Disentangling the Taxonomic Status of Caprella penantis sensu stricto (Amphipoda: Caprellidae) Using an Integrative Approach
Abstract
Despite its importance in intertidal and shallow-water marine ecosystems, Caprella penantis continues to be one of the most taxonomically challenging amphipods in the world. A recent molecular study focusing on C. penantis sensu stricto pointed out the existence of three highly divergent lineages, indicating the possible existence of a process of ongoing speciation and, thus, casting doubt on the taxonomic status of this species. In the present study, we used an integrative approach to continue to shed light on the taxonomy and distribution of this caprellid. To this end, we combined morphological and genetic data (COI and 18S) and included, for the first time, populations from its type locality. Our analyses provide strong evidence of the existence of potentially three distinct species, genetically and geographically restricted, within C. penantis sensu stricto, with the distribution of the true C. penantis sensu stricto restricted to the UK (type locality), the northern coast of the Iberian Peninsula, and the Azores. Results show the co-occurrence of two of these species in a locality of northern Portugal and indicate the existence of distinct evolutionary and diversification patterns along the eastern Atlantic region. Overall, our study highlights the use of an integrative approach to properly assess species boundaries and unravel hidden biodiversity in amphipods.
Full text
Citation: Cabezas, M.P.; Guerra-García, J.M.; Santos, A.M. Disentangling the Taxonomic Status of Caprella penantis sensu stricto (Amphipoda: Caprellidae) Using an Integrative Approach. Life 2022,12, 155. https://doi.org/10.3390/ life12020155 Academic Editors: Pedro Vieira, Filipe Costa and Pedro Soares Received: 15 December 2021 Accepted: 19 January 2022 Published: 21 January 2022 Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations. Copyright: © 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/). life Article Disentangling the Taxonomic Status of Caprella penantis sensu stricto (Amphipoda: Caprellidae) Using an Integrative Approach M. Pilar Cabezas 1,2,* , JoséM. Guerra-García3and António M. Santos 1,2 1Faculdade de Ciências da Universidade do Porto, Rua do Campo Alegre s/n, 4169-007 Porto, Portugal; [email protected] 2CIBIO-InBIO, Centro de Investigação em Biodiversidade e Recursos Genéticos, Universidade do Porto, 4485-661 Vairão, Portugal 3Laboratorio de Biología Marina, Departamento de Zoología, Facultad de Biología, Universidad de Sevilla, 41012 Seville, Spain; [email protected] *Correspondence: pilar[email protected] Abstract: Despite its importance in intertidal and shallow-water marine ecosystems, Caprella penantis continues to be one of the most taxonomically challenging amphipods in the world. A recent molecular study focusing on C. penantis sensu stricto pointed out the existence of three highly divergent lineages, indicating the possible existence of a process of ongoing speciation and, thus, casting doubt on the taxonomic status of this species. In the present study, we used an integrative approach to continue to shed light on the taxonomy and distribution of this caprellid. To this end, we combined morphological and genetic data (COI and 18S) and included, for the first time, populations from its type locality. Our analyses provide strong evidence of the existence of potentially three distinct species, genetically and geographically restricted, within C. penantis sensu stricto, with the distribution of the true C. penantis sensu stricto restricted to the UK (type locality), the northern coast of the Iberian Peninsula, and the Azores. Results show the co-occurrence of two of these species in a locality of northern Portugal and indicate the existence of distinct evolutionary and diversification patterns along the eastern Atlantic region. Overall, our study highlights the use of an integrative approach to properly assess species boundaries and unravel hidden biodiversity in amphipods. Keywords: crustacea; DNA barcoding; hidden diversity; morphology; northeast Atlantic; nuclear DNA; species delimitation; Strait of Gibraltar; taxonomy 1. Introduction Amphipods constitute one of the dominant taxa in shallow-water marine ecosystems worldwide. They can be easily found as epibionts of a wide variety of natural substrata, including macroalgae, hydrozoans, ascidians, sponges, corals, etc. [ 1 – 5 ], but they are also very successful colonisers of artificial substrata such as buoys, ropes, boat hulls, or pontoons, where they can reach high densities under optimal environmental conditions [6–9]. These small, benthic crustaceans form an important trophic link between primary producers and higher trophic levels [ 10 – 12 ], as they are important dietary components for many coastal marine fish species [ 10 ]. Furthermore, recent studies have shown the great usefulness of amphipods as marine bioindicators [ 13 – 17 ] and as potential resources in aquaculture [10,18–22]. At present, amphipods comprise more than 9000 species [ 23 , 24 ]. However, and despite the fact that the study of these invertebrates has increased significantly during the last few decades, amphipods remain one of the least described taxa among crustaceans [ 24 ], which means that even the most recent diversity assessments are likely underestimated [ 24 – 26 ]. In fact, a recent study indicates that about two-thirds of amphipods still remain to be discovered [ 24 ]. As with many other marine invertebrates, knowledge of amphipod biodiversity Life 2022,12, 155. https://doi.org/10.3390/life12020155 https://www.mdpi.com/journal/life
Life 2022,12, 155 2 of 17 is mainly hampered by taxonomic impediment [ 26 ], but also by the realisation that cryptic species (those that are genetically distinct but lack morphological differentiation) are more common than previously anticipated [ 27 – 30 ]. Morphological identification of amphipod species demands the examination of numerous characters, some of which are difficult to observe since they generally display a considerable amount of intraspecific variation (e.g., ontogenetic variation and sexual dimorphism) and are not always considered in taxonomic keys (primarily based on adult males) (e.g., [ 28 , 30 – 32 ]). In addition, amphipods are often neglected or mistakenly identified due to their small size and the morphological uniformity among some closely related species [ 25 , 31 , 33 , 34 ]. Thus, an accurate morphological identification frequently requires experienced taxonomists, most of whom are nearing retirement, and who are becoming increasingly rare due to inadequate funding and the subsequent low recruitment of young scientists into this discipline [26,33,35]. Today, the use of morphological and molecular methods (e.g., DNA barcoding [ 36 ]), in the form of integrative taxonomy, has been suggested as the most effective approach for biodiversity assessments, for which molecular data provide support to accurately identify those taxa not amenable to morphological identification [ 37 , 38 ]. In amphipods, the use of this approach has facilitated the classification and delimitation of specimens with a more complex or hardly accessible morphology (e.g., immature or damaged specimens), allowing also the detection of putative cryptic species and species complexes within widely distributed taxa [25,29,30,39–41]. Caprella penantis Leach, 1814, is considered one of the most taxonomically challenging caprellid amphipods in the world since it has been recorded under several species or subspecies names [ 3 , 42 – 45 ]. Therefore, the need for molecular studies to determine whether it is a species complex or a single species with high intraspecific morphological variability has been highlighted since its first taxonomic studies [ 3 , 43 ]. In his monographs, Mayer [44,45] described 19 forms of the “Caprella acutifrons” group (typica,minor,tabida,tibada,neglecta, gibbosa,andreae,carolinensis,virginia,lusitanica,natalensis,porcellio,simulatrix,testudo,angusta,incisa,verrucosa,borealis, and cristibrachium). Several of these forms have already been given specific ranks [ 43 , 46 – 48 ]. For example, the last four forms were assigned to Caprella incisa Mayer, 1903, C. verrucosa,C. borealis Mayer, 1903, and C. cristibrachium Mayer, 1903, respectively [ 3 , 43 , 46 , 47 ]; form andreae was considered a distinct and valid species, Caprella andreae Mayer, 1890, confirmed by both morphological [ 3 ] and molecular analyses [ 30 ]. Furthermore, a recent study based on mitochondrial and nuclear markers indicated that individuals belonging to the form gibbosa described from Coquimbo by [ 44 , 45 ] could also merit specific rank [ 29 ]. Currently, only the forms carolinensis,virginia,lusitanica,testudo, and simulatrix are still classified as the nominal species C. penantis sensu lato [ 29 , 43 , 49 ], with only the last three being present in the eastern Atlantic and western Mediterranean region [ 29 ] and considered by these authors as C. penantis sensu stricto (hereafter C. penantis s.s.). Interestingly, based on molecular analyses, these authors suggest that the presence or absence of a proximal projection on the propodus of the second gnathopods, which differentiates forms testudo and lusitanica from simulatrix, may correspond to intraspecific variation, and thus, the form simulatrix should not be considered as a valid one [ 29 ]. This molecular study also revealed the existence of three highly divergent lineages in this species (C. penantis s.s.), suggesting the occurrence of a process of ongoing speciation and, thus, casting doubt on the taxonomic status of this marine species. Furthermore, the ultimate clarification of the taxonomical status of C. penantis s.s. was not possible because samples from the type locality were not available. In this study, we used an integrative approach, combining morphological data and molecular species delimitation methods, to continue to shed light on the taxonomy and distribution of C. penantis s.s. To this end, we included for the first-time populations from the type locality of this species and four populations from its distributional range not previously sequenced. These data were integrated with publicly available sequences and morphological information to (1) confirm whether this caprellid is a complex of distinct species and if so, (2) examine their phylogenetic relationships; (3) determine which of them
Life 2022,12, 155 3 of 17 could be considered the true C. penantis; (4) investigate the underlying processes that lead to genetic divergence in this small marine amphipod. 2. Materials and Methods 2.1. Sample Collection and Taxonomic Identification Between 2010 and 2018, a total of 75 specimens of the putative species C. penantis s.s. were by-catch in small samples of intertidal macroalgae at seven localities along the northeast Atlantic coast, including the species’ type locality (Devonshire coast) (Table 1, Figure 1). Most of the samples, except those from Viana do Castelo, were obtained as the result of a series of sampling programs conducted by different members of our research team (see Acknowledgements). At each locality, caprellids were removed by hand and immediately preserved in 96% ethanol. Sampling information, including collection localities, coordinates, and the number of individuals collected, is given in Table 1. Life2022,11,xFORPEERREVIEW3of17 thetypelocalityofthisspeciesandfourpopulationsfromitsdistributionalrangenot previouslysequenced.Thesedatawereintegratedwithpubliclyavailablesequencesand morphologicalinformationto(1)confirmwhetherthiscaprellidisacomplexofdistinct speciesandifso,(2)examinetheirphylogeneticrelationships;(3)determinewhichof themcouldbeconsideredthetrueC.penantis;(4)investigatetheunderlyingprocessesthat leadtogeneticdivergenceinthissmallmarineamphipod. 2.MaterialsandMethods 2.1.SampleCollectionandTaxonomicIdentification Between2010and2018,atotalof75specimensoftheputativespeciesC.penantiss.s. wereby‐catchinsmallsamplesofintertidalmacroalgaeatsevenlocalitiesalongthenortheastAtlanticcoast,includingthespecies’typelocality(Devonshirecoast)(Table1,Figure 1).Mostofthesamples,exceptthosefromVianadoCastelo,wereobtainedastheresult ofaseriesofsamplingprogramsconductedbydifferentmembersofourresearchteam (seeAcknowledgements).Ateachlocality,caprellidswereremovedbyhandand immediatelypreservedin96%ethanol.Samplinginformation,includingcollection localities,coordinates,andthenumberofindividualscollected,isgiveninTable1. Figure1.GeographicdistributionoftheDNAsequencesincludedinthepresentstudy.Stars representlocalitieswhereCaprellapenantissensustrictowascollectedandsequencedforthefirsttime (thiswork),andthoseindicatedbycirclescorrespondtopreviouslysequencedones(sequences availableinGenBank).SeeTables1andS1forlocalitiescodes.Eachlocalityiscolourcoded accordingtothemitochondrialcladetowhichtheybelong:blue—CladeVA,red—CladeVB,and green—CladeVC.Theblackdashedlinescorrespondtothetwophylogeographicbreaksdetected inthepresentstudy. Figure 1. Geographic distribution of the DNA sequences included in the present study. Stars represent localities where Caprella penantis sensu stricto was collected and sequenced for the first time (this work), and those indicated by circles correspond to previously sequenced ones (sequences available in GenBank). See Tables 1and S1 for localities codes. Each locality is colour coded according to the mitochondrial clade to which they belong: blue—Clade VA, red—Clade VB, and green—Clade VC. The black dashed lines correspond to the two phylogeographic breaks detected in the present study. In the laboratory, all specimens were observed under a standard binocular microscope and identified to species level based on the morphological descriptions provided by [3,44,45,50,51]. Special attention was paid to the presence of a short and triangular rostrum on the front of the head, the shape and size of the gills, the robustness of the first antenna, the characteristics of second gnathopods in adult males, including the presence or absence of a proximal projection, and the concavity or convexity of the propodus of pereopods 5–7.
