Taxonomy of North European Lumbricillus (Clitellata, Enchytraeidae)

Abstract Lumbricillus is a genus of clitellate worms with about 80 described species that inhabit marine and limnic habitats. This study follows a recent analysis of the phylogeny of the genus based on 24 species of Lumbricillus collected mainly in Norway and Sweden. We provide the illustrated taxonomic descriptions of all these species and describe two of them as new; Lumbricillus latithecatus sp. n. and L. scandicus sp. n. Using the recent phylogeny, we informally divide Lumbricillus into five distinct morphological groups, into which we also tentatively place the Lumbricillus species not included in this study. Furthermore, we establish Claparedrilus gen. n., with the type species C. semifuscoides sp. n., and transfer Pachydrilus semifuscus Claparède, 1861 (previously referred to Lumbricillus) into said genus.


Introduction
Enchytraeids (Annelida, Clitellata, Enchytraeidae) are small clitellate worms that mainly inhabit terrestrial soils but the family is well represented in the aquatic environment. One of the about 30 genera, Lumbricillus Ørsted, 1844, is primarily found in marine and freshwater habitats, but also in humid soils (Nielsen and Christensen 1959). It was established by Ørsted (1844), and Lumbricus lineatus Müller, 1774, the first enchytraeid ever described (Michaelsen 1889), was later regarded as its type species (Brinkhurst and Jamieson 1971;Coates and Ellis 1981). Ørsted (1842) had earlier referred Lumbricus lineatus to one of three groups constituting the "Lumbricillae", and specifically to the group distinguished by having short, almost straight chaetae, "resembling stitching awls", in both upper and lower bundles. He would later (Ørsted 1844) name this group Lumbricillus. Today, after several emendations, the genus is conceived as having straight to sigmoid chaetae, nephridia with short anteseptale consisting of nephrostome only, and testes enclosed in lobed peritoneal sacs (testis sacs) which, for most species, are structured as regular bunches (Nielsen and Christensen 1959).
Returning to the 19 th Century, Claparède described his Pachydrilus Claparède, 1861 as an assemblage of marine littoral species lacking "hair bristles", but having a single pair of spermathecae in segment V, clitellum covering segments XI-XIII, male pores in XII, and simple vascular and nervous systems. Pachydrilus (and its five representatives) shared these basic traits with the terrestrial species then classified in Enchytraeus Henle, 1837, but was distinguished by the lack of dorsal pores and by generally possessing red-colored blood. Pachydrilus was later redefined by Vejdovský (1879) to include only species with sigmoid chaetae and small nephridial anteseptals and then further restricted by Roule (1888Roule ( , 1889 and Michaelsen (1888Michaelsen ( , 1889, with a diagnosis focusing on the possession of multilobed testes, which led to the selection of P. verrucosus Claparède, 1861 as the type species. It must be mentioned that the very same species epithet had been used by Ørsted in 1844 for his Lumbricillus verrucosus, but due to the lack of an adequate description, it was considered a nomen nudum and incertae sedis (both Vejdovský 1884:45 andMichaelsen 1900:51 believed it should be placed among Tubificidae). The present study will treat P. verrucosus Claparède, 1861 as a valid enchytraeid species.
The name Pachydrilus soon became a competitor for Lumbricillus in the taxonomic literature. Some scientists favored Pachydrilus (Michaelsen 1888(Michaelsen , 1889Ude 1929;Černosvitov 1937), others Lumbricillus (Eisen 1904;Southern 1909;Stephenson 1911Stephenson , 1930. In 1900, Michaelsen attempted to resolve the conflict by accepting the seniority of the name Lumbricillus and placing Pachydrilus verrucosus within this genus (Michaelsen 1900) and proceeded to use this name in following publications (Michaelsen 1905(Michaelsen , 1911. However, he later went back to using Pachydrilus instead of Lumbricillus (Michaelsen 1925(Michaelsen , 1929(Michaelsen , 1934(Michaelsen , 1935; see Rota et al. 2008). The use of two competing names ended in 1959 when Nielsen and Christensen synonymized P. verrucosus with L. lineatus, directly rendering Pachydrilus a junior synonym of Lumbricillus; P. verrucosus had previously been suggested as a form of L. lineatus by Ude (1929) and Černosvitov (1937). Lumbricillus verrucosus has recently been resurrected as a separate species with molecular support (Klinth et al. 2017) and will be given a more extensive morphological description in this paper. It is important to note that by re-instating L. verrucosus as a valid species, we do not change the status of Pachydrilus as a junior synonym to Lumbricillus, as L. verrucosus and L. lineatus, using genetic data, have been found to be closely related to each other (Klinth et al. 2017; see also Fig. 1 herein).
