomsen MS, Olden JD, Wernberg T, Griffin JN, and Silliman BR (2011a) A broad 21 22 23852 framework to organize and compare ecological invasion impacts. Environmental research, 111(7), 24 25853 899-908. 26 27 28854 Thomsen MS, Wernberg T, Olden JD, Griffin JN, and Silliman BR (2011b) A framework to 29 30855 study the context-dependent impacts of marine invasions. Journal of Experimental Marine Biology 31 32 33856 and Ecology, 400(1-2), 322-327. 34 35857 Thomsen MS, Wernberg T, Altieri A, Tuya F, Gulbransen D, McGlathery KJ, Holmer M, 36 37 38858 and Silliman BR (2010) Habitat cascades: the conceptual context and global relevance of 39 40859 facilitation cascades via habitat formation and modification. Integrative and Comparative Biology, 41 42860 50(2), 158-175. 43 44 45861 Tilman D, Isbell F, and Cowles JM (2014) Biodiversity and ecosystem functioning. Annual 46 47862 review of ecology, evolution, and systematics, 45, 471-493. 48 49 50863 Tirasin EM, and Jørgensen T (1999) An evaluation of the precision of diet description. 51 52864 Marine Ecology Progress Series, 182, 243-252. 53 54 55865 Torchin ME, Lafferty KD, and Kuris AM (1996) Infestation of an introduced host, the 56 57866 European green crab, Carcinus maenas, by a symbiotic nemertean egg predator, Carcinonemertes 58 59867 epialti. The Journal of parasitology, 449-453. 60 61 62868 63 64 65

1869 Torres PJ, and González-Pisani X (2016) Primer registro del cangrejo verde, Carcinus 2 3870 maenas (Linnaeus, 1758), en Golfo Nuevo, Argentina: un nuevo límite norte de distribución en 4 5 6871 costas patagónicas. Ecología austral, 26(2), 134-137. 7 8872 Thalinger B, Oehm J, Zeisler C, Vorhauser J, and Traugott M (2018) Sex‐ specific prey 9 10 11873 partitioning in breeding piscivorous birds examined via a novel, noninvasive approach. Ecology and 12 13874 evolution, 8(17), 8985-8998. 14 15 16875 Traugott M, Thalinger B, Wallinger C, and Sint D (2021) Fish as predators and prey: 17 18876 DNA‐ based assessment of their role in food webs. Journal of Fish Biology. 19 20877 https://doi.org/10.1111/jfb.14400 21 22 23878 Traugott M, Kamenova S, Ruess L, Seeber J, and Plantegenest M (2013) Empirically 24 25879 characterising trophic networks: what emerging DNA-based methods, stable isotope and fatty acid 26 27 28880 analyses can offer. In Advances in ecological research (Vol. 49, pp. 177-224). Academic Press. 29 30881 Unger SD, Williams LA, Diaz L, and Jachowski CB (2020) DNA barcoding to assess diet of 31 32 33882 larval eastern hellbenders in North Carolina. Food Webs, 22, e00134. 34 35883 Whitaker MRL, Baker CCM, Salzman SM, Martins DJ, Pierce NE (2019) Combining stable 36 37 38884 isotope analysis with DNA metabarcoding improves inferences of trophic ecology. PLoS ONE 39 40885 14(7): e0219070. https://doi.org/10.1371/journal.pone.0219070 41 42886 White EM, Wilson JC, and Clarke AR (2006) Biotic indirect effects: a neglected concept in 43 44 45887 invasion biology. Diversity and distributions, 12(4), 443-455. 46 47888 Wilcox M, and Rochette R (2015) Does claw morphology of the green crab Carcinus 48 49 50889 maenas vary in relation to its diet on rocky versus fine-sediment shores of southwest New 51 52890 Brunswick, Bay of Fundy, Canada?. Journal of Experimental Marine Biology and Ecology, 465, 53 54 55891 121-129. 56 57892 Yokoyama H, Tamaki A, Harada K, Shimoda K, Koyama K, and Ishihi Y (2005) Variability 58 59893 of diet-tissue isotopic fractionation in estuarine macrobenthos. Marine Ecology Progress Series, 60 61 62894 296, 115-128. 63 64 65

1895 Yorio P, Suárez N, Kasinsky T, Pollicelli M, Ibarra C, and Gatto, A (2020) The introduced 2 3896 green crab (Carcinus maenas) as a novel food resource for the opportunistic kelp gull (Larus 4 5 6897 dominicanus) in Argentine Patagonia. Aquatic Invasions, 15(1), 140-159. 7 8898 Young AM, and Elliott JA (2020) Life History and Population Dynamics of Green Crabs 9 10 11899 (Carcinus maenas). Fishes, 5(1), 4. 12 13900 Zacharia P, Abdurahiman K, and Mohamed K (2004) Methods of stomach content analysis 14 15 16901 of fishes. Winter School on Towards Ecosystem Based Management of Marine Fisheries—Building 17 18902 Mass Balance Trophic and Simulation Models, 1, 148-158. 19 20903 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