Life 2022,12, 155 4 of 17 Table 1. List of sampling locations where Caprella penantis sensu stricto was collected, including codes, source countries, geographical coordinates, collection data, and number of specimens sequenced (N) for each gene (COI/18S). Genetic information regarding the number of haplotypes (H) and respective haplotype codes, the phylogenetic clade each locality belongs to, and the GenBank accession numbers for both sequenced genes are also included. Code Locality Country Coordinates Collection Data N H Haplotype Codes Phylogenetic Clade (COI) GenBank acc. nos. COI/18S 1Sidmouth, Devonshire coast UK 50 ◦ 40 0 29.881 00 N, 3◦14044.99200 W14/08/2018 14 /3 6/1 H17-H22/H2 VB OM057907 − OM057920/OM112220 − OM112222 2 St. Ives UK 50◦1309.68500 N, 5◦28042.21100 W13/08/2018 17/2 14/1 H3-H16/H2 VB OM057890 − OM057906/OM112218 − OM112219 3Wembury, Devonshire coast UK 50 ◦ 18 0 57.287 00 N, 4◦4059.12400 W20/06/2010 1 1 H2 VB OM057889 4 Temara Morocco 33 ◦ 54 0 50.904 00 N, 6◦58049.57700 W31/01/2010 2/1 1/1 H1/H1 VA OM057887−OM057888/OM112217 5 Alteirinhos Portugal 37◦3109.8900 N, 8◦47018.94900 W05/06/2012 12 2 H23-H24 VA OM057921−OM057932 6Monte da Guia, Azores Portugal 38◦3109.10200 N, 28 ◦ 37 0 40.778 00 W 02/12/2014 15/5 13/2 H25-H37/H2 VB OM057933 − OM057947/OM112223 − OM112227 7 Viana do Castelo Portugal 41◦41059.5900 N, 8◦51024.55900 W14/07/2014 14/13 8/1 H38-H45/H2 VA, VB OM057948 − OM057961/OM112228 − OM112240
Life 2022,12, 155 5 of 17 2.2. DNA Extraction, PCR Amplification, and Sequencing Genomic DNA was extracted from caprellid appendages along one side of the body of each specimen, except when individuals were too small (in those cases, the whole specimen was used). We employed the Purelink Genomic DNA Mini Kit (Invitrogen, Paisley, UK) according to the manufacturer’s protocol. A fragment (~658 bp) of the mitochondrial (mt) cytochrome c oxidase subunit I (COI) gene was amplified by polymerase chain reaction (PCR) using the primer pairs jgLCO1490/jgHCO2198 [ 52 ]. PCR amplifications consisted of 25 µ L reaction volumes containing 3 µ L of template DNA, 10x buffer MgCl 2 free (Invitrogen, UK), 3 mM MgCl 2 , 0.2 mM dNTPs, 1 µ M of each primer, 0.3 U Platinum Taq DNA polymerase (Invitrogen, UK) and double-distilled H 2 O to volume. PCR conditions used were as described in [ 53 ]. Furthermore, a 1051 bp fragment of the 18S ribosomal RNA (18S rRNA) nuclear gene was obtained for a subset of 24 individuals (Table 1), using the primers 18S-ai and 18S-bi [ 54 ] and following the PCR conditions specified by [29]. Positive PCR products were purified and bidirectionally sequenced by a commercial company (GENEWIZ, Leipzig, Germany). 2.3. Sequence Analysis The resulting sequences were edited with Sequencher v5.4.6 (Gene Codes Corporation, Ann Arbor, MI, USA) and checked for potential contaminations using GenBank’s BLASTn search [55]. They were thereafter deposited in GenBank (Table 1). To confirm the morphological identification of the sequenced specimens and to shed light on the taxonomy of C. penantis s.s., a total of 79 COI and 15 18S available sequences from this species in GenBank were included in the final data set (Table S1, Figure 1). Furthermore, sequences of closely related species C. dilatata Krøyer, 1843 (24 COI; 2 18S) and C. andreae Mayer, 1890 (45 COI; 6 18S) were also included (Table S1). For the COI gene, all sequences were aligned using MUSCLE [ 56 ], as implemented in MEGA X [ 57 ]. They were also checked for the presence of pseudogenes by translating them into amino acids. On the other hand, 18S rRNA sequences were aligned using MAFFT [ 58 ], and highly variable regions and poorly aligned positions were eliminated from the analysis using Gblocks [59], with default parameters and allowing all gap positions. 2.4. Phylogenetic Reconstruction Analyses were performed using data partitions by codon (1 + 2 +3) for the mitochondrial COI gene, to minimise the saturation effects of codon positions on phylogenetic reconstructions [ 60 ] and to account for different rates of evolution of each one [ 61 ]. For both genes, identical sequences were collapsed into haplotypes to reduce redundancy and facilitate the computational processes, and the species C. simia Mayer, 1903 (accession numbers: COI—KF743437, 18S—KF743485) and C. linearis (Linnaeus, 1767) (accession numbers: COI—FJ581572, 18S—DQ378039) were used as outgroups. Phylogenetic tree reconstructions were performed using two model-based methods of phylogenetic inference: maximum likelihood (ML) in Garli v2.01 [ 62 ] and Bayesian inference (BI) in MrBayes v3.2.6 [ 63 ]. The corrected Akaike Information Criterion (AICc) [ 64 ] implemented in PartitionFinder v2.1.1 [ 20 ] was used to determine the best-fit substitution model for each data set. The resulting models were GTR + I + G (1st position), F81 + I (2nd position), and GTR + G (3rd position) for the COI data set, and K80 + I for the 18S one. ML analysis was performed using 10 independent searchers and 1000 bootstrap replicates. The evaluation of log-likelihood values across searchers allowed checking the convergence between the topologies of the trees generated. The SumTrees command from the Dendropy package [ 65 ] was used to summarise non-parametric bootstrap support (BS) values for the best tree, after generating a majority-rule consensus tree. For the BI analysis, four Markov chain Monte Carlo (MCMC) chains were run twice in parallel for 2×107generations, sampling trees, and parameters every 1000 generations, with the heating parameter set to 0.25. Convergence of the analyses was considered to be reached when the average
