Annelida, Polychaeta, Chaetopteridae

Zoological Journal of the Linnean Society, 2008, 152, 201–225. With 12 ﬁgures

Description of a new species of Mesochaetopterus (Annelida, Polychaeta, Chaetopteridae), with redescription of Mesochaetopterus xerecus and an approach to the phylogeny of the family Downloaded from https://academic.oup.com/zoolinnean/article/152/2/201/2630849 by guest on 25 March 2021 DANIEL MARTIN1*, JOÃO GIL1, JOSEP CARRERAS-CARBONELL1 and MICHEL BHAUD2

1Centre d’Estudis Avançats de Blanes (CSIC), Carrer d’accés a la Cala Sant Francesc 14, 17300 Blanes (Girona), Catalunya, Spain 2Observatoire Océanologique de Banyuls, Université P. et M. Curie – CNRS, BP 44, 66650 Banyuls-sur-Mer, Cedex, France

Received 28 March 2006; accepted for publication 21 May 2007

Mesochaetopterus rogeri sp. nov., a new large chaetopterid polychaete from the Mediterranean Sea, is described. The analyses of partial sequences from the nuclear 18S rRNA (643 bp) and the mitochondrial cytochrome oxidase I (577 bp) genes of representative individuals of all known chaetopterid genera indicated the initial assignment of the new species into Mesochaetopterus. These analyses also supported the monophyly of the family and revealed two well-supported clades: Chaetopterus/Mesochaetopterus and Spiochaetopterus/Phyllochaetopterus. Mesocha- etopterus rogeri sp. nov. is close to Mesochaetopterus xerecus, which is redescribed here from newly collected material. Mesochaetopterus rogeri sp. nov. was characterized as follows: (1) two long tentacles with dorsal transversal black bands of alternating widths (sometimes with two additional longitudinal light-brown bands); (2) region A with nine chaetigers (up to 12), with 13–19 modiﬁed chaetae in the fourth chaetiger; (3) region B with three ﬂat segments, with accessory feeding organs in the second and third segments; (4) sandy straight tubes, 2.5-m long or more, vertically embedded in the sand. In the Bay of Blanes, M. rogeri sp. nov. occurs between 6- and 9-m deep (but also up to 30-m deep), in a patchy distribution (< 1 individual m-2), with maximum densities in April/June (likely to be the result of recruitment events), and minimum densities in September/November (likely to be a behavioural response to increasing sediment dynamics). Although it was originally thought that M. rogeri sp. nov. could be an introduced species, we argue that it is probably a native of the Mediterranean that has been overlooked by scientists up to now. © 2008 The Linnean Society of London, Zoological Journal of the Linnean Society, 2008, 152, 201–225.

ADDITIONAL KEYWORDS: 18S rRNA – behaviour – Chaetopterus – cytochrome oxidase I – distribution – genetics – Mediterranean Sea – Phyllochaetopterus – Spiochaetopterus.

INTRODUCTION data on their presence were these incidental reports, sometimes accompanied by pictures of the living During recent years, scuba divers have occasionally specimens ‘in situ’ (Fig. 1A, B). The oldest known ref- reported the presence of ‘strange’ large worms, with a erence, a worm from Almería, on the south-western pair of long palps and a characteristic striped ‘black Iberian Penninsula, appeared in an encyclopaedia in and white’ colour pattern, in shallow sandy bottoms of 1979 labelled as ‘Spionidae’ (George & George, 1979: the Iberian Mediterranean shoreline, mainly along ﬁg. 8, 353). the Catalan littoral. Up to now, the only available Routine monitoring of the brine discharge from the Blanes desalination plant in the shallow sandy bottoms facing the beach at Punta del Tordera *Corresponding author. E-mail: [email protected] (Catalan coast, north-western Mediterranean)

© 2008 The Linnean Society of London, Zoological Journal of the Linnean Society, 2008, 152, 201–225 201 202 D. MARTIN ET AL. Downloaded from https://academic.oup.com/zoolinnean/article/152/2/201/2630849 by guest on 25 March 2021

Figure 1. Mesochaetopterus rogeri sp. nov. A, B, Two typical positions of the protruded tentacles in living worms (underwater pictures taken by L. Dantart). C, Regions A, B, and C (preserved worm, morphotype with reddish bands). D, Ventral view of region A (preserved worm, morphotype without reddish bands). E, Detail of the anterior end (dorsal view). F, Detail of the anterior end (dorsal view, tentacles removed). Ar, region A; b1–b3, segments of region B; d1–d3, dorsal rami of region-B segments; v1–v3, ventral rami of region-B segments; Cr, region C. Scale bars are in cm. ᭣

revealed a stable population, allowing the collection of shoreline of the Mediterranean coast of the Iberian enough appropriate material. Peninsula (Fig. 2), from October 2001 to September In this study, we used a partial nuclear gene 2004. As a first collection method, we used a PVC Downloaded from https://academic.oup.com/zoolinnean/article/152/2/201/2630849 by guest on 25 March 2021 (18S rRNA) and a mitochondrial gene, cytochrome plate introduced obliquely into the sediment to cut oxidase I (COI), aiming to confirm the assignment of across the worm’s tube, trying to prevent the worms the Catalan specimens into the correct genus within from withdrawing deeper inside the tube. The body in the Chaetopteridae (Annelida, Polychaeta), as well Mesochaetopterus is divided into three differentiated as to assess the phylogenetic relationships within parts: the nine-to-ten anteriormost segments, the the four known genera of the family: Chaetopterus, two-to-three mid-body segments, and an undefined Mesochaetopterus, Spiochaetopterus, and Phyllocha- (but usually very numerous) number of posterior seg- etopterus. ments; these form the regions A, B, and C, respec- The Catalan specimens were revealed to be an tively. Using the PVC plate, the quick reaction of unknown species of Mesochaetopterus. Thus, they are the worm only allowed for the collection of anterior described as a new species, Mesochaetopterus rogeri fragments containing region A and, at most, part of sp. nov, in spite of a lack of data on the morpho- the region B (as well as part of the tube). Successive logy and chaetal arrangement of the posteriormost attempts using either irritating or anaesthetizing segments (because of the logistic difficulties of ex- compounds, and sucking or impelling pumps of tracting complete worms). They were compared with different types, did not result in collection of all known species of the genus, particularly with either entire specimens or larger fragments. The Mesochaetopterus xerecus Petersen & Fanta, 1969, last attempt, carried out in September 2004, involved which closely resembles the new species both in size the archaeological research vessel ‘Tethys’, from the and in having striped palps (Petersen & Fanta, Centre d’Arqueologia Submarina of the Museu 1969). In the case of M. xerecus, however, all material d’Arqueologia de Catalunya. The vessel is equipped was lost (including the types). For this study, new with powerful suction devices, which allowed us to material collected at Paranaguá Bay (Paraná, Brazil) excavate up to 2.5-m deep into the sediment. was used both in genetic and morphologic analyses. However, the tube extended deeper into the sediment, In addition, we provide a brief diagnosis of the soft and only a large fragment of a worm (with ten seg- body of the species, together with a full description ments of region C, in addition to regions A and B) was and pictures of its chaetal arrangement, which was collected after several hours of underwater working. poorly described and illustrated in the original Being the most complete available fragment, this description. specimen was selected as the holotype of the new The seasonal trends of the population in Punta species. del Tordera, habitat preferences, and behaviour of The newly collected specimens of M. xerecus were M. rogeri sp. nov., are also provided, and are com- found in the Ilha do Mel, Paranaguá Bay (Paraná, pared with the known traits of the other known Brazil). The worms were collected by researchers large-sized species of Mesochaetopterus, and are from the Centro de Estudos do Mar (Pontal do Sul, then discussed in an attempt to explain the pres- Brazil) from intertidal soft sediments at low tide ence of such a large unknown worm in the widely using a shovel. investigated shallow waters of the Mediterranean Sea. GENETIC ANALYSIS, DNA EXTRACTION AND SEQUENCING MATERIAL AND METHODS For the genetic analyses, all newly collected speci- SAMPLING PROCEDURE mens were preserved directly in 100% ethanol. Total All observations on living worms and sampling were genomic DNA was extracted from individuals using performed by scuba diving. Most sampling attempts the QIAamp DNA Minikit (Qiagen). A fragment of were carried out off Punta del Tordera on the Catalan 18S rRNA nuclear gene was amplified by polyme-

© 2008 The Linnean Society of London, Zoological Journal of the Linnean Society, 2008, 152, 201–225 204 D. MARTIN ET AL. Downloaded from https://academic.oup.com/zoolinnean/article/152/2/201/2630849 by guest on 25 March 2021

