Taxonomy and distribution of Pectinariidae (Annelida) from Iceland with a comparative analysis of uncinal morphology

Based on samples collected during the BIOICE project off Iceland, four species of marine annelids belonging to the family Pectinariidae were identifi ed: Amphictene auricoma (O.F. Müller, 1776), Cistenides granulata (Linnaeus, 1767), Cistenides hyperborea Malmgren, 1865 and Lagis koreni Malmgren, 1866. Taxonomic remarks and data on geographical and bathymetric distribution are presented. The distribution of each species off Iceland was evaluated and two patterns were defi ned: C. granulata and C. hyperborea were mainly found in waters off the northeast coast, while A. auricoma and L. koreni were found on the southern coast. Several body characters with taxonomic value in this family were reviewed under the stereo microscope and scanning electron microscope, with special emphasis on the neuropodial uncini. Remarks on these special chaetae are included in the diagnoses. European Journal of Taxonomy 666: 1–32 ISSN 2118-9773 https://doi.org/10.5852/ejt.2020.666 www.europeanjournaloftaxonomy.eu 2020 · Parapar J. et al. This work is licensed under a Creative Commons Attribution License (CC BY 4.0). R e s e a r c h a r t i c l e urn:lsid:zoobank.org:pub:2E0FAA1D-DA9A-4486-805F-9DA3DF928539


Introduction
The BIOICE (Benthic Invertebrates of Icelandic Waters) expeditions started in 1992 aiming at characterizing the marine benthic fauna living in the 200-mile economic zone of Iceland in a depth range of 20 to 3500 m. Sampling was carried out at both sides of the Greenland-Iceland-Faeroe Ridge (GIF Ridge) that separates water masses with different oceanographic features (Hansen & Østerhus 2000). In this sense, several water masses associated with submarine ridges are present around Iceland, influencing the flow of several cold and warm water currents (Silva et al. 2017), dividing its coast into three distinct regions (Fig. 1A), and therefore limiting the distribution of the marine species (Brix & Svavarsson 2010;Meißner et al. 2014;Silva et al. 2017).  Müller, 1776). B. Lagis koreni Malmgren, 1866. C. Cistenides granulata (Linnaeus, 1767). D. Cistenides hyperborea Malmgren, 1866. The three main coastal sectors around Iceland are shown in A. The specimens photographed and studied under SEM are indicated with green and red arrowheads, respectively. Abbreviations: H = Höfn; NWF = North-western fjords; RP = Reykjanes Peninsula.
In this paper, we study the diversity and distribution of the Pectinariidae around Iceland after a number of samples collected during the BIOICE cruises. The aforementioned species are reported here again for Iceland and remarks about their taxonomy and ecology are provided. Furthermore, we have also studied the uncini in specimens of different body sizes for all four species; this was done by using the SEM in order to test 1) whether there were morphological differences supposedly related to ontogeny and 2) whether this uncini variability showed any pattern both within a species and among species and therefore could be used as a taxonomic character.

Material and methods
This study is based on material collected in the framework of the BIOICE project. A total of 1066 pectinariid specimens were found in 43 samples; some specimens in eight samples could not be identified because of their small size or bad preservation (e.g., fragmented body). Samples were initially fixed in 10% formalin buffered with borax, and then preserved in 70% ethanol. Specimens were picked from PARAPAR J. et al., Taxonomy and distribution of Pectinariidae (Annelida) from Iceland samples by the staff of the Sandgerdi Marine Centre (SMC, Iceland). Most of the material examined was deposited in the collections of the Icelandic Institute of Natural History (IINH, Reykjavík); several selected specimens were deposited in the Museo Nacional de Ciencias Naturales (MNCN, Madrid).
Abiotic data of all BIOICE samples are available from the following link: http://utgafa.ni.is/greinar/BIOICE_station_list_91-04_Paper_A2.pdf Identification of specimens and stereomicrographs were done using an Olympus SZX9 stereo microscope at the Facultade de Ciencias, Universidade da Coruña (UDC, Spain). Specimens used for examination with SEM were prepared by critical point drying, covered with gold in a BAL-TEC SCD 004 evaporator and examined and photographed under a JEOL JSM-6400 scanning electron microscope (SEM) at the Servizos de Apoio á Investigación (SAI-UDC, Spain).
Nomenclature for morphological characters, genera, species and synonyms mostly follows Holthe (1986) and Hutchings & Peart (2002). Each species is related to water mass(es) as described by Brix & Svavarsson (2010) according to geographic location (coordinates), depth and temperature.