In 1885, Saint-Loup described the species Pachydrilus enchytraeoides Saint-Loup, 1885 from the Marseille harbor. The rather brief description mentioned pouch-like spermathecae, a simple circulatory system and irregularly lobed testes (Saint-Loup 1885). The combination of these characters, interpreted by Saint-Loup's fellow researcher Roule to be intermediate between Lumbricillus and Enchytraeus, convinced Roule (1888) to transfer the species to a new genus, which he named Enchytraeoides Roule, 1888 in acknowledgement of Saint-Loup's work. However, to avoid tautonymy (i.e., when genus name and species epithet are identical), he also unconventionally changed the species epithet to marioni substituting Pachydrilus enchytraeoides with Enchytraeoides marioni. Furthermore, Roule provided an extensive description of the embryology and development of his species along with over a hundred illustrations (Roule 1889). The male anatomy of E. marioni is in many ways reminiscent of some species of Lumbricillus, such as L. arenarius (Michaelsen, 1889), in lacking the typical regularly lobed testis sacs of most Lumbricillus and having sperm funnels several times longer than wide. In fact, L. arenarius has been placed in Enchytraeoides by some authors (Ude 1929;Knöllner 1935;von Bülow 1957). It is not completely clear if Pachydrilus enchytraeoides and Enchytraeoides marioni are indeed the same species. Regardless, the descriptions of both species fit within Lumbricillus, and thus Enchytraeoides should be considered a junior synonym of Lumbricillus (Stephenson 1930). A more extensive account on the intricate taxonomical history of Enchytraeoides and other genera can be found in Rota et al. (2008).
Among the new species described by Claparède (1861) when he established the genus Pachydrilus was P. semifuscus Claparède, 1861. This species was noted to have the body colorless anteriorly and brown posteriorly (due to the strongly pigmented chloragogen tissue), conspicuous nephridia, solid unlobed testes and huge penial bulbs. The species was later transferred to Marionia (a name soon after replaced by Marionina) by Michaelsen (1889) and remained there until Nielsen and Christensen (1959) moved it to Lumbricillus. The brief original description was amended partly by Southern (1909) and more extensively by Stephenson (1911), who reported 4 to 5 pairs of pharyngeal glands, along with details of the chaetae, brain, copulatory glands and more. The species was also described from Iceland by Erséus (1976), who observed that the anteseptale of the nephridia might contain more than just a funnel (also a few coils). It has later been noted that this species (as well as others) lack the typical lobes of the testes and thus might not fit in Lumbricillus (Kossmagk-Stephan 1983).
The molecular phylogenetic study by Erséus et al. (2010) documented that Lumbricillus is a non-monophyletic group, and that a group of species (including L. arenarius) is sister to Grania Southern, 1913 rather than to the remaining Lumbricillus. However, the more recent analysis of the North European Lumbricillus by Klinth et al. (2017) showed ambiguity as to whether three Lumbricillus species (L. arenarius, L. dubius (Stephenson, 1911) and an unidentified species) comprise the sister group of the remaining Lumbricillus or the sister to Grania, where they together with Grania in turn would be the sister to Lumbricillus sensu stricto. The same study also found strong evidence for excluding what they refer to as "L. semifuscus" from Lumbricillus sensu stricto (Klinth et al. 2017). Lately, with increased taxon sampling that species (which we here describe as C. semifuscoides sp. n.) has been found to be more closely related to Globulidrilus Christensen & Dózsa-Farkas, 2012and Bryodrilus Ude, 1892.
The aim of this study is to increase the knowledge of the taxonomy of North European Lumbricillus, based on the most recent phylogenetic reconstruction and molecular delimitation of species (Klinth et al. 2017), and herein, by the addition of morphological studies. However, we will not use the morphological characters for any phylogenetic analysis. We will provide illustrated descriptions of all the included species and establish three new species and one new genus and re-describe L. helgolandicus (Michaelsen, 1927), in order to clear up parts of the taxonomy of this poorly studied group.