Life 2022,12, 155 6 of 17 standard deviation of split frequencies was less than 0.01 and was validated by plotting the log-likelihood values against generation times in Tracer v1.7.1 [ 66 ]. The majority-rule consensus tree was estimated by combining results from duplicated analyses, after discarding 25% of samples as burn-in. The consensus tree inferred for each molecular data set was visualised and rooted using FigTree v1.4.4 [ 67 ] and later edited with Inkscape v1.1 software (https://www.inkscape.org, accessed on 18 December 2021). Clades with BS or BI posterior probability (BPP) greater than 70% or 0.7, respectively, were considered well supported. Furthermore, uncorrected p-distances were calculated in MEGA X [ 57 ] to estimate genetic divergences between the phylogenetic clades obtained from reconstructed trees. Relationships between mitochondrial haplotypes were further examined by building an unrooted network in TCS v1.21 [ 68 ] following the statistical parsimony criterion [ 69 ], with a 95% connection limit. The network was plotted with tcsBU [70]. 2.5. Molecular Species Delimitation To evaluate the clades resulting from the phylogenetic tree reconstructions, species delimitation analyses were performed using three different approaches: two distance-based methods, the Barcode Index Number (BIN) system [ 71 ] and the Assemble Species by Automatic Partitioning (ASAP) method [ 72 ], and one tree-based method, the Bayesian Poisson Tree Process (bPTP) model [ 73 ]. All methods were applied to both COI and 18S data sets, except for the BIN system, which is based only on COI. Using BINs implemented in the Barcode of Life System (BOLD) [ 74 ], COI sequences were clustered into molecular operational taxonomic units (MOTUs) independently on their predefined taxonomic classification. Each cluster was assigned to a unique alphanumeric identifier or BIN that closely corresponds to a specific species. For the ASAP method, we applied the p-distances model implemented on the ASAP Web (https://bioinfo.mnhn.fr/abi/public/asap/asapweb.html, accessed on 19 November 2021). By building partitions from single-locus sequence alignments, it provides a score for each defined partition and sorts the sequences into putative species [ 72 ]. Finally, the bPTP model was performed on the bPTP server (https://species.h-its.org/ptp/, accessed on 19 November 2021) using the Bayesian tree as input, running 100,000 MCMC generations, and with the burn-in set to 25%. Unlike BIN and ASAP, bPTP infers putative species based on a non-ultrametric phylogenetic tree by identifying the transition points between inter- and intraspecific branching events [73]. 2.6. Genetic Structure Analyses To test the genetic differentiation between locations, and between regions, Fst estimations were calculated in Arlequin v3.5.1.2 [ 75 ]. For the latter, populations were grouped according to the geographical region to which they belong: continental Portugal (Alteirinhos, Viana do Castelo, Mindelo, Labruge, Castelejo, and Praia Azul) and Strait of Gibraltar (Torreguadiaro, Tarifa Island, Puerto de Ceuta, Grottes d’Hercules, Ksar-es Seghir, Estepona, El Chorrillo, Punta Carnero, Benzú, and Punta Almina) for Clade VA; UK (Sidmouth and St. Ives), northern Spain (Baleo, Cetarea, and Oyambre), the Azores (Monte da Guia), and continental Portugal (Viana do Castelo) for Clade VB. Locations with less than three individuals were excluded from the analyses. The significance of pairwise Fst values was determined by performing 10,000 permutations, under the null hypothesis of no differentiation. In addition, an analysis of molecular variance (AMOVA) [ 76 ] was performed to examine the overall genetic structure within each clade, using pairwise differences as a distance measure in Arlequin v3.5.1.2 [75]. Finally, the potential effect of isolation by distance was tested on the mitochondrial dataset applying the Mantel test to genetic and geographical distances matrices in IBD v1.52 [ 77 ] using 10,000 randomisations. Genetic distances, corresponding to pairwise Fst values, were estimated in Arlequin v3.5.1.2 [ 76 ] through 10,000 permutations, and geographic distances were assessed through the GeoDataSource platform (https://www. geodatasource.com/demo#latlong, accessed on 9 January 2022).