Figure 2. Mesochaetopterus rogeri sp. nov., location of reports. A, Iberian Peninsula. B, Catalan coast. C, Blanes litoral. Key: , seasonal monitoring at the Punta del Tordera; 1, Almería; 2, Alicante; 3, Valencia. rase chain reaction (PCR) using newly designed pri- (Genotek), and 20–30 ng genomic DNA. PCR was mers (Chae18SF, 5′-AAACGGCTACCACATCCAAG- performed in a Primus 96 plus (MWG Biotech). PCR 3′; Chae18SR, 5′-AACTAAGAACGGCCATGCAC-3′). products were cleaned with the QIAquick PCR Puri- Cycle parameters consisted of first a denaturing step fication Kit (Qiagen) or ethanol precipitation, and at 94 °C for 2 min, followed by 35 cycles of 1 min at then sequenced with the BigDye Sequencing Kit ABI 94 °C, 1 min at 55 °C, and 1 min at 72 °C, and then a Prism. PCR products were purified by ethanol pre- final extension at 72 °C for 7 min. The mitochondrial cipitation and analysed on an ABI 3700 automatic COI gene was amplified using primers HCO2198 sequencer (Applied Biosystems) from the Serveis (5′-TAAACTTCAGGGTGACCAAAAAATCA-3′) and Cientifico-Tècnics of the Universitat de Barcelona. LCO1490 (5′-TCAACAAATCATAAAGATATTGG-3′) Representatives of all known genera of Chaetop- (Folmer et al., 1994). The amplification profile was teridae (i.e. Chaetopterus, Mesochaetopterus, Spiocha- optimized for each extraction, optionally with a touch- etopterus,andPhyllochaetopterus) were used for the down of five cycles of 94 °C for 60 s, 45 °C for 90 s, genetic analysis, and the new sequences obtained 72 °C for 60 s, and then 35 cycles of 94 °C for 30–40 s, were deposited in the European Molecular Biology 51 °C for 30–90 s, 72 °C for 60 s, with an initial single Laboratory (EMBL) (Table 1). For 18S rRNA, the denaturation step at 94 °C for 60–120 s, and a final GenBank sequences of Chaetopterus variopedatus single extension step at 72 °C for 5–7 min. Both gene from the North Sea (U67324), Chaetopterus amplifications were carried out in 20 mL of total pugaporcinus (DQ209224), Mesochaetopterus taylori volume with 1X reaction buffer (Genotek), 2 mM (DQ209217), and Prionospio sp. (DQ209226) were

MgCl2, 250 mM deoxyribonucleotide triphosphate also used. For COI, we have also included sequences (dNTP), 0.25 mM of each primer, 1 U Taq polymerase from GenBank of C. pugaporcinus (DQ209257),

Table 1. Accession numbers for the newly sampled and sequenced representative species of the family Chaetopteridae

Species Collection locality N * 18S EMBL† COI EMBL‡

Chaetopterus variopedatus (Renier, 1804) Naples (Italy) 2 AJ966758 AM503094 Banyuls-sur-Mer (France) 2 AJ966759 AM503095 Norwich, Norfolk (UK) 2 U67324 AM503096 Spiochaetopterus solitarius (Rioja, 1917) Port Vendres (France) 2 AJ966760 nda Mesochaetopterus xerecus Petersen & Fanta, 1969 Ilha do Mel (Brazil) 3 AJ966763 AM503097 Mesochaetopterus rogeri sp. nov. Blanes (Spain) 3 AJ966762 AM503098 Phyllochaetopterus socialis Claparède, 1870 Banyuls-sur-Mer (France) 2 AJ966761 DQ209247 Downloaded from https://academic.oup.com/zoolinnean/article/152/2/201/2630849 by guest on 25 March 2021

*N, number of individuals sequenced. †18S rRNAfrom the European Molecular Biology Laboratory (EMBL). ‡Cytochrome oxidase I (COI) from EMBL.

M. taylori (DQ209251), Phyllochaetopterus socialis transversions (Tv)andTs + Tv versus genetic diver- (DQ209247), and Prionospio sp. (DQ209266). Priono- gence for all pairwise comparisons in each gene inde- spio sp. was used as the outgroup. Specimens belong- pendently were plotted. ing to the two morphotypes of M. rogeri sp. nov. (i.e. Each gene was analysed individually, and both with or without reddish bands in the palps) were genes were joined creating a new data set. Phyloge- specifically selected for the analyses. netic trees were inferred by Bayesian inference (BI) using Mr Bayes 3.1.2 (Huelsenbeck & Ronquist, 2001), as this method appears to be the best for inferring phylogenetic relationships between species SEQUENCE ANALYSIS AND PHYLOGENETIC (Alfaro, Zoller & Lutzoni, 2003; Carreras-Carbonell, RECONSTRUCTION Macpherson & Pascual, 2005). The computer Sequences were edited and aligned with SeqMan II program MODELTEST version 3.06 (Posada & Cran- (DNASTAR, Inc.) and ClustalX (Thompson et al., dall, 1998) was used to choose the best-fit evolution 1997) under default settings, and were then verified model under the Akaike information criterion (AIC) visually. Gblocks software was used to check the for each gene separately, and was then subsequently alignments (Castresana, 2000), as regions that are used in the BI analyses (Posada & Buckley, 2004). not well conserved may not be homologous or may The Markov chain Monte Carlo (MCMC) algorithm have been saturated by multiple substitutions, and with four Markov chains was run for 1 500 000 gen- the exclusion of poorly aligned positions and highly erations, sampled every 100 generations, resulting in divergent regions aids phylogenetic reconstruction. 15 000 trees. The first 1500 trees were eliminated The method makes the final alignment more suitable because they did not reach stationarity for the like- for phylogenetic analysis, by selecting blocks of posi- lihood values, and the rest were used to construct the tions that meet a simple set of requirements regard- consensus tree and to obtain the posterior probabili- ing the number of contiguous conserved positions, ties of the branches. lack of gaps, and the degree of conservation of flank- ing positions. Genetic divergence estimated as the percentage of haplotype sequence differences between MORPHOLOGICAL OBSERVATIONS species was calculated using the program PAUP* Light microscope (LM) micrographs were made with version 4.0b10 (Swofford, 2001). Zeiss Axioplan (body and tube) and Zeiss Stemi The homogeneity of base composition across taxa 2000-c (chaetae) stereomicroscopes equipped with the was assessed using the goodness-of-fit (c2) test, and SPOT hardware and software (SP100 KAF1400 the incongruence length difference (ILD) test (Farris digital camera and software version 2.1) from Diag- et al., 1994) was used to assess analytical differences nostic Instruments Inc. For the LM micrographs of between genes; both tests are implemented in chaetae, each parapodia of region A was dissected PAUP* ver. 4.0b10. In the latter test, only parsimony- from the body and isolated in a Petri dish. Parapodia informative characters were included, and heuristic were carefully dissected, and representatives of each searches were performed with ten random stepwise chaetal type were removed with fine forceps, under a additions, with tree–bisection–reconnection (TBR) compound microscope, prior to mounting on a slide. branch swapping and 1000 randomizations. In order LM micrographs of uncini were made by dissecting to assess the degree of saturation, transitions (Ts), a fragment from a parapodium and squashing it

© 2008 The Linnean Society of London, Zoological Journal of the Linnean Society, 2008, 152, 201–225 206 D. MARTIN ET AL. directly onto a slide. The type series of the new using 18S rRNA has shown that the species selected species and the newly collected specimens of M. xere- to represent each chaetopterid genera form a mono- cus were deposited in the Museo Nacional de Ciencias phyletic group with two well-supported clades corre- Naturales (MNCN) of Madrid. sponding to: (1) Chaetopterus and Mesochaetopterus, and (2) Spiochaetopterus and Phyllochaetopterus (Fig. 3A). The genetic distance between the two clades ECOLOGICAL DATA (mean, 3.25%) was three times greater than within Data on population abundances in the Punta del clades (mean, 1.36 and 1.24% between Chaetopterus Tordera were collected monthly, by scuba diving and Mesochaetopterus, and between Spiochaetopterus during routine brine discharge monitoring from Feb- and Phyllochaetopterus, respectively) (Table 2). The ruary 2002 to January 2004. To quantify the abun- mean genetic distance between species from the dif- Downloaded from https://academic.oup.com/zoolinnean/article/152/2/201/2630849 by guest on 25 March 2021 dance of worms, the number of tentacle pairs ferent Chaetopteridae genera was 2.09%. No genetic protruding from the tube was noted along 50-m long differences were found between the two morphotypes and 2-m wide transects. The abundances were moni- of M. rogeri sp. nov. (i.e. with and without reddish tored at two locations (south and facing the Tordera bands in the palps), which were included in the River), with two random sites per location and eight Chaetopterus/Mesochaetopterus clade closest to M. random transects per site. The seasonal pattern is xerecus. Two diagnostic sites were detected (G–T in represented as monthly averages of density (indi- the 145-bp position, and C–T in the 207-bp position) viduals m-2). Water temperature was registered in between M. xerecus and M. rogeri sp. nov., with a parallel with observations on the worms. Data on genetic distance of 0.31% (italics, Table 2). This was wave height were obtained from the mooring buoy less than the distance between Chaetopterus and (DATAWELL Waverider) controlled by the Labora- Mesochaetopterus genera (mean 1.36%), but was tori d’Ingenieria Maritima of the Universitat Politèc- greater than the distance between the two Mediter- nica de Catalunya, located 1 mile off Punta del ranean populations of C. variopedatus (0.16%). The Tordera (41°38.81′N; 02°48.93′E). Wave height was genetic distance between both Mediterranean and expressed as the significant wave height in cm North Sea C. variopedatus populations was greater (height corresponding to the average of a third of than that between the two Mesochaetopterus species the highest waves). (bold numbers, Table 2). The phylogenetic tree constructed using the COI gene yielded two well-supported clades: one contain- RESULTS ing all Chaetopterus species, and the other contain- GENETIC ANALYSIS ing M. xerecus and M. rogeri sp. nov. However, A total of 1220 bp was analysed for all genes com- M. taylori was clustered, with a low node-support bined: 643 bp for 18S rRNA and 577 bp for COI. The value, in the Chaetopterus clade. All Mesocha- 18S rRNA gene presented 54 variable and 24 parsi- etopterus and Chaetopterus species form a well- mony informative sites (8.40 and 3.73%, respectively), supported group clearly differentiated from whereas the COI gene showed a greater proportion, P. socialis (Fig. 3B). No genetic differences were yielding 277 variable and 205 parsimony informative found between the two morphotypes of M. rogeri sites (48 and 35.53%, respectively). The Ts/Tv was 1.46 sp. nov. (i.e. with and without reddish bands in the for 18S rRNA and 1.01 for COI. Saturation tests palps). Between this species and M. xerecus the carried out for each gene independently showed no genetic distance was 1.21% (italics, Table 2). This evidence of sequence saturation in these genes (data distance was greater than between the two Medi- not shown). The goodness-of-fit test for each gene terranean populations of C. variopedatus (0.35%). showed homogeneous base composition across taxa However, the genetic distance between both Medi- (P = 1.00), and the partition homogeneity test showed terranean and North Sea C. variopedatus popula- no significant heterogeneity between genes (PILD = tions was greater than that between the two 0.46). Although there is no generally accepted P value Mesochaetopterus species (bold numbers, Table 2). for significant results, most authors agree data should The phylogenetic tree combining both genes was be combined when P values are greater than 0.05 reconstructed, yielding high node-support values. (Cristescu & Hebert, 2002; Russello & Amato, 2004). Two well-supported clades were found. The first one The models selected according to the AIC and applied included Chaetopterus and Mesochaetopterus, with to the tree reconstruction were as follows: TrN + I the species belonging to each genus being clearly (g=0.002) for the 18S rRNA and GTR + I + G grouped, and forming two different and highly sup- (g=3.38) for COI. ported groups. The second clade only included P. so- We obtained a tree for each gene independently and cialis, as the COI gene could not be sequenced for any combining both genes. Phylogenetic reconstruction Spiochaetopterus species (Fig. 3C).