1904 Figure Legends 2 3905 4 5 6906 Figure 1. Frequency of occurrence (FO) boxplots of major prey groups in green crab diet from 7 8907 Patagonia, Argentina for: a) visual and b) metabarcoding analyses. The lower and upper hinges 9 10 11908 correspond to the first and third quartiles (the 25th and 75th percentiles). Points are outliers (data > 12 13909 1.5* IQR) and horizontal lines are medians. 14 15 16910 17 18911 Figure 2. Differences between male and female diets in green crabs. a) Ordination of samples 19 20912 based on COI metabarcoding data (OTUs, NCBI-based assignments), depicted by sex. Distance 21 22 23913 measure: Jaccard index. b) Distribution of OTU richness (number of different OTUs) by sex. 24 25914 Letters "a" and "b" indicate statistically significant differences between sexes (p<0.05; Kruskal- 26 27 28915 Wallis rank-sum test). 29 30 31916 Figure 3. Posterior estimates from Stable Isotope Analysis of the green crabs at Golfo Nuevo, 32 33 34917 Patagonia, Argentina. a) Trophic position (TP) for female and male green crabs b) Alpha (α) for 35 36918 female and male green crabs. Means are indicated by dotted lines (males: blue; females: pink; both 37 38 39919 sexes: black). 40 41 42920 Figure 4. Stable Isotope Analysis of the green crabs at Golfo Nuevo, Patagonia, Argentina. a) 43 44 15 13 45921 δ N‰ and δ C‰ values for female and male green crabs. Ellipses represent the standard ellipses 46 47922 for each sex corrected for sample size. b) Standard ellipse area (‰2) estimated by Bayesian 48 49923 methods (SEAb) for male and female green crabs. 50 51 52924 53 54 55 56 57 58 59 60 61 62 63 64 65

1925 Figure 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58926 59 60 61927 Figure 2 62 63 64 65

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24928 25 26 27 28929 29 30 31930 32 33 34 35931 36 37 38932 39 40 41 42933 43 44 45934 46 47 48 49935 50 51 52936 53 54 55 56937 57 58 59938 60 61 62 63939 Figure 3 64 65

1940 2 3 4941 5 6 7 8942 9 10 11943 12 13 14 15944 16 17 18945 19 20 21 22946 23 24 25947 26 27 28 29948 30 31 32949 33 34 35 36950 37 38 39951 40 41 42 43952 44 45 46953 47 48 49 50954 51 52 53955 54 55 56956 57 58 59 60 61 62 63 64 65

1957 Figure 4 2 3 4 958 5 6 7 8 959 9 10 11 960 12 13 14 15 961 16 17 18 962 19 20 21 22 963 23 24 25 964 26 27 28 29 965 30 31 32 966 33 34 35 36 967 37 38 39 968 40 41 42 43 969 44 45 46 970 47 48 49 50 971 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

Table 1 bioRxiv preprint doi: https://doi.org/10.1101/2020.08.13.249896; this version posted MayClick 28, 2021. here The to copyright access/download;Table;Table holder for this preprint 1.odt (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

Table 1. List of taxa in green crab stomach content determined by metabarcoding and visual analyses.

Frequency of occurrence (FO%) for metabarcoding (met) and visual inspection (vis) of each taxa are shown.

*Indicates taxa with no records in Punta Este intertidal but that can inhabit nearby terrestrial areas.

**Indicates taxa that inhabit freshwater several kilometers away from the sampled area. Both * and ** are

probably incorporated as detritus by green crabs.

Phylum Subphylu Class Subclass Order/Subcla Family/Suborde Genus/Specie FO FO m ss r (Met) (Vis) Mollusca 74.28 46.31 Bivalvia 68.57 35.57 Mytilida 17.14 35.57 Mytilidae 17.14 35.57 Aulacomya 2.86 - Brachidontes/ 5.71 - Aulacomya/ Mytilus Brachidontes/Peru 8.57 - mytilus Heterodont 62.86 - a Heterodonta 62.86 - Heterodonta - Heterodonta Unid. 25.71 - 1 Heterodonta Unid. 45.71 - 2 Myoida 2.86 - Myoida Unid. 1 2.86 - Myoida Unid. 2 2.86 - Gastropoda 8.57 18.12 Trochidae 8.57 12.75 Tegulidae 8.57 12.75 Tegula 8.57 12.75 Siphonariida - 4.03 Siphonariidae - 4.03 Siphonaria lessoni - 4.03 Polyplacophor - 0.67 a Chitonida - 0.67 Chaetopleuridae - 0.67 Chaetopleura - 0.67 Annelida 22.87 16.78 Clitellata 11.43 - Haplotaxida 11.43 - Naididae 11.43 - Nadididae Unid. 1 8.57 - Nadididae Unid. 2 2.86 - Haplotaxida 5.71 - Unid. Haplotaxida Unid. 2.86 - 1 Haplotaxida Unid. 2.86 - 2 Polychaeta 22.86 - Phyllodocida 14.28 4.70 Nereididae 3.36 Platynereis 5.71 - Nereidae Unid - 3.36 Phyllodocidae 2.86 - bioRxiv preprint doi: https://doi.org/10.1101/2020.08.13.249896; this version posted May 28, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