Life 2022,12, 155 7 of 17 3. Results 3.1. Morphological Analysis All caprellids collected were confirmed as C. penantis s.s. mainly based on the presence of a short and triangular rostrum on the front of the head, peduncle of antenna 1 slender, small, and elongated gills, and pereopods 5–7 propodi palm slightly concave (Figure 2) [3,50,51]. However, some small morphological differences were found among specimens from different populations (Figure 2, Table S2). Individuals collected from the UK and the Azores were characterised by smaller and elongated gills and more elongated second gnathopods, with abundant setae and a larger proximal projection (Figure 2a,b,d), while those from continental Portugal and Morocco showed rounder and larger gills and rounder second gnathopods (Figure 2c,e). Interestingly, individuals from Viana do Castelo (Figure 2e) depicted intermediate characteristics, with more rounded second gnathopods, similar to those from Portugal and Morocco, but densely setose, similar to those from the UK and Azores populations. We were unable to include any photos of specimens from Wembury (UK) and Temara (Morocco), due to their degraded state. Life2022,11,xFORPEERREVIEW7of17 gnathopods,similartothosefromPortugalandMorocco,butdenselysetose,similarto thosefromtheUKandAzorespopulations.Wewereunabletoincludeanyphotosof specimensfromWembury(UK)andTemara(Morocco),duetotheirdegradedstate. Figure2.Adultmale’sphotographsofCaprellapenantissensustrictofromthelocalitiessequencedin thisstudy:(a)Sidmouth(UK);(b)St.Ives(UK);(c)Alteirinhos(continentalPortugal);(d)Monteda Guia(Azores,Portugal);(e)VianadoCastelo(continentalPortugal).Thesmallerpictures(insidea whitesquare)representthesecondgnathopods.Scalebars:1mm. 3.2.PhylogeneticInferenceandSpeciesDelimitation Afterqualityfiltering,thelengthofthefinalalignmentwas537bpand926bpfor COIand18S,respectively.TheCOIdatasetincludedatotalof225sequences,75fromthis studyand150availablefromGenBank(Tables1andS1),whereasatotalof49sequences comprisedthe18Sdataset,24ofthemsequencedforthefirsttimeinthepresentstudy (Table1).Nostopcodons,insertions,ordeletionswereobserved. PhylogeneticanalysesoftheCOIdatasetbasedonBIandMLapproaches,recovered treeswithverysimilartopologies,differingonlyinthepositionofafewhaplotypeswithin innergroups(Figures3andS1).Bothanalyseshighlysupportedthemonophylyof Caprellapenantiss.s.andtheexistenceofthreemajordistinctandgeographicallyseparated mtDNAcladeswithinthisspecies(Figure3),henceforthnamedasCladeV A ,V B ,andV C , asthisnomenclaturewasalreadyestablishedinthestudyof[29].CladeV A includedall haplotypesoccurringintheStraitofGibraltarandTemara(AtlanticMorocco)populations,andmosthaplotypesfromcontinentalPortugal(Figure3,Tables1andS1).Clade V B groupedallhaplotypesfromtheUK(includingthespecies’typelocality),northern Figure 2. Adult male’s photographs of Caprella penantis sensu stricto from the localities sequenced in this study: ( a ) Sidmouth (UK); ( b ) St. Ives (UK); ( c ) Alteirinhos (continental Portugal); ( d ) Monte da Guia (Azores, Portugal); ( e ) Viana do Castelo (continental Portugal). The smaller pictures (inside a white square) represent the second gnathopods. Scale bars: 1 mm. 3.2. Phylogenetic Inference and Species Delimitation After quality filtering, the length of the final alignment was 537 bp and 926 bp for COI and 18S, respectively. The COI data set included a total of 225 sequences, 75 from this study and 150 available from GenBank (Tables 1and S1), whereas a total of 49 sequences
Life 2022,12, 155 8 of 17 comprised the 18S data set, 24 of them sequenced for the first time in the present study (Table 1). No stop codons, insertions, or deletions were observed. Phylogenetic analyses of the COI data set based on BI and ML approaches, recovered trees with very similar topologies, differing only in the position of a few haplotypes within inner groups (Figure 3and Figure S1). Both analyses highly supported the monophyly of Caprella penantis s.s. and the existence of three major distinct and geographically separated mtDNA clades within this species (Figure 3), henceforth named as Clade VA, VB, and VC, as this nomenclature was already established in the study of [ 29 ]. Clade VA included all haplotypes occurring in the Strait of Gibraltar and Temara (Atlantic Morocco) populations, and most haplotypes from continental Portugal (Figure 3, Tables 1and S1). Clade VB grouped all haplotypes from the UK (including the species’ type locality), northern Spain, and the Azores archipelago (Figure 3, Tables 1and S1). Interestingly, this clade also comprised two haplotypes (H40 and H41) observed in the Viana do Castelo population (continental Portugal). Finally, in the Clade VC, only the haplotypes present in Safi (Atlantic Morocco) were included (Figure 3, Table S1). Genetic divergence (uncorrected