Taxonomy
The study of the samples collected during the BIOICE project yielded four pectinariid species belonging to three genera. The most abundant species was Lagis koreni (801 specs; 75.1% of the total pectinariid specimens identified), followed by Amphictene auricoma (117 specs; 11.0%), Cistenides hyperborea (125 specs; 11.7%) and C. granulata (23 specs; 2.2%). All four species show a well-defined geographical distribution around Iceland corresponding to narrow ranges of temperature and depth, but being mostly present in temperate waters at depths of between 100 and 300 m (Figs 1-2); there were some exceptions to this pattern corresponding to samples with C. hyperborea and L. koreni. Abundance and distribution around Iceland of each species and taxonomic remarks are presented and discussed below.

Remarks
The genus is characterized according to Hutchings & Peart (2002) by having neurochaetal uncini with major teeth arranged in two rows. However, an examination of BIOICE specimens of A. auricoma suggests that this character could be reassessed (see Discussion below).  (Linnaeus, 1767

Remarks
Species of Amphictene bear a characteristic crenulated opercular rim, that is not present in other genera (Figs 4A, 5A). Amphictene auricoma -type locality: Denmark (Holthe 1986) -has been the only species of the genus reported in North Atlantic waters and also bears thick paleae with sharp tips (Fig. 5A), scaphe with crenulated margins (Fig. 6D), 17 uniramous and 13 biramous chaetigers, and distally curved scaphal hooks (Fig. 6E). However, Hutchings & Peart (2002: table 1) pointed out that there is "a considerable variation recorded for this species" in the North Atlantic. General morphology of BIOICE specimens agree well with previous descriptions and the aforementioned characters; however, SEM examination revealed that several pygidial and uncini features differ from what was described for A. auricoma. First, Holthe (1986: fig. 4c) and Hartmann-Schröder (1996: fig. 234c) mentioned a long "anal tongue" and "anal cirrus", corresponding to the "anal lobe" and "anal papilla" (after Hutchings & Peart 2002), that are much shorter in BIOICE specimens (Fig. 6D).
On the other hand, the uncini of A. auricoma have been described as having 1-2 vertical rows of main teeth ( . This pattern is even more evident in the last unciniger, where the capitium teeth seem mixed with those of the subrostral process and leaving a small opening through which the tip of the rostrum is visible (Fig. 7E).
Finally, we have found two longitudinal rows of ciliated tufts/patches along the whole ventral body surface ( Fig. 5B) that have not been mentioned in the literature to the best of our knowledge, and those tufts may be related to water irrigation within the tube.

Remarks
All diagnostic characters for this species -type locality: Northern Europe (Holthe 1986) -as stated by Holthe (1986) are present in BIOICE material: shape and number of paleae and cephalic veil (Figs 4C, 8A, 9A), scaphal shape, anal lobe and anal papilla (Fig. 8F). Ventral uncini agree well with those present in the genus, not showing variations within uncinigers nor along the body both in large ( Fig. 8B-E) and small ( Fig. 9B-F) specimens; there is only one row of teeth becoming progressively smaller in size from the uncinus base (at the level of the subrostral process) towards the distal end. This lack of variation, which has been considered the typical pattern in the family, was only found in C. granulata among the four pectinariid species in BIOICE samples.