© 2008 The Linnean Society of London, Zoological Journal of the Linnean Society, 2008, 152, 201–225 NEW MESOCHAETOPTERUS FROM THE MEDITERRANEAN 207 Downloaded from https://academic.oup.com/zoolinnean/article/152/2/201/2630849 by guest on 25 March 2021

Figure 3. Individual gene trees generated by Bayesian inference for 18S rRNA (A), cytochrome oxidase I (COI) (B), and by combining both genes (C). Only posterior probability node support values greater than 0.75 were represented. Med, Mediterranean; NS, North Sea.

SYSTEMATIC ACCOUNT 12) chaetigerous segments. Parapodia with notopo- ORDER CANALIPALPATA dial lobes only, bearing capillary and lanceolate (dor- SUBORDER SPIONIDA sally) to oar-like and sickle-like (ventrally) chaetae. Fourth notopodia with usually 18 (ranging from FAMILY CHAETOPTERIDAE 13 to 19) lanceolate knife-like and stout modified MESOCHAETOPTERUS ROGERI SP. NOV. chaetae. Ventral glandular shield long, pale brown- (FIGS 1, 4–6) ish. Region B with three elongated, flattened seg- Diagnosis: Large Mesochaetopterus with well- ments with biramous parapodia. Notopodia wing- developed peristomium, with a pair of long peri- like, bearing about 15 capillary chaetae. Neuropodia stomial palps with a successive series of dorsal unilobed (segment 1) and bilobed (segments 2 and transversal black stripes, alternating one thick and 3), bearing uncini with eight (nine) teeth. Associated wide with one to several thin narrow ones, some- feeding organs in segments 2 and 3. Region C with times with two longitudinal dorsal and ventral, all segments nearly similar, bearing biramous orange to light-brown stripes, and a lateral black parapodia with unilobed chaetigerous notopodia and stripe (most often present in the first basal half of bilobed uncinigerous neuropodia. Associated feeding the palp), and two longitudinal ciliated grooves. organs absent. Tube longer than 2.5 m, parchment- Second pair of palps and eyes are absent. Region A like internally, externally fully covered by grains of with usually nine (more rarely ranging from 10 to sand. Tube ending unknown.

Table 2. Matrix of pairwise genetic divergence within and between the species representative of all Chaetopteridae genera

C. variopedatus C. variopedatus M. rogeri sp. nov. M. xerecus M. taylori (Med. B) (Med. N)

M. rogeri sp. nov. – 1.21 22.70 24.26 24.61 M. xerecus 0.31 – 22.18 23.92 24.26 M. taylori 1.56 1.24 – 22.01 22.36 C. variopedatus (Med. B) 1.56 1.40 2.18 – 0.35 C. variopedatus (Med. N) 1.71 1.56 2.33 0.16 – C. variopedatus (NS) 1.40 1.56 2.33 0.78 0.93 Downloaded from https://academic.oup.com/zoolinnean/article/152/2/201/2630849 by guest on 25 March 2021 C. pugaporcinus 1.71 1.56 2.33 0.47 0.62 P. socialis 3.27 3.27 3.42 3.42 3.58 S. solitarius 2.80 2.80 3.27 3.11 3.27 Prionospio sp. 4.82 4.67 4.67 4.98 5.13

C. variopedatus (Atl) C. pugaporcinus P. socialis S. solitarius Prionospio sp.

M. rogeri sp. nov. 22.70 23.92 27.29 nda 29.64 M. xerecus 22.36 23.57 26.92 nda 28.77 M. taylori 21.14 20.80 26.02 nda 29.29 C. variopedatus (Med. B) 19.93 21.66 27.11 nda 28.94 C. variopedatus (Med. N) 19.93 21.66 27.11 nda 29.29 C. variopedatus (NS) – 19.58 25.49 nda 30.68 C. pugaporcinus 0.93 – 26.39 nda 30.33 P. socialis 3.89 3.27 – nda 29.01 S. solitarius 3.27 2.96 1.24 – nda Prionospio sp. 5.13 5.13 5.91 5.44 –

M., Mesochaetopterus; C., Chaetopterus; S., Spiochaetopterus; P., Phyllochaetopterus; NS, North Sea; Med. N, Mediter- ranean (Naples); Med. B, Mediterranean (Banyuls); nda, no data available. Above and below diagonal values for cytochrome oxidase I (COI) and 18S rRNA gene values, respectively.

Region A: Holotype with nine segments (N = 14), but the first basal half of the palps, just at the lateral also found with ten (N = 2) or 12 (N = 1) segments, limits of the orange ventral bands. Longitudinal 0.8–1-cm long. Prostomium small, triangular, light- orange bands absent in some specimens. Two longi- brown dorsally, with a rounded, entire anterior tudinal ciliated grooves (one dorsal and one ventral) border. Eyespots absent. Peristomium extended, twice on each palp. Second pair of small antennae or palps as long as, and completely surrounding, the prosto- absent. Eyes absent. Mouth as a vertical slit, below mium, contracted in fixed worms. Two peristomial the prostomium and surrounded by the peristomium. lips, separated by a mid-ventral notch, with variable Anterior end of the dorsal ciliated groove just behind brownish dorsal pigmentation. Two long dorsally the prostomium, between the basis of the palps, grooved palps arise dorsally just behind the junction forming a small triangular lip. Dorsal ciliated faecal of the lateroposterior peristomial borders, up to five groove running from the mouth, through the median times as long as region A in preserved worms (up to line, to the posterior end. Ventral plastron long, 15-cm long ‘in vivo’). Palps with a characteristic pale brownish in colour (often darker posteriorly), colour pattern composed of: (1) two longitudinal restricted to the ventral side of region A, without orange to light-brown stripes (one dorsal and one secretory crescents, but showing a characteristic dis- ventral), covering the whole palp in the last third; (2) tinct epithelium. several successive series of dorsal or dorsolateral Parapodia of region A uniramous, short, with noto- transversal black stripes, alternating one thick and podia only. Chaetae yellowish to pale orange (up to wide with one to several thin and narrow ones (less dark brown in A4), occurring dorsolaterally on two than one third of the thickness, and from half to one irregular question mark shaped rows on segments third of the width of the broad stripes); and (3) a A1–A3 and A5–A6, and in a single, irregular, question longitudinal black stripe of variable length, usually in mark shaped row on segments A4 and A7–A12.

© 2008 The Linnean Society of London, Zoological Journal of the Linnean Society, 2008, 152, 201–225 NEW MESOCHAETOPTERUS FROM THE MEDITERRANEAN 209 Downloaded from https://academic.oup.com/zoolinnean/article/152/2/201/2630849 by guest on 25 March 2021

Figure 4. Mesochaetopterus rogeri sp. nov. A, Dorsalmost capillary chaeta from A1. B, Tip for the dorsalmost capillary chaeta from A1. C, Dorsal fine-lanceolate chaeta from A1. D, Tip of the dorsal fine-lanceolate chaeta from A1. E, Ventral oar-like chaeta from A1. F, Tip of the ventral oar-like chaeta from A1. G, Ventral oar-like chaeta from A3. H, Tip of the ventral oar-like chaeta from chaetiger A3. I, Dorsal fine-lanceolate chaeta from A8. J, Tip of the dorsal fine-lanceolate chaeta from A8. K, Ventral sickle-like chaeta from A8. L, Tip of the ventral sickle-like chaeta from A8. Scale bars are in mm.