p-distances) between these clades ranged from 10.2% (Clades VA and VB) to 10.6% (Clades VB and VC) (Table 2), which is similar to that found between the closely related species C. dilatata and C. andreae (11.1%). The divergence between C. penantis clades and the congeneric species included in the analyses were slightly higher, ranging from 13.2 to 22.6%, with the lowest value found between Clade VC and C. dilatata, and the highest between Clade VA and C. linearis (Table 2). Life2022,11,xFORPEERREVIEW8of17 Spain,andtheAzoresarchipelago(Figure3,Tables1andS1).Interestingly,thiscladealso comprisedtwohaplotypes(H40andH41)observedintheVianadoCastelopopulation (continentalPortugal).Finally,intheCladeV C, onlythehaplotypespresentinSafi(AtlanticMorocco)wereincluded(Figure3,TableS1).Geneticdivergence(uncorrectedpdistances)betweenthesecladesrangedfrom10.2%(CladesV A andV B )to10.6%(Clades V B andV C ) (Table2),whichissimilartothatfoundbetweenthecloselyrelatedspeciesC. dilatataandC.andreae(11.1%).ThedivergencebetweenC.penantiscladesandthecongenericspeciesincludedintheanalyseswereslightlyhigher,rangingfrom13.2to22.6%, withthelowestvaluefoundbetweenCladeV C andC.dilatata,andthehighestbetween CladeV A andC.linearis(Table2). Figure3.BayesianconsensustreeofCaprellapenantissensustrictoandrelatedtaxabasedonmitochondrialCOIsequences.Bayesianposteriorprobabilities(BPP)over0.8arerepresentedbyred circlesatnodesandvaluescorrespondtobootstrapsupport(above50)givenbythemaximumlikelihoodanalyses.CladesVA–VCareidentified.ThetreewasrootedwithC.simiaandC.linearis(sequencesavailableinGenBank:KF743437andFJ581572,respectively).Thetwohaplotypes(H40and H41)fromVianadoCastelogroupedinCladeVBarehighlightedinbold.Colourbarsontheright representresultsfromthespeciesdelimitationanalysesusingBarcodeIndexNumber(BIN),AssembleSpeciesbyAutomaticPartitioning(ASAP),andBayesianPoissonTreeProcess(bPTP)models. Table2.Averagep‐distancevalues(%)betweenmajormitochondrialCOI(belowdiagonal)and nuclear18S(abovediagonal)cladesidentifiedinthephylogeneticanalyses. CladeVACladeVBCladeVCC.dilatataC.andreaeC.simiaC.linearis CladeVA ‐‐ ‐‐ ‐‐ ‐‐ ‐‐ ‐‐ CladeVB10.2 0.40.91.11.61.3 CladeVC10.310.6 0.81.01.51.2 C.dilatata13.413.413.2 1.31.51.4 C.andreae15.014.314.211.12.31.6 C.simia19.219.419.617.517.00.9 C.linearis22.621.321.119.820.919 Figure 3. Bayesian consensus tree of Caprella penantis sensu stricto and related taxa based on mitochondrial COI sequences. Bayesian posterior probabilities (BPP) over 0.8 are represented by red circles at nodes and values correspond to bootstrap support (above 50) given by the maximum likelihood analyses. Clades VA–VC are identified. The tree was rooted with C. simia and C. linearis (sequences available in GenBank: KF743437 and FJ581572, respectively). The two haplotypes (H40 and H41) from Viana do Castelo grouped in Clade VB are highlighted in bold. Colour bars on the right represent results from the species delimitation analyses using Barcode Index Number (BIN), Assemble Species by Automatic Partitioning (ASAP), and Bayesian Poisson Tree Process (bPTP) models.
Life 2022,12, 155 9 of 17 Table 2. Average p-distance values (%) between major mitochondrial COI (below diagonal) and nuclear 18S (above diagonal) clades identified in the phylogenetic analyses. Clade VA Clade VB Clade VC C. dilatata C. andreae C. simia C. linearis Clade VA – – – – – – Clade VB 10.2 0.4 0.9 1.1 1.6 1.3 Clade VC 10.3 10.6 0.8 1.0 1.5 1.2 C. dilatata 13.4 13.4 13.2 1.3 1.5 1.4 C. andreae 15.0 14.3 14.2 11.1 2.3 1.6 C. simia 19.2 19.4 19.6 17.5 17.0 0.9 C. linearis 22.6 21.3 21.1 19.8 20.9 19 For the 18S gene, the ML and BI analyses rendered trees with identical topologies (Figure 4). Unlike the COI gene, only two main clades were identified within C. penantis s.s.—a first clade containing the haplotypes from the Safi population (Clade VC), and a second clade where the remaining haplotypes were grouped (Clade VA + VB) (Figure 4, Tables 1and S1). According to these analyses, the Clade VC is sister to the species C. andreae, thus indicating that C. penantis appears to be paraphyletic, although it showed little bootstrap support (Figure 4). Within C. penantis, genetic divergence between the two clades was 0.4% (Table 2). The divergence between this species and the other species of the genus ranged from 0.8% (Clade VC and C. dilatata) to 1.6% (Clade VB and C. simia) (Table 2). The interspecific divergence between Caprella species, excluding C. penantis, was in the range 0.9–2.3%, with the lowest divergence found between C. linearis and C. simia and the highest between the latter and C. andreae (Table 2). Life2022,11,xFORPEERREVIEW9of17 Forthe18Sgene,theMLandBIanalysesrenderedtreeswithidenticaltopologies (Figure4).UnliketheCOIgene,onlytwomaincladeswereidentifiedwithinC.penantis s.s.