Remarks
Cistenides hyperborea -type locality: Greenland and Spitzbergen (Holthe 1986) -is a poorly known species; the original description does not indicate the number of rows of teeth present on the uncini, although 1 to 3 rows have previously been mentioned for specimens assigned to this taxon (Hutchings & Peart 2002). Because type material was never designated (Holthe 1986), Hutchings & Peart (2002) suggested that specimens from the type locality need to be re-examined to ascertain the generic identity of this species.
The examination of the Icelandic specimens here identified as C. hyperborea, covering a wide range of body sizes, shows a considerable variation in the shape of the neuropodial uncini, contrary to the expected, according to the description of the species. BIOICE specimens agree with previous descriptions (e.g., Pettibone 1954;Holthe 1986;Hutchings & Peart 2002) in: 1) having a smooth dorsal brim and cephalic veil free from operculum, provided with several well-defined cirri (Fig. 10A); 2) the relation between chaetigers with notochaetae versus those with also ventral uncini is 17/12; 3) the paleae are long, numerous, straight and with pointed tips (Figs 10A, 12A); 4) the scaphe has small lateral lobes (Fig. 11A) with a short and rounded anal lobe and anal papilla (Fig. 11B), and scaphal hooks are distally bent (Fig. 11C-E). On the other hand, the aforementioned characters show, however, differences related to the size of the specimen. For instance, medium-sized and small individuals show a strait scaphe border with a verrucose anal lobe ( Fig. 11E-F), thinner paleae (Figs 12A, 13A) and the tip of the scaphal hooks is strongly curved (compare Figs 12F, 13F, 14F).
The number and arrangement of teeth in the neuropodial uncini shows differences among BIOICE individuals, contrary to that observed in C. granulata. Two different types of uncini can be distinguished: 1) Uncini with 2-3 main teeth arranged in a vertical row. This is the typical teeth arrangement in the pectinate type for this species and the genus as well. We found this pattern in large specimens (Figs 3D, 10).
2) Uncini provided with a large basal tooth (rostrum) and an upper group of short teeth (capitium). This agrees with the typical avicular type present in many species of Terebellomorpha and found here in smaller specimens (Figs 3D, 12-13).
SEM examination of BIOICE specimens seems to support the hypothesis that these differences are related to body size even though we did not examine the whole range of sizes (Fig. 3D). Uncini of large specimens (sample 3252, Fig. 10) do not show a variation either within an individual torus or along the body; on the contrary, mid-sized specimens (sample 2660, Fig. 12) show a progressive increase in the number of upper teeth, mostly in ventral uncini (Fig. 12C, E) and especially in posterior chaetigers (Fig. 12E), the latter having a well-defined capitium; dorsal uncini show, in turn, the usual shape as   found in the genus (Fig. 12B, D). This pattern of variation is more evident in small specimens (sample 2041, Fig. 13), where the dorsal uncini are also of avicular type (Fig. 13B-D), similar to ventral ones (Fig. 13E). Therefore, these results suggest that the BIOICE specimens of C. hyperborea show changes in the uncini shape during ontogenetic development and that the uncini type varies accordingly depending on the age of the specimen. This variation, as explained above, occurs within a torus and along the body; thus, as small specimens with the avicular type of uncini grow, and uncini are being replaced along the torus, there is a progressive decrease in the number of upper teeth and a loss of the rostrum finally resulting in a pectinate type with the typical formula of the genus (i.e., MF:1:1:1).
Specimens from sample 2060 (Fig. 2D) found in a shallow bottom (48 m) in North Iceland, were also identified as C. hyperborea. Nevertheless, they show thicker paleae (cf. Fig. 14 vs Fig. 13) than similarsized specimens from other samples, and dorsal uncini are similar to those of larger specimens found elsewhere (Fig. 14B, D), while ventral ones (Fig. 14C, E) correspond to those of specimens of their size such as those of sample 2660 (Fig. 12C, E). Finally, scaphe hooks appear more strongly curved than would be expected for this species (cf. Fig. 14F vs Figs 11E, 12F, 13F). Therefore, this variation suggests that specimens from sample 2060 may correspond to another species (probably undescribed yet) within a potential species complex present in NE Atlantic waters, that should be assessed in future in combination with molecular studies.
Finally, SEM micrographs show that the tip of the notochaetae is serrated (Fig. 15A-B), and reveal the presence of perforations in the cuticle of the dorsal region anterior to the scaphe (Fig. 15C-D) and of two types of ciliated fields: 1) rounded fields located before the scaphe and among the cuticular perforations ( Fig. 15C-D) and 2) one large field located dorsal to the anal lobe and anterior to the anal papilla ( Fig. 15E-F). These fields may be related to water irrigation, but certainly a future in-depth study is necessary.