© 2008 The Linnean Society of London, Zoological Journal of the Linnean Society, 2008, 152, 201–225 210 D. MARTIN ET AL. Downloaded from https://academic.oup.com/zoolinnean/article/152/2/201/2630849 by guest on 25 March 2021

Figure 5. Mesochaetopterus rogeri sp. nov. A, Chaetal arrangement of the modified A4 chaetiger. B–L, Chaetae. B, Tip of the dorsal finely pointed capillary chaeta ‘d’. (C) Dorsolateral transparent knife-like chaeta ‘ld1’. D, Tip of the dorsolateral transparent knife-like chaeta ‘ld1’. E, Dorsolateral brownish knife-like chaeta ‘ld2’. F, Tip of the dorsolateral brownish knife-like chaeta ‘ld2’. G, I, K, Whole view of the typical modified chaetae ‘lv1’ (lateral), ‘lv2’ (lateroventral), and ‘v’ (ventralmost). H, J, L, Tips of G, I, K, respectively. Scale bars are in mm.

© 2008 The Linnean Society of London, Zoological Journal of the Linnean Society, 2008, 152, 201–225 NEW MESOCHAETOPTERUS FROM THE MEDITERRANEAN 211 Downloaded from https://academic.oup.com/zoolinnean/article/152/2/201/2630849 by guest on 25 March 2021

Figure 6. Mesochaetopterus rogeri sp. nov. A, Dorsalmost uncini from neuropodia B1. B, Ventralmost uncini from neuropodia B1. C, Uncini from notopodia B2. D, Uncini from neuropodia B2. E, Uncini from notopodia B3. F, Uncini from neuropodia B3. G, Uncini from the anterior region of notopodia C2. H, Uncini from the posterior region of notopodia C2. I, Uncini from the anterior region of neuropodia C2. J, Uncini from the posterior region of neuropodia C2. Scale Bars are in mm.

Chaetal arrangement changing in segments A1, lateral ones, and twice as wide as the A1 ventral A2–A6 (except A4) and A7–A12; notopodia with 15 (A1) chaetae; from A7 to A12, the ventral notopodial oar- to 70 (A7–A12) long and fine lanceolate dorsal chaetae, like chaetae being replaced with sickle-like chaetae (up becoming progressively capillary when more dorsal; to 60 chaetae per parapodia). notopodia of A1 with about 30 small lanceolate, oar- A4 notopodia with up to 20 yellowish, transparent, like chaetae lateroventrally; notopodia of A2–A3 and finely pointed capillary chaetae in dorsal position (d); A5–A6 with up to 35 oar-like chaetae lateroventrally, four or five yellowish, transparent knife-like chaetae the ventralmost chaetae twice as wide and long as the more than five times wider than the lanceolate ones

(ld1); between one and three dark yellow to brownish D-shaped, with a single row of eight or nine minute knife-like chaetae, stouter that the previous ones, teeth (most commonly eight), slightly smaller in pos- with the tip of the curved edge slightly serrated (ld2); teriormost segments. 13–19 (typically 18) asymmetrical, knob-like, stout modiﬁed chaetae having serrated tips, the ventral Tube: Known part of the tube straight, completely chaetae (v) smaller than the lateral ones (lv1–lv2). buried into the sediment, except for the aperture, Modiﬁed chaetae dark brown, often partly embedded which protruded 1–2 cm from the sediment surface in the notopodia. and was completely coated with grains of sand (Fig. 1A). From the surface opening the tube followed a vertical path downward for more than 2.5 m. Total Region B: Always with three segments, 2.5–3.5-cm

tube length and shape of end still unknown. Tube Downloaded from https://academic.oup.com/zoolinnean/article/152/2/201/2630849 by guest on 25 March 2021 long, as a flat plate-like region. Flattened and elon- structure very similar all along its length, with gated segments with their flanks dorsally glandular, a relatively thin-walled, parchment-like material all them similar in size, longer than the segments embedded with a thick external layer of sand grains, in region C; B1 slightly narrower than B2 and B3. with a few (often only one) ramifications of the main Associated feeding organs on posterior part of B2 tube at nonregular intervals, shorter than the main and B3, only one per segment. Parapodia of region B tube, filled with sand and closed at the junction by the biramous. Notopodia unilobed: B1 long, pointed, main tube wall. Tube colour changes from yellowish digitiform, distally slightly swollen (d1); B2 and B3 to blackish at around 30 cm below the surface of the wider than B1, triangular, with a groove on the ante- sediment. Detached fragments of sediment-filled or rior side, distally slightly swollen, and with a dark collapsed tubes of different diameters occur around pigmented band just before the distal swelling (d2, the inhabited ones. Tube surrounded by a thick d3). From 11 to 13 extremely long and thin notocha- cylinder of sand all along its length, about 6 cm in etae, scarcely protruding from the tip of the notopo- diameter, more compact than the remaining sedi- dia; about 11 have tape-like tips, whereas two or ment, and particularly evident when drilling around three are capillary. Neuropodia unilobed in B1, with a the inhabited tubes. single low ventral lobe (v1), and bilobed in B2 and B3, with a short, slightly anteriorly orientated dorsolat- Etymology: Species name dedicated to Roger Martin eral lobe, and an elongate, posteriorly orientated (the first author’s elder son). ventral lobe (v2, v3). Uncini roughly D-shaped, with a single row of eight or nine minute teeth (most com- Material examined: Holotype: incomplete specimen, monly eight); dorsal lobe with fewer uncini (about 60) with regions A and B, and only ten segments of than in the ventral lobe (over 400) in B2 and B3. region C, measuring 110-mm long by 11 mm of Uncini of dorsal and ventral lobes in three or more maximum width, MNCN 6.01/10145. Paratypes: 16 irregular rows, with adjacent uncini somewhat dis- incomplete specimens (lacking region C), 13–15-m placed either up or down in relation to each other. deep, Punta del Tordera, Blanes (Girona, Catalunya, Uncini of dorsal lobes with teeth directed posteriorly, Spain, 41.40°N, 2.48°E), collected by D. Martin, whereas those of ventral lobes have anteriorly MNCN 16.01/10146. One incomplete specimen directed teeth. Uncini of dorsal lobes smaller than (lacking regions B and C), 7–10-m deep at Badalona those on the ventral ones. (Barcelona, Catalunya, Spain, 41.27°N, 2.15°E), collected by L. Dantart and G. Álvarez, MNCN 16.01/ Region C: Known only from the holotype, incomplete, 10147. consisting of ten segments for about 5 cm in length. First segment longer than the remaining ones, but Known geographical distribution: Distributional area shorter than those of region B. Parapodia biramous, of the species (Fig. 2) based on underwater observa- as a flat plate-like region with glandular lateral epi- tions. Andalucía, south-western Iberian Penninsula: thelium. Gut markedly protruding from the body 20–30-m deep on the west coast near Hotel La Parra, plan, dark green in living and recently preserved Almería (observed by A. Svoboda, see George & specimens. Notopodia poorly developed, nearly trian- George, 1979); 5-m deep at Cabo de Gata, Almeria gular or wing-like, with three types of chaetae: (observed by J. Junoy). Valencia, western Iberian between five and ten very long and thin, scarcely Penninsula: 10–15-m deep at Cullera, south of the protruding from the tip of the notopodia, between river Jucar (observed by J. Tena’s team); 11-m deep at three and nine with tape-like tips, and between one Canet d’En Berenguer (observed by J. Tena’s team). and three capillary. Associated feeding organs absent. Alicante, western Iberian Penninsula: 20-m deep at Neuropodia all bilobed, similar to B2 and B3 in shape, Punta del Rincón de Lois, Benidorm (observed by J. distribution, and number of uncini. Uncini roughly Tena’s team). Catalunya, north-western Iberian Pen-

© 2008 The Linnean Society of London, Zoological Journal of the Linnean Society, 2008, 152, 201–225 NEW MESOCHAETOPTERUS FROM THE MEDITERRANEAN 213 ninsula: 5–7-m deep at Punta de la Mora, Garraf the lower part, closed at the lower extremity, and (observed by B. Weitzmann); 10–15-m deep at Bada- ending blindly in a nearly rounded apex. lona (observed by L. Dantart, G. Álvarez, and X. Males with elongate sperm having a long ﬂagellum. Turón); 30-m deep at Mataró (observed by L. Dantart); Females with oocytes of about 200 mm in diameter, 20-m deep at Arenys de Mar (observed by B. Weitz- which are present in all segments of region C. Popu- mann); 10-m deep at Malgrat (observed by L. Dantart); lation densities at Ilha do Mel reach about 100 indi- 6-m deep at Blanes Harbour (observed by D. Martin); viduals m-2. 6–10-m deep at Cala Sant Francesc, Blanes (observed by different CEAB divers); 6–15-m deep at Cala Region A: Between 9 and 14 segments, more fre- S’Aguia, Pinya de Rosa, Blanes (observed by different quently 9 (N = 3) or 13 (N = 3). Parapodia all