—afirstcladecontainingthehaplotypesfromtheSafipopulation(CladeV C ),anda secondcladewheretheremaininghaplotypesweregrouped(CladeV A +V B )(Figure4, Tables1andS1).Accordingtotheseanalyses,theCladeV C issistertothespeciesC.andreae,thusindicatingthatC.penantisappearstobeparaphyletic,althoughitshowedlittle bootstrapsupport(Figure4).WithinC.penantis,geneticdivergencebetweenthetwo cladeswas0.4%(Table2).Thedivergencebetweenthisspeciesandtheotherspeciesof thegenusrangedfrom0.8%(CladeV C andC.dilatata)to1.6%(CladeV B andC.simia) (Table2).TheinterspecificdivergencebetweenCaprellaspecies,excludingC.penantis,was intherange0.9–2.3%,withthelowestdivergencefoundbetweenC.linearisandC.simia andthehighestbetweenthelatterandC.andreae(Table2). Figure4.BayesianconsensustreeofCaprellapenantissensustrictoandrelatedtaxabasedonnuclear 18Ssequences.Bayesianposteriorprobabilities(BPP)over0.8arerepresentedbyredcirclesatnodes andvaluescorrespondtobootstrapsupport(above50)givenbythemaximumlikelihoodanalyses. CladesVA–VCareidentified.ThetreewasrootedwithC.simiaandC.linearis(sequencesavailable inGenBank:KF743485andDQ378039,respectively).Colourbarsontherightrepresentresultsfrom thespeciesdelimitationanalyses:AssembleSpeciesbyAutomaticPartitioning(ASAP)andBayesianPoissonTreeProcess(bPTP)models. Finally,speciesdelimitationanalyses(BIN,ASAP,andbPTP)clusteredtheCOIsequencesofC.penantiss.s.inthreedistinctMOTUS(Figure3),whilefor18S,atotaloftwo MOTUSwereobserved(Figure4).Theseresultsagreewiththosefromphylogeneticanalysesand,therefore,stronglysupporttheexistenceofatleasttwopotentiallydistinctspecieswithinC.penantiss.s. 3.3.PhylogeographicStructure ThemtDNACOInetworkforC.penantiss.s.retrievedthreeseparatenetworksthat couldnotbeconnectedusingthe95%parsimonyconnectionlimit(Figure5).Thus,supportingtheexistenceofthreedivergentgroupswithinthisspecies,whichmatchedthe distinctcladesobservedinthephylogeneticanalyses(Figure3).Atotalof86haplotypes wasobservedforthe154individualsconsidered:35inCladeV A ,49inCladeV B, and2in CladeV C .Fromthesehaplotypes,79wererestrictedtoasinglelocation,andonly7were Figure 4. Bayesian consensus tree of Caprella penantis sensu stricto and related taxa based on nuclear 18S sequences. Bayesian posterior probabilities (BPP) over 0.8 are represented by red circles at nodes and values correspond to bootstrap support (above 50) given by the maximum likelihood analyses. Clades VA–VC are identified. The tree was rooted with C. simia and C. linearis (sequences available in GenBank: KF743485 and DQ378039, respectively). Colour bars on the right represent results from the species delimitation analyses: Assemble Species by Automatic Partitioning (ASAP) and Bayesian Poisson Tree Process (bPTP) models.
Life 2022,12, 155 16 of 17 65. Sukumaran, J.; Holder, M.T. DendroPy: A Python library for phylogenetic computing. Bioinformatics 2010 ,26, 1569–1571. [CrossRef] 66. Rambaut, A.; Drummond, A.J.; Xie, D.; Baele, G.; Suchard, M.A. Posterior summarization in Bayesian phylogenetics using Tracer 1.7. Syst. Biol. 2018,67, 901–904. [CrossRef] 67. Rambaut, A. FigTree-Version 1.4.4, a Graphical Viewer of Phylogenetic Trees. 2018. Available online: https://github.com/ rambaut/figtree/releases (accessed on 20 November 2021). 68. Clement, M.; Posada, D.; Crandall, K.A. TCS: A computer program to estimate gene genealogies. Mol. Ecol. 2000 ,9, 1657–1659. [CrossRef] [PubMed] 69. Templeton, A.R.; Crandall, K.A.; Sing, C.F. A cladistic analysis of phenotypic associations with haplotypes inferred from restriction endonuclease mapping and DNA sequence data. III. Cladogram estimation. Genetics 1992,132, 619–633. [CrossRef] 70. Santos, A.M.; Cabezas, M.P.; Tavares, A.I.; Xavier, R.; Branco, M. TcsBU: A Tool to extend TCS network layout and visualization. Bioinformatics 2016,32, 627–628. [CrossRef] 71. Ratnasingham, S.; Hebert, P.D.N. A DNA-based registry for all animal species: The Barcode Index Number (BIN) System. PLoS ONE 2013,8, e66213. [CrossRef] [PubMed] 72. Puillandre, N.; Brouillet, S.; Achaz, G. ASAP: Assemble Species by Automatic Partitioning. Mol. Ecol. Resour. 2021 ,21, 609–620. [CrossRef] [PubMed] 73. Zhang, J.; Kapli, P.; Pavlidis, P.; Stamatakis, A. A general species delimitation method with applications to phylogenetic placements. Bioinformatics 2013,29, 2869–2876. [CrossRef] 74. Ratnasingham, S.; Hebert, P.D.N. BOLD: The Barcode of Life Data System (Http://Www.Barcodinglife.Org). Mol. Ecol. Notes 2007,7, 355–364. [CrossRef] 75. Excoffier, L.; Lischer, H.E.L. Arlequin Suite Ver 3.5: A new series of programs to perform population genetics analyses under Linux and Windows. Mol. Ecol. Resour. 2010,10, 564–567. [CrossRef] [PubMed] 76. Excoffier, L.; Smouse, P.E.; Quattro, J.M. Analysis of Molecular Variance inferred from metric distances among DNA haplotypes: Application to human mitochondrial DNA restriction data. Genetics 1992,131, 479–491. [CrossRef] [PubMed] 77. Bohonak, A.J. IBD (Isolation by Distance): A program for analyses of isolation by distance. J. Hered. 