Remarks
The diagnosis of the genus by Hutchings & Peart (2002) states that neurochaetal uncini have large teeth arranged in two or more rows. Examination of BIOICE specimens of L. koreni shows again that there are changes in teeth shape with increasing size of an individual, reflected in a progressive reduction in the number of teeth rows with the age of the animal. This still fits the diagnosis of the genus ("two or more rows") but a better definition of uncini shape is needed. In contrast to the other species studied here, the uncini are, in all body sizes studied, of the pectinate type (see below).

Remarks
Body and chaetal features match well those of L. koreni provided by Holthe (1986), including the shape and number of paleae (Figs 4I, K, 16A, 17A), smooth opercular rim (Fig. 4I, K), cirrate cephalic veil ( Fig. 17A-B), a crenulated scaphe (Fig. 16F) provided with slightly distally curved hooks (Figs 16F insert, 17F), and 16 uniramous and 12 biramous segments (see Holthe 1986). However, SEM examination again reveals differences between the uncini features and those reported in the literature, that seem related to morphological variability according to size/age. Large and mid-sized specimens ( Fig. 16C-E) bear uncini as described in the literature, i.e., having 3-4 vertical rows with many teeth (6-8) of similar size each and showing no variation across the body or within an individual torus; by contrast, uncini of small specimens (Fig. 17C-E) have more rows of teeth (5-7) in frontal view ( Fig. 16E vs Fig. 17D-E).
Finally, SEM micrographs also show circular areas under the parapodia with segmentary arrangement (Fig. 16B); these are likely nephridial or glandular and may have a role in tube building, although this should be confirmed in future morphological studies.   Hutchings & Peart (2002), in their revision of the Australian Pectinariidae, stated that "Individuals belonging to the Pectinariidae are easy to recognise, but the systematics of this widely distributed family has been neglected. The confused identity of the common European species suggests that many species may have been misidentified or perhaps several species confused under one name". This is also true for the NE Atlantic fauna, where the most recent monograph was provided by Holthe (1986); this contrasts with other areas such as China, Brazil and Australia where knowledge on the family has substantially been improved recently (e.g., Sun & Qiu 2012;Zhang & Qiu 2017;Zhang et al. 2015Zhang et al. , 2019Nogueira et al. 2019).

Discussion
vertical series of teeth while in young ones there is a large number arranged in transverse series. The aforementioned authors also affirm that the extent of this variability, both within the same neuropodial row and between segments along the body, has never been widely appreciated, although Duchène & Bhaud (1988) previously reported similar patterns in six terebellid species. Because of that, and its apparent correlation with other relevant taxonomic characters, Glasby & Glasby (2006) proposed to consider these variations in future phylogenetic studies of the group. Garraffoni & Camargo (2006) also stated that Terebellidae uncini "… is a rich source of taxonomic information that can be used to help assess evolutionary relationships within the family". Our study demonstrates that this approach should be extended to the Pectinariidae in general, and especially to the genus Cistenides.
Recently, Nogueira et al. (2019), following Hutchings & Peart (2002), considered the uncini of Cistenides as being "pectinate, with single longitudinal row of teeth" and reported for the first time in the family the presence of avicular uncini when diagnosing the genus Petta: "the uncini are breviavicular rather than bipectinate (sensu Holthe 1986), with transverse series of progressively shorter distal wards secondary teeth on top of the main fang." (Nogueira et al. 2019: 491). Our observations suggest that: 1) in Cistenides the uncini type change as the specimen grows, and this should be considered in the diagnosis of the genus; 2) the presence of avicular uncini in Petta should be also confirmed in small specimens; 3) avicular uncini are not exclusive of Petta but are also present in Icelandic specimens of C. hyperborea as demonstrated here for young specimens of the latter. Nogueira et al. (2019) also pointed out that SEM examination is needed to truly assess the uncini features, by comparing SEM micrographs of Petta alissoni Nogueira et al., 2019 with those obtained with light microscope (cf. Nogueira et al., 2019: fig. 11d vs fig. 12f-g). Thus, images obtained with the latter show similar uncini to those described and illustrated by Holthe (1986: fig. 7f) as having a "rostrum surmounted by a large tooth", but the use of SEM demonstrated that this is a wrong interpretation of the actual uncini shape. Unfortunately, specimens of Petta pusilla Malmgren, 1866, the type species of the genus, were not found among the BIOICE material in order to test this hypothesis and therefore future work is needed to address this.