Material and methods
Worms were collected in marine, brackish and limnic habitats, mainly in Norway and Sweden (Appendix 1), by decantation of suspended organic material from bottom substrates and killed and preserved in 80 % ethanol. After sorting under a dissecting microscope, the posterior end of each worm was cut away for DNA analyses (partly published by Klinth et al. 2017) while the anterior end was stained in paracarmine, dehydrated in xylene and mounted in Canada balsam on microscope slides; for a more detailed description of procedures see Erséus (1994). Using a compound microscope, specimens on slides were identified to species using the primary taxonomic literature, and largely adhering to the list of accepted species in Schmelz and Collado (2012), with the following exceptions: L. aegialites Stephenson, 1922and L. georgiensis Tynen, 1969, both synonymized with L. pagenstecheri (Ratzel, 1868 by Coates & Ellis, 1981, L. magdalenae Nurminen, 1965 which we consider a synonym of L. arenarius (Michaelsen, 1889), and finally L. cervisiae Kossmagk-Stephan, 1983 and L. christenseni Tynen, 1966, which we consider separate species from L. knoellneri Nielsen & Christensen, 1959 with which they were previously synonymized. In the species descriptions we provided brief chresonymy lists of the references we found most relevant for each species. Morphological characters were drawn using a camera lucida and the images were treated by Gimp 2.8.10 software. All specimens studied are vouchers of DNA sequences (Appendix 1), including the COI-barcodes generated by Klinth et al. (2017) study, which were also used in this study to find matches between our specimens and records in the Barcoding of Life Database (BOLD), to better estimate the geographical ranges of the species. Where available, Barcode Index Numbers or BIN:s have been noted in each species description. These BIN:s refer to clusters of COI-barcodes on BOLD that are considered to comprise specimens of the same species. Finally, the only remaining syntype of Pachydrilus helgolandicus Michaelsen, 1927, herein designated as the lectotype, was borrowed from the Zoological Museum in Hamburg (ZMH). It has now been mounted and morphologically examined. Types and other voucher specimens are deposited in the Swedish Museum of Natural History (SMNH) and the Zoological Museum, University of Bergen (ZMBN).

General notes
All descriptions are based on fixed worms mounted on slides. This has some disadvantages for discerning the shape of certain internal organs such as the nephridia, but is not an unusual method for marine worms, and it improves the description of other characters such as chaetae. Nevertheless, morphology can differ from descriptions in the literature based on living specimens. Welch (1914) noted that fixation in alcohol can reduce Lumbricillus body length from about 15-19 mm in living specimens, to about 9-14 mm in fixed specimens. Finogenova and Streltsov (1978) noted that the ratio between the length and width of the sperm funnels was similarly reduced in fixed specimens. They reported sperm funnels about 2-4 times longer than wide in living specimens and a ratio of about 1.2-1.5:1 after fixation. Furthermore, Finogenova and Streltsov (1978) as well as Southern (1909) reported the ratio length:width of the sperm funnels to vary in living specimens due to body contractions. Southern (1909) further questioned the importance of the midventral subneural glands (previously referred to as copulatory glands) for separating species, as these glands seem to vary in size and sometimes in segmental distribution between individuals of the same species. We also observed great variation in these features. Lastly, as the colour cannot be distinguished after the staining, remarks on the colour of the worms is based on the notes made by the collector, prior to preservation and mounting.
All specimens in this study are amputated of their posterior segments (used for DNA extraction). Therefore, comparisons of total length and segment number with original descriptions have not been possible. When available, the length of the fifteen first segments as well as the width at the clitellum of the worms has been used to compare the general body size of the species.