CEAB divers); 10–15-m deep at Cala Pola, Tossa de Downloaded from https://academic.oup.com/zoolinnean/article/152/2/201/2630849 by guest on 25 March 2021 uniramous, with notopodia only. Chaetae yellowish to Mar (observed by the CEAB Caulerpa team); west of pale orange (dark brown in A4), occurring dorsolater- Urbanization Rosamar, Sant Feliu de Guixols, ally in several irregular, question mark shaped rows 20–30-m deep (observed by A. Svoboda). (a single row in A4). Chaetal arrangement changing in segments A1, A2–A6 (except the A4), A7–A9, and MESOCHAETOPTERUS XERECUS PETERSEN &FANTA, A10–A14; notopodia of Al with two types of chaetae, 1969 (FIGS 7–11) up to 25 very long and fine lanceolate to capillary dorsal chaetae, becoming finer as progressing to the Diagnosis: Based on Petersen & Fanta (1969) and on dorsum, and about 30 small lanceolate, oar-like lat- observations of newly collected specimens. Large cha- eroventral chaetae; notopodia of A2–A14 with up to etopterid (reaching up to 60-cm long ‘in vivo’) with a 30 very long and fine lanceolate dorsal chaetae, their well-developed peristomium (about one-fifth of the flattened ends becoming shorter when more dorsal; length of region A) surrounding the prostomium, and notopodia from A2–A3 and A5–A6 with up to 35 a pair of long peristomial palps up to twice as long as oar-like lateroventral chaetae, the ventralmost region A in preserved worms. Palps with transversal chaetae twice as wide and long as the lateral chaetae; dark pigmented rings (dark-orange to greenish- notopodia from A7 to A9, with up to 30 bayonet brown) of different widths, without a clear alternating ventral chaetae and a few hooked lateroventral pattern, and a pair of longitudinal ciliated grooves chaetae; the bayonet chaetae becoming smaller and (one dorsal and one ventral). Second pair of small progressively replaced by the hooked ones in the antennae or palps absent. Eyes present between posteriormost segments; from A10 to A14, the hooked insertion of the palps and the peristomium. Dorsal chaetae fully replacing the bayonet chaetae. ciliated faecal groove running from mouth to posterior A4 notopodia with up to 15 yellowish, transparent, end (anus), along the median body line. Ventral plas- finely pointed lancet-like dorsal chaetae (d); five yel- tron occupying the whole of region A and up to one lowish, transparent, knife-like chaetae more than third of the first segment of region B. twice as wide as the lancet-like ones (ld1); between Region A 1–1.5-cm long, with 8–14 chaetigerous two and five dark-yellow to brownish knife-like segments (typically 11 or 12). Parapodia uniramous, chaetae, stouter than the previous ones, with the tip notopodial lobes short; ventral glandular shield long, of the curved edge slightly serrated (ld2, ld3); 12–14 uniformly greenish in colour. Region B about 3-cm (typically 11) asymmetrical, knob-like, stout modified long, with four (but up to seven) elongate segments, chaetae having serrated tips, the ventral chaetae (v) and with associated feeding organs or cupules usually smaller than the lateral ones (lv1–lv3). Modified in segments 2–4 (but found up to segment 7). Parapo- chaetae dark brown, often partly embedded in the dia biramous, as a flat plate-like region, with glan- notopodia. dular lateral epithelium. Notopodia poorly developed, with a nearly triangular or wing-like shape. Uncini- gerous neuropodia unilobed in segment 1, and bilobed Region B: Usually with four (up to seven) segments in segments 2 and 3. Region C up to 55-cm long for having biramous parapodia. Between 12 and 16 90–120 segments. Parapodia biramous, with unilobed extremely long, thin, notochaetae, scarcely protruding notopodia and bilobed uncinigerous neuropodia. All from the tip of the notopodia; about 12 with flattened- segments similar, except for some in the posterior- to-lanceolate tips, and two or three with pointed tips. most pygidial region, which are shorter and have Neuropodia unilobed in segment B1, and bilobed in reduced parapodia. B2 and B3, with several hundreds of uncini irregu- Tube longer than 1 m, parchment-like, externally larly disposed in two or three rows, roughly triangu- covered by grains of sand (inconspicuous in worms lar, with seven (between six and ten) long curved from muddy sandy bottoms), vertical or J-shaped, teeth plus a few small ones (often difficult to distin- with a transverse partition with three perforations in guish) in the anterior and posterior ends of the ser-

© 2008 The Linnean Society of London, Zoological Journal of the Linnean Society, 2008, 152, 201–225 214 D. MARTIN ET AL. Downloaded from https://academic.oup.com/zoolinnean/article/152/2/201/2630849 by guest on 25 March 2021

Figure 7. Mesochaetopterus xerecus. A, Dorsal view of region A (preserved). B, Ventral view of region A (preserved). C–J, Chaetae. C, Dorsalmost capillary chaeta from A1. D, Dorsolateral fine-lanceolate chaeta from A1. E, Ventral oar-like chaeta from A1. F, Tip of the ventral oar-like chaeta from A1. G, Dorsalmost fine-lanceolate chaeta from A3. H, Tip of the dorsalmost fine-lanceolate chaeta from A3. I, Ventral oar-like chaeta from A3. J, Tip of the ventral oar-like chaeta from A3. Scale bars are in cm in A and B, and are in mm in C–J. rated edge. All uncini of B1 similar in size; dorsal chaetae (shorter than the notochaetae of B region), uncini of B2 and B3 smaller than ventral uncini. scarcely protruding from the tip of the notopodia, with flattened to lanceolate tips. Neuropodia all Region C: Incomplete, with more than 50 segments, bilobed, with uncini irregularly disposed in two or all biramous. Notopodia with about ten long and thin three rows, similar in shape and teeth arrangement

intertidal sandy beach, Ilha do Mel, Paranaguá Bay (Paraná, Brazil), 25°34′S, 48°20′W, October 30 2001, collected by V. Radashevsky, MNCN 16.01/10148.

Habitat, behaviour and population density of Mesochaeopterus rogeri sp. nov. Mesochaetopterus rogeri sp. nov. commonly inhabits ﬁne to coarse grain sandy bottoms (of 400–600 mmin grain diameter) between 5- and 30-m deep (but most commonly between 6- and 15-m deep). These locations

are often subject to medium-speed currents, which Downloaded from https://academic.oup.com/zoolinnean/article/152/2/201/2630849 by guest on 25 March 2021 may facilitate the suspension-feeding behaviour of the worm. Different kinds of particles (possibly potential food) transported by the water currents were observed ‘in situ’ when they came into contact with the tentacles: they remained attached to the surface of the tentacles and began to be transported towards the mouth with the help of the ciliated grooves. When undisturbed, the tentacles protruded from the tube opening and formed a V, with both tips arranged in a spiral (Fig. 1A). Changes in current speed or direction did not trigger any response from the worm, except for the modification of the tentacle position induced by the new conditions, which may lead toward a less regular tentacle arrangement in the case of stronger currents (Fig. 1B). Any contact with the surrounding sediment induced the worm to retract inside the tube. A slow constant retraction occurred after a subtle contact. If the contact was not repeated, the worm stopped the reaction and re-acquired the typical position after a few minutes. Either repeated or violent contacts with the surrounding sediment caused a fast retraction inside the tube. However, after several minutes, the worm protruded again to adopt its usual position. The worms were able to expel both introduced liquids and small particles (such as sand grains) from the interior of their tubes, by means of a powerful exhalation current. In normal conditions, with the worms completely inside the tube, there was a regular exhalation/inhalation water flow, probably generated by peristaltic movements of the segments in region C. This flow seemed similar to those described as a part of the feeding mode and ventila- Figure 8. Mesochaetopterus xerecus, scheme of a mature tion system for other chaetopterid polychaetes, sperm. including Mesochaetopterus (Barnes, 1965; Sendall, Fontaine & O’Foighil, 1995). to those of region B. Uncini differing in size both The population density ranged between 1 and dorsoventrally and from anterior (smaller) to poste- 3 individuals 40 m-2, but may reach up to 1 rior (larger) sections of each neuropodia, clearer individual 10 m-2 (Fig. 12B). However, they were fre- in anteriormost segments, and progressively less quently found in clusters of 2 or 3 individuals m-2, evident in the posteriormost segments, becoming separated from their neighbours by large unoccupied basically similar in size around segment C30. areas. Maximum densities tended to occur from April to June, together with rising temperatures, and could Material examined: Ten incomplete specimens perhaps be related to recruitment events, as proposed (reaching up to 55 segments in region C) from a low for Mediterranean soft-bottom invertebrates (Sardá

© 2008 The Linnean Society of London, Zoological Journal of the Linnean Society, 2008, 152, 201–225 216 D. MARTIN ET AL. Downloaded from https://academic.oup.com/zoolinnean/article/152/2/201/2630849 by guest on 25 March 2021

Figure 9. Mesochaetopterus xerecus. A, Dorsalmost capillary chaeta from A8. B, Dorsolateral ﬁne-lanceolate chaeta from A8. C, Lateroventral sickle-like chaeta from A8. D, Tip of the lateroventral sickle-like chaeta from A8. E, Ventralmost lancet-like chaeta from A8. F, Tip of the ventralmost lancet-like chaeta from A8. G, Ventralmost sickle-like chaeta from A10. Scale bars are in mm.