Dorsal chaetae versus ventral uncini (dc/vu)
Again, following Hutchings & Peart (2002), some generic characters seem to overlap in Pectinariidae, and therefore it seems advisable to find additional diagnostic elements. In fact, these authors propose to include the number of uniramous vs biramous segments, i.e., the number of chaetigers with dorsal chaetae (dc) vs those with ventral uncini (vu), suggesting the following values for each genus: Amphictene (17/13), Cistenides (17/13), Lagis (16/12), Pectinaria (17/13) and Petta (18/14), although they recognize that the validity of this character needs further testing because some species have reduced posterior notopodia that may easily be overlooked. The study of BIOICE specimens of Cistenides confirms the actual number of chaetigers with thoracic notochaetae (17), but we also suggest to modify the number of neuropodial uncinigerous tori from 12 to 13 (see below). Hutchings & Peart (2002) include the character "notochaetal ornamentation" among those most relevant for Pectinariidae genera; in the case of Cistenides, this feature was not described. Our examination of C. granulata and C. hyperborea reveals that the distal half of the notochaetae is clearly serrated and this should be added to the genus diagnosis. The same authors also stated that the actual size of the opercular paleae may not be diagnostic because they may be worn by digging and can presumably be replaced during life. From our examination of BIOICE specimens we think that, even considering that the distal end may be worn, other characters remain unchanged during development, including length, width or shape of distal end of paleae. In fact, the paleal general shape is proposed here as a reliable cue to recognize potential different species of C. hyperborea.

PARAPAR J. et al., Taxonomy and distribution of Pectinariidae (Annelida) from Iceland
Delimitation of genera Following Hutchings & Peart (2002), the characters traditionally used to separate the pectinariid genera are: 1) the degree of fusion of the cephalic veil to the operculum; 2) the marginal ornamentation of the cephalic veil and the operculum; 3) the number of uncinigerous segments; 4) the number of vertical rows of teeth on the uncini, and 5) the degree of separation of the scaphe from the abdomen. As we noted above, Hutchings & Peart (2002) also included details of the notochaetae for the genera represented in Australian waters. Hutchings & Peart (2002) diagnozed the genus Cistenides as having 13 uncinigers and proposed to transfer several species to Pectinaria because of having two rows of teeth in uncini instead of one. Following Gil (2011), the genus Cistenides as erected by Malmgren (1866) was originally diagnosed as having 12 uncinigers; this diagnosis was followed by many authors such as Fauvel (1927) and Annenkova (1929) and is also followed here. Gil (2011) revised the descriptions of the species attributed by Hutchings & Peart (2002) to this genus (some described as having 13 uncinigers while others with 12), and considered that the genus should be referred as presenting 12-13 uncinigers; however, Gil (2011) also provided a key for pectinariid genera where Cistenides is characterized as having 12 uncinigers and Pectinaria 13 because this condition was found in specimens of both European species of this genus (both present in BIOICE material), one of them being C. granulata, the type species of the genus.
Gil (2011) agrees with Hutchings & Peart (2002) in considering the number of rows of teeth in the uncini as a relevant diagnostic character at the genus level and thus uncini of Cistenides would have large teeth arranged in 1-2 rows. In this context, we suggest that this character only corresponds to large specimens and until the ontogeny of all species of Pectinariidae is known, the shape of the uncini should be used with caution when separating genera and species.
It is likely that an in-depth revision of the family in the NE Atlantic, including a molecular approach, may reveal a higher diversity. For instance, Jolly et al. (2005) observed different patterns of genetic structure in different populations for L. koreni along Brittany and the English Channel; they found a "surprisingly deep phylogeographic break" which was associated with a biogeographic boundary along the western coast of Brittany and suggested the occurrence of potential cryptic or sibling species. In the same vein, Nygren et al. (2018) demonstrated the presence of more than 25 species of the genus Terebellides Sars, 1835 (Trichobranchidae) along the NE Atlantic coast.