In fixed Lumbricillus worms the origin of the dorsal vessel can be difficult to establish since vessel expansions are more or less conspicuous according to the peristaltic movement of the blood at the time of fixation. Thus, due to the varying conditions when animals were killed and fixed, the dorsal vessel may appear to originate in different segments.  Klinth et al. (2017), together with comparisons of some morphological characters. Furthermore, the Lumbricillus species not included in the present study are listed with tentative placements to one of these groups. Note that it is likely that some of the species not studied could upon closer examination prove to be synonyms to the included species in this study. * L. latithecatus sp. n. was referred to as L. sp. E in Klinth et al. (2017). ** L. scandicus was referred to as L. cf. helgolandicus in Klinth et al. (2017).  Fig. 1). The lineatus, pagenstecheri and "tuba" groups all have testis sacs with several large lobes in a bunch-like arrangement, characteristic for the majority of Lumbricillus species. The testis sacs of the buelowi and arenarius groups appear as a more or less compact irregular mass, which can still be lobed but not bunch-like. Interestingly, we noted that in the former three groups the upper and lower chaetal bundles are arranged almost symmetrically around the body (as dorsolateral and ventrolateral bundles), whereas in the buelowi and arenarius groups the upper bundles tend to be closer to the lateral lines (observed by all three authors). The number of chaetae varies within each group and is usually 3-6 in each bundle but can reach 10 or more, except in the buelowi and arenarius groups where there are rarely more than 2-3 chaetae per bundle. The length/width ratio of the sperm funnels varies within and among the species in each group, but the funnels are usually only a few times longer than wide in all groups except for the lineatus and arenarius groups, where they can be 5-10 times longer than wide. The penial bulbs are round and compact in all groups except the arenarius group where they can also be bilobed. Finally, the spermathecae are spindle-shaped without a clear distinction between the ectal duct and the ampulla in the lineatus group; have a clear distinction between ectal duct and ampulla in the pagenstecheri, "tuba" and buelowi groups, with gland cells also along the duct in the pagenstecheri group; or have a gradually widening ectal duct, making the ampulla more or less indistinct in the arenarius group (Table 1, Fig. 1). Using these combinations of characters we made a preliminary placement of the remaining Lumbricillus species, not studied in this paper, into any of these five groups (Table 1). Note that, as we have not been able to include these species in our molecular phylogeny, we cannot be certain that our species groups would remain monophyletic with the addition of these species.  Klinth et al. (2017). Species tree based on 12S, 16S, COI, 18S, 28S, ITS and H3 genes, estimated using Bayesian inference under the multispecies coalescent model in *BEAST. Posterior probabilities higher than 0.9 shown as support values. Scale shows expected number of changes per site in COI with all other genes relative to it. Lumbricillus tuba has been added to the tree at the most probable position given the other gene and species trees from Klinth et al. (2017). The tree depicts the five morpho-groups that are used in the present study to discuss the relationships within Lumbricillus. The general morphology of the spermathecae and testes/testis sacs of these groups are also shown.
Geographical distribution including BOLD data. Genetically identified from the Netherlands, Norway, Spain and Sweden; also recognized from Canada (BIN-numbers BOLD:AAF9630 & BOLD:ACV7068). This species has historically been widely reported from Europe and North America, and even from the southern hemisphere (Stephenson 1932b).
Remarks. Lumbricillus lineatus, possibly the first ever described enchytraeid, has an interesting history and was for a long time poorly defined as a species. It was given the name Lumbricus lineatus by Müller (1774) who described it as abundant on the shores of the Baltic Sea. Müller classified the worm under "Setosa", and gave a brief description of its circulatory system and short protruding chaetae. In the same work, he refers to a worm that he previously described (Müller 1771), which he found on sandy shores of the Baltic and particularly among the rotting seaweed by the ramparts of Copenhagen, where he did most of his work. However, he classified that worm in Gordius, referred to it as a Faden Wurm (nematode) said to lack any segments or ring, and he failed to mention anything about chaetae. It is therefore difficult to say if Müller simply missed those characters or if the worm from 1771 was something different. Nevertheless, as Müller worked with live material, it is most plausible that the type locality for the first described Lumbricillus lineatus is around Copenhagen. Unfortunately, we do not have any specimens from Copenhagen, but we have found that our molecularly defined L. lineatus is common throughout the Baltic Sea. For the sake of finally defining the true L. lineatus and connecting it to a molecular profile, we decided to designate one of our specimens from Öland in the Baltic Sea as a neotype.