Eulalia 2.86 - Polynoidae - 0.67 Polynoidae Unid. - 0.67 Glyceridae - 0.67 Glyceridae Unid. - 0.67 Palpata 2.86 - Sabellida 2.86 - Sabellida Unid. 2.86 - Eunicida - 0.67 Eunicidae - 0.67 Eunicidae Unid. - 0.67 Terebellida 2.86 - Terebellidae 2.86 - Terebellidae Unid. 2.86 - Other Polychaeta Unid. - 11.41 Polychaeta Unid. Nemertea 11.43 - Pilidiophora 11.43 - Heteronemerte 11.43 - a Heteronemertea 11.43 - Heteronemertea 11.43 - Unid. Arthropoda 94.28 20.80 Crustacea 85.71 20.13 Malacostraca 82.86 6.03 Amphipoda 82.86 1.34 Gammaridae 82.86 - Gammaridae Unid. 82.86 - Corophiidae - 1.34 Monocorophium - 1.34 insidiosum Decapoda 5.71 - Porcellanidae 5.71 - Pachycheles 5.71 - Isopoda 5.71 0.67 Sphaeromatidae 5.71 - Exosphaeroma/Pse 2.86 - udosphaeroma Sphaeroma 2.86 - Isopoda Unid. - 0.67 Euphausiacea - 0.67 Euphausiacea - 0.67 Unid. Euphausiacea Unid. - 0.67 Tanaidacea - 2.68 Tanaididae - 2.68 Tanais dulongii - 2.68 Cumacea - 0.67 Cumacea Unid. - 0.67 Cumacea Unid - 0.67 Hexanauplia 2.86 - Copepoda 2.86 - Cyclopoida 2.86 - Oithonidae 2.86 - Oithona similis 2.86 - Cirripedia - 1.32 Cirripedia - 1.32 Unid. Cirripedia Unid. - 1.32 Cirripedia Unid. - 1.32 Other Crustacea Unid. - 18.79 Crustacea Unid. Insecta 91.43 0.67 Colleoptera 14.28 - Chrysomelidae 2.86 - Chrysomelidae 2.86 - Unid.* bioRxiv preprint doi: https://doi.org/10.1101/2020.08.13.249896; this version posted May 28, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

Dytiscidae 2.86 - Dytiscidae Unid.* 2.86 - Elateridae 2.86 - Elateridae Unid.* 2.86 - Elmidae 8.57 - Elmidae Unid. 1* 2.86 - Elmidae Unid. 2* 8.57 - Diptera 88.57 0.67 Anthomyiidae 2.86 - Anthomyiidae* 2.86 - Calliphoridae 8.57 - Calliphoridae Unid. 8.57 - * Ceratopogonidae 2.86 - Ceratopogonidae 2.86 - Unid.* Chironomidae 54.28 0.67 Chironomus 2.86 - Chironomidae 2.86 - Unid. 1 Eukiefferiella 5.71 - Chironomidae 8.57 - Unid. 2 Chironomidae 2.86 - Unid. 3 Chironomidae Unid 37.14 - 4. Polypedilum 2.86 - Tanytarsus 2.86 - Chironomidae 31.43 - Unid. 5 Chironomidae - 0.67 Unid. 6 (larvae) Simuliidae 82.86 - Simulium* 82.86 - Dolichopodidae 11.43 - Dolichopodidae 5.71 - Unid 1.* Dolichopodidae 5.71 - Unid 2.* Dolichopodidae 5.71 - Unid 3.* Micropezidae 2.86 - Micropezidae 2.86 - Unid.* Tipulidae 2.86 - Tipulidae Unid.* 2.86 - Tabanidae 11.43 - Chrysops* 11.43 - Tachinidae 2.86 - Tachinidae Unid.* 2.86 - Asilidae 5.71 - Atomosia* 2.86 - Laphriinae Unid.* 5.71 - Odonata 11.43 - Aeshnidae 2.86 - Aeshnidae Unid** 2.86 - Coenagrionidae 8.57 - Coenagrionidae 8.57 - Unid.** Lepidoptera 5.71 - Erebidae 5.71 - Ctenucha** 5.71 - Trichoptera 20.00 - Hydropsychidae 11.43 - bioRxiv preprint doi: https://doi.org/10.1101/2020.08.13.249896; this version posted May 28, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