© 2008 The Linnean Society of London, Zoological Journal of the Linnean Society, 2008, 152, 201–225 NEW MESOCHAETOPTERUS FROM THE MEDITERRANEAN 217 Downloaded from https://academic.oup.com/zoolinnean/article/152/2/201/2630849 by guest on 25 March 2021

Figure 10. Mesochaetopterus xerecus. A, Chaetal arrangement of the modified A4 chaetiger. B–J, Chaetae. B, Tip of the dorsal finely pointed lancet-like chaeta ‘d’. C, Dorsolateral transparent knife-like chaeta ‘ld1’. D, Dorsolateral transparent knife-like chaeta ‘ld2’. E, Dorsolateral brownish knife-like chaeta ‘ld3‘. F, Lateroventral brownish knife-like chaeta ‘ld3’. G, H, I, Typical modified chaetae: lv2, lateral; lv3, lateroventral; v, ventralmost. J, Tip for the ventralmost modified chaeta ‘v’. Scale bars are in mm.

© 2008 The Linnean Society of London, Zoological Journal of the Linnean Society, 2008, 152, 201–225 218 D. MARTIN ET AL. Downloaded from https://academic.oup.com/zoolinnean/article/152/2/201/2630849 by guest on 25 March 2021

Figure 11. Mesochaetopterus xerecus. A, Dorsalmost uncini from neuropodia B1. B, Ventralmost uncini from neuropodia B1. C, Uncini from notopodia B2. D, Uncini from neuropodia B2. E, Uncini from notopodia B3. F, Uncini from neuropodia B3. G, Uncini from the anterior region of notopodia C2. H, Uncini from the posterior region of notopodia C2. I, Uncini from the anterior region of neuropodia C2. J, Uncini from the posterior region of neuropodia C2. K, Uncini from notopodia C30. L, Uncini from neuropodia C30. Scale bars are in mm.

© 2008 The Linnean Society of London, Zoological Journal of the Linnean Society, 2008, 152, 201–225 NEW MESOCHAETOPTERUS FROM THE MEDITERRANEAN 219 Downloaded from https://academic.oup.com/zoolinnean/article/152/2/201/2630849 by guest on 25 March 2021

Figure 12. Mesochaetopterus rogeri sp. nov., seasonal pattern in abundance at Punta del Tordera, in relation to water temperature and wave height. A, Water temperature at 20-m deep and significant wave height. B, Population density. et al., 1995, Sardá, Pinedo & Martin, 1999). However, tubes, abandoned by the worms either during their the seasonal pattern of this species differed from the normal growth or as a result of the aforementioned general pattern proposed by Sardá et al. in that adaptive process. Once the storms ended, the worms remarkable minimum densities occurred between Sep- built new tube sections to reach the surface, so that tember and November (particularly in October). This there would be as many empty tubes as stormy events could not be explained by an increasing mortality after during the lifetime of each worm. recruitment events leading to basal adult densities, as This mode of life supports the potential influence of previously proposed (Sardá et al., 1995, 1999). In M. rogeri sp. nov. in structuring the surrounding sedi- M. rogeri sp. nov., the density decrease did not seem to ments, as a sediment stabilizer (dense populations be related to mortality, but to an adaptation to survive inside extremely long tubes, palisades of empty under the increasing instability of the sediment during tubes). Also, the presence of a cylinder of compact the stormy autumnal period characteristic of a Medi- sediment surrounding the tube all along its length terranean environment (Fig. 12A). In fact, the extraor- (i.e. more than 2.5-m deep into the sediment, and dinary length of the tube, as well as the ability to reaching up to 6 cm in diameter) in M. rogeri sp. nov. quickly retract inside, could represent an adaptation to (also reported for M. taylori) has been attributed to an such an unstable sedimentary regime. Together with ‘aerobic halo’ generated by the presence of the tube the increase of instability of the water column (as (Sendall et al., 1995). Therefore, a possible significant expressed by the increase in wave height, Fig. 12A), bioturbating potential can also be attributed to the the sediments tended to become more and more M. rogeri sp. nov. populations. mobile, to the extent that the worms were not able to protrude above the surface, and survived by hiding Morphological comparison between inside the tube, at a depth where the sediment was Mesochaeopterus rogeri sp. nov. and all known stable. This could explain the autumnal decrease in species of the genus density (Fig. 12B), as the divers could not count the Previous studies on the intra- and interspecific vari- worms that did not protrude from the sediment ability in the genus Mesochaetopterus show that hard surface. Moreover, this could also explain the occa- structures (i.e. chaetae and uncini) show a continuity sional presence of empty ramifications and the huge both in shape and size, from larvae to juveniles and number of empty tubes found during the excavation adults (Bhaud, 2005). This allows the small-sized collection. Accordingly, these were probably functional species of the genus – viz. Mesochaeopterus minutus

Table 3. Comparison of the main characteristics of the currently known large-sized Mesochaetopterus species

M. xerecus M. rickettsii M. rogeri M. taylori Petersen and Berkeley and Characters sp. nov. (Potts, 1914) Fanta, 1969* Berkeley, 1941

Length of living worm 8.3 (for regions A, 60 60 35 (cm) B, and the 10 segments of C) Width of worm (cm) 0.8–1 1 0.8–1.5 1 Tentacle colour Dark pigmented rings Uniform light Dark pigmented ? and bands, and red yellow rings Downloaded from https://academic.oup.com/zoolinnean/article/152/2/201/2630849 by guest on 25 March 2021 longitudinal bands Relative length of 3–5:1 1:1 2:1 2:1 tentacles: A region Eyes Absent Absent Present ? Parapodia of region A 9–12 (9) 9–11 (9) 8–14 (11) 10 Chaetae of the A4 13–19 (18) 10 7–17 (11) ? segment Shape of the modified Short, strong with Short, strong with Short, strong with Fimbriated, with A4 chaetae finely serrated tip truncated tip finely serrated tip oblique edge Types of chaetae in the 33?4? a region (except A4) Parapodia of region B 3 3 4 (up to 7) 21 Accessory feeding B2 B2–B3 B2–B3 (up to C4) B21 organs Glandular lateral B1-C? B1–B3 B1–B4 region B epithelium Length of the largest 98–111 115 106–134 ? uncini of the B region (mm) Number of teeth on the 8 (8–9) 9 (7–10) 7 (6–10) 8–10 uncini of region B Relative length of teeth 2/5 of uncini width 1/3 of uncini width 1/2 of uncini width ? on the uncini of region B Outline of uncini of Roughly D-shaped, Roughly triangular, Roughly triangular, Roughly D-shaped, region B with slightly curved with markedly with straight with straight edges curved edges edges edges Tube Straight up to 2.5 m L-shaped, Vertical or slightly Straight up to deep, sand covered, parchment-like J-shaped, 1 m, 2 m deep, sand lower end unknown with sand grains parchment -like covered, lower embedded, straight with sand grains end unknown section 10–20 cm embedded, lower end deep blid and rounded Type locality Mediterranean California Brazil California

Potts, 1914 Mesochaeopterus capensis (McIntosh, sp. nov. measured about 11-cm long (lacking most of 1885), Mesochaeopterus laevis Hartmann-Schroeder, region C), and their uncinal plates are slightly longer 1960, Mesochaeopterus crypticus Ben-Eliahu, 1976, than 110 mm. and Mesochaeopterus xejubus Petersen & Fanta, 1969 Among the large-sized Mesochaetopterus species – to be discarded for the purpose of the taxonomical (Table 3), M. rogeri sp. nov. resembled M. taylori and comparison with M. rogeri sp. nov. All these species Mesochaeopterus alipes (Monro, 1828) in having three seldom reach 3.5-cm long (see Table 1 in Nishi, 1999), parapodia on region B. In addition to geographical and the largest uncinal plates are shorter than 70 mm distribution and tentacle colour pattern, the new (Bhaud, 2005), whereas the holotype of M. rogeri species differs from M. taylori in the number of A4

Table 3. Continued

M. alipes M. japonicus M. selangolus M. mexicanus Characters (Monro, 1828) (Fujiwara, 1934) (Rullier, 1976) Kudenov, 1975

Length of living worm 80 25 25 (up to 60) > 9.5 (cm) Width of worm (cm) 6 1 1 0.5 Tentacle colour Opaque white, with Unpigmented Uniform pale brownish ? traces of brown Relative length of 2:1 1:1 1.5:1 ? tentacles: A region Downloaded from https://academic.oup.com/zoolinnean/article/152/2/201/2630849 by guest on 25 March 2021 Eyes Absent Absent Absent Present Parapodia of region A 9 9 9–10 13 Chaetae of the A4 ? 8–13 8–15 ? segment Shape of the modiﬁed Short, strong with Truncated, spatulate, Truncated, spatulate, Oblique, distally A4 chaetae truncated edge with serrations with serrations dentate Types of chaetae in ?? ? 2 region A (except A4) Parapodia of region B 3 2 2 4 Accessory feeding B2 B2 B2 B1–B4 organs Glandular lateral B1 B1 and anterior B2 B1 and anterior B2 ? epithelium Length of the largest ? ? 72–85 140 uncini of region B (mm) Number of teeth on the 8 9–10 7–8 8–9 uncini of region B Relative length of teeth ? ? 1/2 of uncini width 1/2 of uncini width on the uncini of region B Outline of uncini of ? ? Roughly D-shaped, with Roughly triangular, region B slightly curved edges with markedly and serrated basis curved edges Tube Parchment-like Strait, parchment- J-shaped, parchment- Parchment-like like, with blind lower like with embedded end sand and periodic annulated rings and perforations at the lower end Type locality Panama Japan Malaysia Mexico

*Data emended on the basis of the studied material.

chaetae, the shape of uncini from region B, and the tion of material from Invertebrate Zoology, Royal tube shape, and differs from M. alipes in the shape of British Columbia Museum, and the Natural History the A4 chaetae and the tube composition. Mesocha- Museum of Los Angeles seems to support the exist- etopterus alipes was found in Panama and is only ence of different species along the different geographi- known from the original description, which lacked cal locations where the species has been reported, many relevant data. Mesochaetopterus taylori was some of them being probably new to science (M. originally described from the Paciﬁc coasts of British Bhaud, unpubl. data). The data used for the compari- Columbia by Potts (1914), and has been subsequently son between M. taylori and M. rogeri sp. nov. are thus reported along the Paciﬁc and Atlantic coasts of based on the original description of the species America (Gilbert, 1984; Blake, 1996). The examina- (Table 3).