The specimens of L. lineatus in this study were smaller than the ones in the re-description by Nielsen and Christensen (1959) and the sperm funnels were shorter in relation to their length. However, the shape of the spermathecae, number of chaetae and segments conformed well with these authors' description. Furthermore, both the vesiculate and avesiculate forms of the species were observed, as has previously been noted to occur in L. lineatus (Nielsen andChristensen 1959, Christensen 1961), further supporting the designation of a neotype from our material. Christensen and O'Connor (1958) were the first to describe diploid vesiculate and triploid avesiculate forms of L. lineatus and their intriguing life histories (see also Christensen 1960). While the diploid form has normal gametogenesis and is amphimictic, the triploid form is dependent on copulation with the diploid form, which acts as sperm donor, to produce offspring. Furthermore, the sperm from the diploid form does merely activate the egg of the triploid individual, without fertilizing it. Instead, oogenesis in the triploid cytotype follows a peculiar pattern of chromosome divisions and mergings that results ultimately in the restoration of triploid nuclei. The relationship between the two forms has been described as an obligatory co-existence (Christensen 1960). To further complicate things, tetra-and pentaploids have also been observed within L. lineatus. The tetra-and pentaploids have testis sacs that are smaller than those of diploids but larger than those of triploids. Furthermore, the sperm funnels are smaller in tetra-pentaploids than in di-and triploids. Apparently, the testes sacs of tetra-and pentaploids sometimes produce sperm which can activate the eggs of all polyploid forms (Christensen et al. 1978). Unfortunately, we did not determine the ploidy level of our sampled specimens, but we did not observe any genetic distinction between the vesiculate and avesiculate forms, neither in COI nor in ITS sequences (Klinth at al. 2017).
Geographical distribution including BOLD data. Originally described from USA, now genetically identified from Norway, Sweden and United Kingdom; also recognized from Canada and Switzerland (BIN-number: BOLD:ACV8942).
Remarks. Our newly sampled material matches Welch's (1914) description of Lumbricillus rutilus well in all characters, except for the sperm funnel shape, where our specimens had funnels less elongate in relation to their width. The material in this study was collected in Sweden and the United Kingdom, whereas this species was originally described from a sewage treatment plant in Chicago, USA. Interestingly, some of our sampled specimens also come from two such plants, in Sweden and the UK, respectively. The species was additionally collected in littoral and freshwater environments in Sweden, and it is likely to be an opportunist that thrives in nutrient-rich habitats. Specimens found in the treatment plants showed increased body size and reduced number of chaetae per bundle, compared to the specimens sampled in the sea, possibly a side effect of living in such rich environments.
Lumbricillus rutilus is genetically most closely related to L. lineatus and L. latithecatus sp. n. (Fig. 1). However, morphologically it is more similar to L. rivalis, particularly in the shape of the spermathecae. Lumbricillus rutilus has on average fewer chaetae per bundle compared with L. rivalis and sperm funnels that are shorter in relation to their width. Paratype. ZMBN 107941 (CE12042), one whole-mounted sexually mature specimen from the type locality.
Etymology. Named from the Latin latus meaning wide and theca for spermatheca. Diagnosis. This species can be distinguished from other Lumbricillus species by the shape of the spermathecae, which do not gradually widen from the ectal pore but instead originates from a very wide pore followed by an ectal duct and ampulla of even width throughout. This makes the duct and ampulla of the spermathecae virtually indistinguishable. There is at least a superficial similarity to the spermathecae of Lumbricillus lineatus and L. verrucosus with a midway constriction or bend and sperm aggregated in the ental part of the ampulla (Fig 3). However, the spermathecae of L. lineatus and L. verrucosus have ectal pores that are much smaller than the diameter of their ampullae, giving the impression of a rapid widening of the spermathecae after the ectal pore, even though the duct and ampulla are difficult to distinguish also in these species.
Male genitalia paired (Fig. 5D). Testes originating in XI, extending forwards into IX, with testis sacs forming regular club-shaped lobes. Sperm funnels in XI, sometimes extending into X or XII, 360-1300 µm long, 155-235 µm wide, making them about 2-6 times longer than wide, funnels tapering towards vasa deferentia. Most of vasa irregularly coiled in XII, 15-30 µm wide. Penial bulbs round, 155-285 µm in diameter, possibly with a small ventral lobe set off from the rest of the bulb. Ovaries in XII. One to four mature eggs present at a time.