Hydropsychidae 2.86 - Unid. 1** Hydropsychidae 11.43 - Unid. 2** Limnephilidae 2.86 - Limnephilidae 2.86 - Unid.** Phryganeidae/Li 8.57 - mnephilidae Phryganeidae/Limn 8.57 - ephilidae Unid.** Trichoptera 2.86 - Unid. Trichoptera 2.86 - Unid.** Plecoptera 5.71 - Perlodidae 5.71 - Perlodidae Unid.** 5.71 - Hemiptera 5.71 - Aphididae 2.86 - Aphididae Unid.* 2.86 - Corixidae 2.86 - Corixidae Unid.* 2.86 - Megaloptera 2.86 - Sialidae 2.86 - Sialidae Unid.** 2.86 - Ephemeropter 2.86 - a Caenidae 2.86 - Caenis** 2.86 - Collembola 25.71 - Poduromorpha 25.71 - Neanuridae 25.71 - Neanuridae Unid. 25.71 - Arachnida 2.86 - Trombidiform 2.86 - es Demodicidae 2.86 - Demodex 2.86 - Echinodermat - 0.67 a Echinoidea - 0.67 Camarodonta - 0.67 Temnopleuridae - 0.67 Pseudechinus - 0.67 magellanicus Cnidaria 2.86 - Hydrozoa 2.86 - Anthoathecata 2.86 - Pandeidae 2.86 - Amphinema 2.86 - Porifera 14.28 - Demospongiae 14.28 - Demospongiae 14.28 - Unid. Chordata 8.57 1.32 Ascidiacea 2.86 - Enterogona 2.86 - Enterogona 2.86 - Unid. Enterogona Unid 1. 2.86 - Actinopterygii 5.71 - Blenniiformes 2.86 - Blenniiformes 2.86 - Unid. 1 bioRxiv preprint doi: https://doi.org/10.1101/2020.08.13.249896; this version posted May 28, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

Actinopterygii 5.71 - Unid. Actinopterygii 2.86 - Unid. 1 Actinopterygii 5.71 - Unid. 2 Fishes Unid. - 1.32 Metazoa 5.71 - Unid. Metazoa Unid. 5.71 - Ochrophyta 82.86 2.68 Phaeophyceae 82.86 2.68 Ectocarpales 48.57 - Acinetosporacea 11.43 - e Acinetosporaceae 2.86 - Unid. 1. Acinetosporaceae 8.57 Unid. 2 Chordariaceae 40.00 - Hecatonema 14.28 - Myrionema 8.57 - Microspongium 8.57 - Chordariaceae 2.86 - Unid. 1 Chordariaceae 2.86 - Unid. 2 Chordariaceae 2.86 - Unid. 3 Chordariaceae 2.86 - Unid. 4 Ectocarpaceae 17.14 - Ectocarpus Unid. 1 14.28 - Ectocarpus Unid. 2 5.71 Scytosiphonacea 14.28 - e Scytosiphon 5.71 - Petalonia 5.71 - Planosiphon 2.86 - Ectocarpales 5.71 - Unid. Ectocarpales Unid. 2.86 - 1 Ectocarpales Unid. 2.86 - 2 Dictyotales 40.00 - Dictyotaceae 40.00 - Dictyota 40.00 - Ralfsiales 40.00 - Ralfsiales Unid. 40.00 - Ralfsiales Unid. 1 40.00 - Ralfsiales Unid. 2 2.86 - Sphacelariales 2.86 - Sphacelariales 2.86 - Unid. bioRxiv preprint doi: https://doi.org/10.1101/2020.08.13.249896; this version posted May 28, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

Sphacelariales 2.86 - Unid. 1 Laminariales 34.28 - Alariaceae 34.28 - Undaria pinnatifida 34.28 - Rhodophyta 31.43 2.68 Florideophycea 28.57 - e Ceramiales 22.86 - Rhodomelaceae 20.00 - Polysiphonia 2.86 - Streblocladia 2.86 Gredgaria 17.14 Ceramiales 2.86 - Unid. Ceramiales Unid. 2.86 - Corallinales 5.71 - Corallinaceae 2.86 - Corallina Unid. 2.86 - Corallinacea 2.86 - Unid. Corallinacea Unid. 2.86 - Bangiophyceae 5.71 - Bangiales 5.71 - Bangiaceae 5.71 - Bangiales Unid. 2.86 - Bangiaceae Unid. 2.86 Chlorophyta 2.86 0.67 Ulvophyceae 2.86 - Ulvales 2.86 - Ulvaceae 2.86 - Ulva 2.86 - Other Algae Algae Unid. - 23.49 Unid.