The possible coincidence of tentacle colour pattern etopterus was more than twice (Banyuls) or even (somewhat based on dark rings or bands), as well as three times (Naples) greater, indicating that they are the genetic analyses, lead us particularly to compare probably different species, with the North Sea one the new species with M. xerecus. However, the origi- incorrectly identiﬁed as C. variopedatus. nal type material of this species was lost, and the When analysing a faster evolving gene (COI), comparisons are based on newly collected material. In genetic divergence values between species obviously addition, this helped to complete the original descrip- increase, with the values between Chaetopterus tion by Petersen & Fanta (1969). However, the striped species ranging from 18 to 21% (Osborn et al., 2007), pattern, as well as the additional reddish longitudinal similar to the values found in the present study. stripes present in some specimens and the tentacle Conversely, the genetic distance between M. rogeri

length, clearly identifies M. rogeri sp. nov. In addition, sp. nov. and M. xerecus was only 1.21% (i.e. smaller Downloaded from https://academic.oup.com/zoolinnean/article/152/2/201/2630849 by guest on 25 March 2021 this species may be separated from M. xerecus by: (1) than that between the Mediterranean and North Sea the chaetal arrangement of region A (including the A4 populations of C. variopedatus). segment), and (2) the number and size of teeth, and Both 18S and COI genes suggested that the Medi- the shape and outline of uncini from regions B and C terranean Chaetopterus belonged to the same species, (Table 3). whereas the North Sea specimens must be considered Tube morphology of M. rogeri sp. nov. seems to coin- as different. In fact, several different species of Cha- cide with that of Mesochaeopterus rickettsii Berkeley etopterus have been described from European coasts, & Berkeley, 1941, at least in the known sections (2.5 both Mediterranean and Atlantic, probably with some and 2 m, respectively). However, both species can be of them being valid in addition to C. variopedatus easily distinguished by the number of segments in (Petersen, 1984). Further studies focused on Cha- region B (Table 3). etopterus are needed, involving more individuals and locations, to asses the taxonomical status of the C. variopedatus complex. DISCUSSION The monophyly of the Chaetopteridae has been PHYLOGENETIC APPROACH previously demonstrated (Osborn et al., 2007), and our results confirmed this finding. However, the The main purpose of the molecular analysis was to monophyly at the genus level does not seem to be well assign the new species (M. rogeri sp. nov.) to its resolved. According to our results, Mesochaetopterus correct genus. A second, but no less relevant conse- is a sister group of Chaetopterus, with both appearing quence, was that the phylogenetic relationships to be monophyletic using either the 18S or the com- between Chaetopteridae genera could also be eluci- bined 18S and COI analysis, and supported by high dated. All our analyses placed M. rogeri sp. node support values (Fig. 3A, C). Conversely, M. tay- nov. within a Mesochaetopterus/Chaetopterus clade, lori was first joined to the Chaetopterus clade when always closely related to M. xerecus, demonstrating using COI alone, although with low node support that the new species described here belonged to values (Fig. 3B). In turn, the monophyly of Phyllocha- Mesochaetopterus, coinciding with the results of the etopterus and Spiochaetopterus cannot be assessed, as morphological study. only one species per genus has been available for the The 18S rRNA gene is a slowly evolving sequence analyses. Furthermore, it was impossible to obtain a (Hillis & Dixon, 1991; Avise, 1994), and has been COI sequence for the Spiochaetopterus species analy- regarded as a very conservative marker of speciation sed in this paper [Spiochaetopterus solitarius (Rioja, events. With this gene, M. xerecus and M. rogeri 1917)]. In all studied cases, however, the specimens of sp. nov. appeared closely related (0.31%), but as both genera were closer to each other, and were distinct clades within the Mesochaetopterus clade clearly more distant from the Mesochaetopterus and (Table 2; Fig. 3A). Osborn et al. (2007) found similar Chaetopterus clades. values between Chaetopterus species (ranging from 0.4 to 1.6%). By including specimens of C. variopedatus from the Mediterranean (newly collected TAXONOMY, DISTRIBUTION, AND BEHAVIOUR OF in Naples and Banyuls) and the North Sea (from MESOCHAEOPTERUS ROGERI SP. NOV. GenBank, collected in Norwich, Norfolk, UK) coasts We initially thought that M. rogeri sp. nov. could be in the genetic analysis, a result that could have been an introduced exotic species, which led us to the expected from a biogeographical point of view was comparison with M. xerecus. Comparison among all revealed. Although the genetic distance between the known species of the genus clearly pointed out differ- Mediterranean Chaetopterus was less than half that ences between them and M. rogeri sp. nov.; however, of the two species of Mesochaetopterus, the distance the large-sized Brazilian species apparently had between the Mediterranean and North Sea Cha- the most similar morphology (Table 3), which is

© 2008 The Linnean Society of London, Zoological Journal of the Linnean Society, 2008, 152, 201–225 NEW MESOCHAETOPTERUS FROM THE MEDITERRANEAN 223 supported by the genetic analyses. Moreover, the scuba divers in recent years: there were fewer than long-living planktonic larvae typical of Chaetop- 100 divers associated with the Catalan Federation of teridae might have easily been introduced to the Underwater Activities (FECDAS) in 1954, whereas Mediterranean through ballast water discharge from the current number exceeds 12 000. a ship coming from Brazil. The discovery of a new From the point of view of traditional soft-bottom population of M. xerecus in the Ilha do Mel (Pontal do sampling, the behaviour and tube structure of Sul, Brazil), close to the coasts from where the species M. rogeri sp. nov. may also have contributed to was originally described (Rio de Janeiro), allowed us ‘hiding’ the presence of the species in the widely to discard the hypothesis that the Mediterranean investigated shallow waters of the Mediterranean species was the same as the Brazilian one. However, Sea. The prompt reaction of M. rogeri sp. nov. to any

taking into account the conspicuous appearance of contact with the surrounding sediment, as well as the Downloaded from https://academic.oup.com/zoolinnean/article/152/2/201/2630849 by guest on 25 March 2021 M. rogeri sp. nov. and the long-term background extreme length of the tube, make its collection virtu- knowledge on the polychaete fauna of the Mediterra- ally impossible. The initial contact of a grab or dredge nean Sea, the possibility of an exotic origin for with the sediment surface triggers a fast retraction M. rogeri sp. nov. (from a foreign, but still unknown, inside the tube, so that only fragments of empty tubes region) cannot be dismissed. On the other hand, most are collected (R. Sardá, S. Pinedo & D. Martin, pers. known introduced species of marine invertebrates, comm.), as has been repeatedly observed during the including polychaetes, tend to occur as concentrated long-term monitoring carried out in the Punta del populations in geographically restricted areas, often Tordera (Sardá et al., 1995, 1999, 2000; Pinedo, in sheltered environments, such as harbours or Sardá & Martin, 1996). Other large-sized species of coastal lagoons (Goulletquer et al., 2002). According to Mesochaetopterus also build very long tubes, but they numerous underwater observations, M. rogeri sp. nov. are either L-shaped, as in M. taylori (Sendall et al., is currently widely distributed along the Iberian 1995), or J-shaped, as is the case of M. xerecus (V. Mediterranean coast, particularly in Catalunya, Radashevsky, pers. comm.; Petersen & Fanta, 1969). which is contradictory to a hypothetical introduced Moreover, they rarely extend into the sediment origin for the species. deeper than 30 cm or 1 m, respectively. In addition, Independently of the origin of the species, a second they live in tidal environments so that collecting question occurs when trying to explain the increasing specimens at low tide is an easy task, in comparison number of records in recent years. Could it be caused with M. rogeri sp. nov. In fact, all attempts to collect by a recent increase in the distributional area of the entire specimens failed, and only the last one involv- species? If so, this could be related either to a spread- ing a large research vessel succeeded in obtaining the ing of the species from warm southern waters (like longest fragment (i.e. the holotype), which included a those at Almería, from where the species was first few segments of region C. A similar situation was recorded) caused by an increase in seawater tempera- reported for M. rickettsii, the tubes of which extended ture (perhaps related to climate change), or to the vertically into the sediments for about 2 m, so that decline of shellfish harvesting efforts resulting from the entire tube was never collected, and the orienta- the overfishing of the smooth venus Callista chione tion and structure of the lower end still remains (Linnaeus, 1758), which has led to a reduction in unknown (MacGinitie & MacGinitie, 1949; Sendall habitat disturbance. Although stimulating, explora- et al., 1995). tion of the possible relationships between these Taking this into account, and the fact that M. rogeri hypotheses and the postulated expanding presence of sp. nov. seems to be perfectly adapted to the peculiar M. rogeri sp. nov. does not seems to be possible at characteristics of its seasonally unstable sandy present, and falls out of the scope of this study. habitat, it seems more appropriate to argue that this However, there is a simple framework allowing species is more likely to be a native Mediterranean explanation of the increase of reports. Mesocha- one that has been overlooked by scientists until now. etopterus rogeri sp. nov. inhabits ‘uninteresting’ sandy bottoms, which are seldom visited by divers. One of ACKNOWLEDGEMENTS these visits (and the subsequent observation of the worm) occurred once in the 1970 (George & George, With this paper we honour Lluís Dantart, who died in 1979), and a gap of about 20 years passed until the a car accident in spring 2005. He took the first recent second recording in 1995. The sand coverage of the underwater picture and collected the first fragment of tubes and the colour pattern of tentacles may help M. rogeri sp. nov., starting the sequence of data that the species to remain camouflaged within the sedi- led to the description of the new species. We wish to ment, so that direct observations are a matter of thank Guillermo Álvarez, Xavier Turón, Boris Weitz- chance. Nevertheless, the probability of observation mann, Juan Junoy, the José Tena and the CEAB’s rose with the increase in scientific and naturalist Caulerpa teams, as well as many other divers of the