Spermathecae (Figs 3B, 5C) in V, pouch-shaped, without distinct ampulla. Ectal duct seemingly indistinguishable from ampulla as the wide pore is followed by a lumen that remains about the same width or possibly widening slightly. Ectal pore surrounded by mass of gland cells forming compact body 140-325 µm in diameter at its widest part. Ampulla, with possible constriction midway dividing it into two sections; ental connection with oesophagus. Sperm filling middle of ectal duct, heads of spermatozoa embedded and forming aggregates mainly in ental part of ampulla. Each spermatheca altogether 220-410 µm long, 65-160 µm wide at widest part of ampulla. Up to three Details of holotype. The largest specimen of the lot. Length 7.8 mm (amputated specimen), first 15 segments 4.7 mm long, width at clitellum 0.85 mm. 27 segments. Dorsal bundles with 5-8 chaetae anterior to clitellum, 4-7 chaetae in postclitellar segments. Ventral bundles with 5-10 chaetae anterior to clitellum, 5-9 chaetae posteriorly. Longest measured chaetae 75 µm long, about 5 µm wide. Clitellum extending over XII-1/2XIII.
Remarks. The measured lengths of the sperm funnels are probably underestimated due to the difficulty of tracing them through the worms and due to folding. The two Swedish specimens were somewhat smaller than the Norwegian ones, and their funnels folded and only measurable for about 360 µm, but the length:width ratio was close to 4-6:1, as noted for the Norwegian specimens.
The description of Pachydrilus lineatus by Backlund (1947), from a drainpipe in Southern Sweden, in some ways reminds of our new species. Backlund was uncertain if his species belonged to P. lineatus because of the very wide spermathecal duct (which seemed as wide at the pore as in its medial part), the lack of a distinct ampulla and the possession of a large gland around the ectal pore. Furthermore, he described the penial bulbs as bilobed with a larger dorsal and a smaller ventral lobe. The description of the spermathecae sounds like the one of those of L. latithecatus, and after having examined the penial bulbs in the whole-mounted specimens of the latter species, it seems as if there could be a small lobe hidden behind the large spherical lobe (when viewed laterally). However, we would need transverse sections to truly compare this character to that which Backlund described. Lastly, Backlund reported small pharyngeal glands without dorsal development, which does not match with what we have observed for our species. Therefore, we are not certain as to the identity of Backlund's P. lineatus.
Spermathecae (Figs 3C-D, 6C) in V, spindle-shaped, without distinct ampulla. Ectal duct short, widening into ampulla. Ampulla with constriction midway to two thirds of the length, dividing it into two sections, the inner one of which connecting with oesophagus. Sperm filling lumen of ectal duct, heads of spermatozoa embedded in inner part of ampulla, sometimes also in outer part, forming aggregates. Spermathecae 180-300 µm long, 65-110 µm wide at widest part of ampulla. Gland cells surrounding ectal pore, forming compact mass, whole glandular body 95-180 µm in diameter at its widest part. One or two midventral subneural glands in XIII-XIV, 90-125 µm and 95-125 µm long, respectively; gland in XIV not observed in all specimens. Geographical distribution. Originally described from the United Kingdom, now genetically identified from Norway and Sweden. The full extent of this species' distribution is difficult to ascertain since it was previously synonymized with L. lineatus, a species distributed worldwide. BIN-number BOLD:ACV7714.
Remarks. Lumbricillus verrucosus was originally described by Claparède (1861) and later synonymized with L. lineatus by Nielsen and Christensen (1959), probably following Černosvitov (1937) who had downgraded the former to a form of the latter. Specimens from the two species examined in this study are indeed very similar when considering the shape of the spermathecae (Fig. 3) and most body measurements, but they differ in body colour, chaetal number and the proportions of the sperm funnels: indeed Lumbricillus verrucosus was described as being pale yellow, having 3-5 chaetae and sperm funnels about three times longer than wide (Claparède 1861), against L. lineatus being orange-red, having more chaetae and sperm funnels about five times longer than wide. Furthermore, the spermathecae of L. verrucosus seem to have an ampulla that is slightly longer and wider in the part ectal to the constriction, but the importance of this character remains to be proved. Even though the two species may be very difficult to separate morphologically, except perhaps by body colour, they are supported as separate species molecularly and avesiculate specimens have been found by us only in L. lineatus, not in L. verrucosus.