2002 ,93, 153–154. [CrossRef] [PubMed] 78. Horton, T.; Lowry, J.; De Broyer, C.; Bellan-Santini, D.; Coleman, C.O.; Corbari, L.; Costello, M.J.; Daneliya, M.; Dauvin, J.-C.; Fišer, C.; et al. WoRMS—World Register of Marine Species. Caprella Lamarck 1801. Available online: https://www.marinespecies.org/ aphia.php?p=taxdetails&id=101430 (accessed on 11 December 2021). 79. Guerra-García, J.M.; Tierno de Figueroa, J.M. What do caprellids (Crustacea: Amphipoda) feed on? Mar. Biol. 2009 ,156, 1881–1890. [CrossRef] 80. Cabezas, M.P. Taxonomía Molecular y Biogeografía de los Caprélidos (Crustacea: Amphipoda) del Estrecho de Gibraltar. Ph.D. Thesis, University of Seville, Seville, Spain, October 2013. 81. Cabezas, M.P.; Xavier, R.; Branco, M.; Santos, A.M.; Guerra-García, J.M. Invasion history of Caprella scaura Templeton, 1836 (Amphipoda: Caprellidae) in the Iberian Peninsula: Multiple introductions revealed by mitochondrial sequence data. Biol. Invasions 2014,16, 2221–2245. [CrossRef] 82. Baird, H.P.; Miller, K.J.; Stark, J.S. Evidence of hidden biodiversity, ongoing speciation and diverse patterns of genetic structure in giant Antarctic amphipods. Mol. Ecol. 2011,20, 3439–3454. [CrossRef] 83. Witt, J.D.S.; Threloff, D.L.; Hebert, P.D.N. DNA barcoding reveals extraordinary cryptic diversity in an amphipod genus: Implications for desert spring conservation. Mol. Ecol. 2006,15, 3073–3082. [CrossRef] 84. Mohrbeck, I.; Horton, T.; Ja˙ zd˙ zewska, A.M.; Martínez Arbizu, P. DNA barcoding and cryptic diversity of deep-sea scavenging amphipods in the Clarion-Clipperton zone (Eastern Equatorial Pacific). Mar. Biodivers. 2021,51, 26. [CrossRef] 85. Vieira, P.E.; Desiderato, A.; Holdich, D.M.; Soares, P.; Creer, S.; Carvalho, G.R.; Costa, F.O.; Queiroga, H. Deep segregation in the open ocean: Macaronesia as an evolutionary hotspot for low dispersal marine invertebrates. Mol. Ecol. 2019 ,28, 1784–1800. [CrossRef] [PubMed] 86. Barrowclough, G.F.; Zink, R.M. Funds enough, and time: mtDNA, nuDNA and the discovery of divergence. Mol. Ecol. 2009 ,18, 2934–2936. [CrossRef] 87. Maggs, C.A.; Castilho, R.; Foltz, D.; Henzler, C.; Jolly, M.T.; Kelly, J.; Olsen, J.; Perez, K.E.; Stam, W.; Väinölä, R.; et al. Evaluating signatures of glacial refugia for North Atlantic benthic marine taxa. Ecology 2008,89, S108–S122. [CrossRef] 88. Palumbi, S.R. Marine speciation on a small planet. Trends Ecol. Evol. 1992,7, 114–118. [CrossRef] 89. Patarnello, T.; Volckaert, F.A.M.J.; Castilho, R. Pillars of Hercules: Is the Atlantic-Mediterranean transition a phylogeographical break? Mol. Ecol. 2007,16, 4426–4444. [CrossRef] 90. Xavier, R.; Zenboudji, S.; Lima, F.P.; Harris, D.J.; Santos, A.M.; Branco, M. Phylogeography of the marine isopod Stenosoma nadejda (Rezig, 1989) in north African Atlantic and western Mediterranean coasts reveals complex differentiation patterns and a new Species. Biol. J. Linn. Soc. 2011,104, 419–431. [CrossRef] 91. Desiderato, A.; Costa, F.O.; Serejo, C.S.; Abbiati, M.; Queiroga, H.; Vieira, P.E. Macaronesian islands as promoters of diversification in amphipods: The remarkable case of the Family Hyalidae (Crustacea, Amphipoda). Zool. Scr. 2019,48, 359–375. [CrossRef]
Life 2022,12, 155 17 of 17 92. Vieira, P.E.; Desiderato, A.; Azevedo, S.L.; Esquete, P.; Costa, F.O.; Queiroga, H. Molecular evidence for sweeping discontinuity between peracarid (Crustacea) fauna of Macaronesian islands and nearby continental coasts: Over fifty candidate endemic species. bioRxiv 2021. [CrossRef] 93. Virgilio, M.; Fauvelot, C.; Costantini, F.; Abbiati, M.; Backeljau, T. Phylogeography of the common ragworm Hediste diversicolor (Polychaeta: Nereididae) reveals cryptic diversity and multiple colonization events across its distribution. Mol. Ecol. 2009 ,18, 1980–1994. [CrossRef] [PubMed] 94. Deli´c, T.; Stoch, F.; Borko, Š.; Flot, J.-F.; Fišer, C. How did subterranean amphipods cross the Adriatic Sea? Phylogenetic evidence for dispersal–vicariance interplay mediated by marine regression–transgression cycles. J. Biogeogr. 2020 ,47, 1875–1887. [CrossRef] 95. Gómez-Gesteira, M.; deCastro, M.; Alvarez, I.; Gómez-Gesteira, J.L. Coastal sea surface temperature warming trend along the continental part of the Atlantic Arc (1985–2005). J. Geophys. Res. Ocean. 2008,113, 1–9. [CrossRef] 96. Savolainen, O.; Lascoux, M.; Merilä, J. Ecological genomics of local adaptation. Nat. Rev. Genet. 2013 ,14, 807–820. [CrossRef] [PubMed] 97. Gittenberger, E. What about non-adaptive radiation? Biol. J. Linn. Soc. 1991,43, 263–272. [CrossRef] 98. Mathews, L.M. Cryptic biodiversity and phylogeographical patterns in a snapping shrimp species complex. Mol. Ecol. 2006,15, 4049–4063. [CrossRef] [PubMed]