CEAB who reported the presence of the new species. Cristescu MEA, Hebert PDN. 2002. Phylogeny and adapta- We are particularly grateful to David George and tive radiation in the Onychopoda (Crustacea, Cladocera): Armin Svoboda, who contributed with the ﬁrst known evidence from multiple gene sequences. Journal of Evolu- report, and to Xavier Nieto, from the Centre tionary Biology 15: 838–849. d’Arqueologia Submarina of the Museu d’Arqueologia Farris JS, Kallersjo M, Kluge AG, Bult C. 1994. Testing de Catalunya and the cruise of the archaeological RV signiﬁcance of incongruence. Cladistics 10: 315–319. ‘Tethys’, for his insightful help in collecting the holo- Folmer OM, Black W, Hoeh R, Lutz R, Vrijenhoek R. type. Thanks also to all Brazilian colleagues involved 1994. DNA primers for ampliWcation of mitochondrial cytochrome c oxidase subunit I from diverse metazoan inverte- in the search of the type series of M. xerecus, espe- brates. Molecular Marine Biology and Biotechnology 3: 294– cially to Paulo Lana and Vasily Radashevsky, who 299.

also collected the new specimens. Enrique Macpher- Downloaded from https://academic.oup.com/zoolinnean/article/152/2/201/2630849 by guest on 25 March 2021 George JD, George JJ. 1979. Marine life an illustrated son and Nuria Raventos kindly authorized the use of encyclopaedia of invertebrates in the sea. London: Harrap. their data on the seasonal survey at the Punta del Gilbert KM. 1984. Chapter 11. Family Chaetopteridae Tordera. Temir A. Britayev, Helmut Zibrowius, and Malmgren, 1867. In: Uebelacker JM, Johnson PG, eds. the colleagues of the International Polychaete Asso- Taxonomic guide to the polychaetes of the northern of gulf of ciation provided insightful comments and discussion Mexico. Metairie, LA: Minerals Management Service, U.S. on collection methods and distribution hypothesis, Department of the Interior, Gulf of Mexico Regional Office, and the anonymous reviewer and the Associated 11/1–11/13. Editor highly improved the quality of the manuscript. Goulletquer P, Bachelet G, Sauriau PG, Noël P. 2002. Marta Pascual kindly checked the genetic sections. Open Atlantic coast of Europe. A century of introduced The Catalan Federation of underwater activities pro- species into French waters. In: Leppäkoskii E, Gollasch S, vided the data on the increasing number of divers in Oleiin S, eds. Invasive aquatic species of Europe: distribu- Catalunya. This study has been partly financed by a tion, impacts and management. Dordrecht: Kluwer Aca- research contract between the CEAB (CSIC) and demic Publishers, 276–290. CREOCEAN. This paper is a contribution to the Hillis DM, Dixon MT. 1991. Ribosomal DNA: molecular research projects INTAS–OPEN-97–0916 and EVK3- evolution and phylogenetic inference. Quarterly Review of 2000-22038. Biology 66: 411–453. Huelsenbeck JP, Ronquist FR. 2001. MrBayes: Bayesian REFERENCES inference of phylogenetic trees. Bioinformatics 17: 754–755. MacGinitie GE, MacGinitie N. 1949. Natural history of Alfaro ME, Zoller S, Lutzoni F. 2003. Bayes or Bootstrap? marine animals. New York: MacGraw-Hill Book Co. A simulation comparing the performance of Bayesian Nishi E. 1999. Redescription of Mesochaetopterus seanglosus Markov Chain Monte Carlo sampling and bootstrapping in (Polychaeta: Chaetopteridae), based on type specimens and assessing phylogenetic confidence. Molecular Biology and recently collected material from Morib Beach, Malaysia. Evolution 20: 255–266. Pacific Science University of Hawaii Press 53: 24–36. Avise JC. 1994. Molecular markers, natural history and evo- Osborn KJ, Rouse GW, Goffredi SK, Robison BH. 2007. lution. New York: Chapman & Hall. Description and relationships of Chaetopterus pugaporci- Barnes RD. 1965. Tube-building and feeding in chaetopterid nus, an unusual pelagic polychaete (Annelida, Chaetop- polychaetes. Biological Bulletin Marine Biological Labora- teridae). Biological Bulletin 212: 40–54. tory, Woods Hole 129: 217–233. Petersen ME. 1984. Chaetopterus variopedatus (Annelida, Bhaud M. 2005. Evidence of a geographical variation in Polychaeta): another victim of the ‘characteristic species’ Mesochaetopterus (Polychaeta: Chaetopteridae) from the disease. American Zoologist 24: 62A. Pacific Ocean. Journal of the Marine Biological Association Petersen JA, Fanta ES. 1969. On two new species of of the United Kingdom 85: 1409–1423. Mesochaetopterus (Polychaeta) from the Brazilian coast. Blake JA. 1996. Family Chaetopteridae Malmgren, 1867. In: Beiträge zur Neotropischen Fauna 6: 120–136. Blake JA, Hilbig B, Scott PH, eds. Taxonomic atlas of the Pinedo S, Sardá R, Martin D. 1996. Seasonal dynamics and benthic fauna of the Santa Maria basin and western Santa structure of soft-bottom assemblages in the Bay of Blanes Barbara channel. Santa Barbara, CA: Santa Barbara (Western Mediterranean Sea). Publicaciones Especiales del Museum of Natural History, 233–251. Instituto Español de Oceanografía 22: 61–70. Carreras-Carbonell J, Macpherson E, Pascual M. 2005. Posada D, Buckley TR. 2004. Model selection and model Rapid radiation and cryptic speciation in Mediterranean averaging in phylogenetics: advantages of Akaike Informa- triplefin blennies (Pisces: Tripterygiidae) combining mul- tion Criterion and Bayesian approaches over Likelihood tiple genes. Molecular Phylogenetics and Evolution 37: 751– ratio tests. Systematic Biology 53: 793–808. 761. Posada D, Crandall KA. 1998. MODELTEST: testing the Castresana J. 2000. Selection of conserved blocks from mul- model of DNA substitution. Bioinformatics 14: 817–818. tiple alignments for their use in phylogenetic analysis. Potts FA. 1914. Polychaetes from the N. E. Pacific: Chaetop- Molecular Biology and Evolution 17: 540–552. teridae. With an account of the phenomenon of asexual

reproduction in Phyllochaetopterus and the description Sardá R, Pinedo S, Martin D. 1999. Seasonal dynamics of of two new species of Chaetopteridae from the Atlantic. macroinfaunal key species inhabiting shallow bottoms in Proceedings of the Zoological Society of London 67: 955–994. the Bay of Blanes (NW Mediterranean). Acta Oecologica 20: Russello MA, Amato G. 2004. A molecular phylogeny of 315–326. Amazona: implications for Noetropical parrot biogeography, Sendall KA, Fontaine AR, O’Foighil D. 1995. Tube mor- taxonomy and conservation. Molecular Phylogenetics and phology and activity patterns related to feeding and tube- Evolution 30: 421–437. building in the polychaete Mesochaetopterus taylori Potts. Sardá R, Martin D, Pinedo S, Dueso A, Cardell MJ. 1995. Canadian Journal of Zoology 73: 509–517. Seasonal dynamics of shallow-water benthic communities in Swofford DL. 2001. PAUP* phylogenetic analysis using par- the Western Mediterranean. In: Nicolaidou A, ed. The simony (* and other methods), Version 4. Sunderland, MA: biology and ecology of shallow coastal waters. Fredensborg: Sinauer Associates. Olssen & Olssen, 191–198. Thompson JD, Gibson TJ, Plewniak F, Jeanmougin Downloaded from https://academic.oup.com/zoolinnean/article/152/2/201/2630849 by guest on 25 March 2021 Sardá R, Pinedo S, Grémare A, Taboada S. 2000. Changes F, Higgins DG. 1997. The ClustalX windows interface: in the dynamics of shallow sandy-bottom assemblages due ﬂexible strategies for multiple sequence alignment aided by to sand extraction in the Catalan Western Mediterranean quality analysis tools. Nucleic Acids Research 25: 4876– Sea. ICES Journal of Marine Science 57: 1446–1453. 4882.