Interestingly, Lumbricillus verrucosus is genetically most closely related to L. rivalis (Levinsen, 1883) and not to L. lineatus (Fig. 1). However, these three species are well supported as closely related to each other. ( (CE22602), three mature specimens from Norway. For information on specimen collection localities and Gen-Bank accession numbers see Appendix 1.
Geographical distribution. Genetically identified from Greenland, Norway and Sweden; also recognized from Canada (BOLD:AAF9076). This species was originally described from Denmark and is considered well distributed throughout Europe, and North America.
Remarks. Pachydrilus rivalis was originally defined by Levinsen (1884) as a species with up to 9 chaetae per bundle, spermatheca formed by a large, red, pear-shaped container (Danish "beholder"), ending ectally with a glandular rosette, but bearing no glands on duct. The nephridial efferent duct originated at the posterior end of postseptale; the postseptals had red spots; the body color was red. Michaelsen (1889Michaelsen ( , 1900 placed the species in synonymy with L. lineatus along with his own Pachydrilus germanicus Michaelsen, 1886, but Ditlevsen (1904 reinvestigated Levinsen's type locality (Langelinie, a pier in the port of Copenhagen) and found that only one species conformed to Levinsen's short primary diagnosis, a species that was different from L. lineatus. He thus expanded the description of rivalis by adding the following traits: body length 15-20 mm, copulatory glands in XIII-XV, dorsal vessel from XIV, chaetal number 6-9 dorsally, 8-11 ventrally. Nevertheless, Welch (1917) and Černosvitov (1937) adhered to Michaelsen's opinion and the species had to await Nielsen and Christensen (1959) to be revalidated.
Our new specimens fit well with the more detailed re-descriptions, particularly in the number of chaetae and the shape of the spermathecae. The only differences concern the body size, where our worms are much smaller than those reported before, and the length/width ratio of the sperm funnels, which Nielsen and Christensen described as up to 10 times longer than wide, compared to our observed 4.5 times. These differences could be explained by our examination of fixed instead of live material.
Lumbricillus rivalis is genetically most closely related to L. verrucosus ( Fig. 1), but it is morphologically more alike L. rutilus (see remarks for that species) and L. enteromorphae, although it can be distinguished from the latter by lacking an atrium-like part where the vas deferens meets the penial bulb (see remarks for L. rubidus). Fig. 8 Lumbricillus sp. G; Klinth et al. 2017.

Lumbricillus sp. G
Material examined. Specimen SMNH 152834 (CE2246) & SMNH 152835 (CE2661), one mature and one immature specimen from the United Kingdom, and ZMBN 107942 (CE23373), one immature specimen from Norway. For information on specimen collection localities and GenBank accession numbers see Appendix 1.
Geographical distribution. Only known from Norway and the United Kingdom.
Remarks. Unfortunately only one mature specimen was available for this study making it difficult to assign it to a known species. On the other hand, the description of a new species on such limited material would be premature. Lumbricillus sp. G is genetically placed within the lineatus group (Fig.1). Fig. 9 Lumbricillus kaloensis Nielsen & Christensen, 1959: p. 100, figs 113-114;Erséus et al. 2010;Klinth et al. 2017.

Lumbricillus kaloensis Nielsen & Christensen, 1959
Type material. Typus amissus (Nomenclatura Oligochaetologica). Type locality: Kalø Vig, Denmark. We did not designate a neotype as we do not have material from the type locality. Material examined. SMNH 152733 (CE978), one mature specimen from Sweden & ZMBN 107842 (CE5412), one immature specimen from Norway. For information on specimen collection localities and GenBank accession numbers see Appendix 1.
Male genitalia paired (Fig. 9D). Testes originating in XI, extending forwards into IX, in one specimen also extending backwards into XII, with testis sacs forming regular club-shaped lobes. Sperm funnels in XI, 185 µm long, 155 µm wide, making them slightly longer than wide, funnels tapering towards vasa deferentia. Most of vasa irregularly coiled in XII, 15 µm wide. Penial bulbs round, 80 µm in diameter. Ovaries in XII. No mature eggs observed. Spermathecae (Fig. 9C) in V, spindle-shaped, without distinct ampulla, gradually widening, entally connecting with oesophagus. Sperm evenly embedded in ampulla. Spermathecae 290 µm long, 65 µm wide at widest part of ampulla. Gland cells surrounding ectal pore divided into at least three somewhat separated lobes, one lobe significantly larger than the others, whole glandular body 120 µm in diameter at its widest part. Two midventral subneural glands in XIII-XIV, 95 µm and 75 µm long, respectively.
Geographical distribution including BOLD data. Originally described from Denmark, now genetically identified from Norway and Sweden. BIN-number: BOLD:AAU0152.
Remarks. The specimens examined in this study match the original description by Nielsen and Christensen (1959) in the majority of the characters. Our worms were smaller than theirs, with slightly fewer chaetae (particularly in postclitellar bundles) and sperm funnels shorter in relation to their width, the latter of which could be explained by the difference between live and mounted material. We also observed two subneural glands which were not originally reported by Nielsen and Christensen. The spermathecae, despite being slightly damaged in our mature specimen, exhibit the typical large asymmetrical ectal gland subdivided into flap-like lobes, one of which is clearly larger than the others. Asymmetrical ectal glands have also been observed in L. enteromorphae von Bülow, 1957, and to some extent in L. rubidus Finogenova & Streltsov, 1978. The former can be distinguished from L. kaloensis by having sperm funnels 8 times longer than wide, and an atrium-like part where the vasa deferentia meet the penial bulbs. Lumbricillus rubidus, while having ectal glands on the spermathecae that may appear asymmetrical, does not have the glands subdivided into such clearly asymmetrical flaps as in L. kaloensis. Furthermore, L. rubidus has a more distinct musculature covering the ectal duct of the spermathecae, sometimes appearing as a muscular bulb; L. kaloensis also has a muscular coating of the duct, but it is thinner.
Lumbricillus kaloensis is genetically placed within the lineatus group. Its possible sister-group relationship with Lumbricillus sp. F (suggested by the tree in Fig. 1) is not statistically supported by the DNA data. Material examined. SMNH 152832 (CE2659), one mature specimen from the United Kingdom. For information on specimen collection locality and GenBank accession number see Appendix 1.
Male genitalia paired (Fig. 10D). Testes originating in XI, extending forwards into X, with testis sacs forming regular club-shaped lobes. Sperm funnels in XI, 175 µm long, 90 µm wide, making them about twice as long as wide. Vasa deferentia not observed. Penial bulbs round, 85 µm in diameter. Ovaries in XII. About two or three mature eggs present. Spermathecae (Fig. 10C) in V, spindle-shaped, without distinct ampulla. Ectal duct short, widening into ampulla. Ampulla with constriction midway, dividing it into two sections, the inner of which connecting with oesophagus. Sperm concentrated in inner part of ampulla, embedded in wall of ampulla. Spermathecae 210 µm long, 80 µm wide at widest part of ampulla. Gland cells surrounding ectal pore, forming compact mass, 110 µm in diameter at its widest part. Possibly one subneural gland in XIV, 50 µm long.
Geographical distribution. Only known from the United Kingdom and Norway (unpublished results).
Remarks. Unfortunately only one mature specimen was available for this study, making it difficult to assign it to a known species. On the other hand, the description of a new species on such limited material is not recommended. Lumbricillus sp. F is genetically placed within the lineatus group; it is possibly phylogenetically close to L. kaloensis, but this is not strongly supported (Fig.1). Stephenson, 1932 Fig. 11 Lumbricillus pumilio Stephenson, 1932a (CE3437), seven mature specimens from the United Kingdom. For information on specimen collection localities and GenBank accession numbers see Appendix 1.
Coelomocytes difficult to identify as such in this species, but small round, oval or spindle-shaped granulated cells about 5-7 µm long were noted in the coelomic cavity. Paired pharyngeal glands present in IV, V and VI; each pair converging dorsally (Fig.  11B). Dorsal vessel originating in XII or XIII. Nephridia observed in IX and XV-XVIII, about 45-50 µm long. Anteseptale small, consisting of funnel only. Postseptale oval, tapering posteriorly into efferent duct. Brain longer than wide, with a marked posterior incision creating two horn-like structures.
Geographical distribution. Genetically identified from the United Kingdom. Also reported and briefly described from Iceland (Erséus 1976).
Remarks. Our measurements of the coelomic corpuscles (5-7 µm long) contradict the original description where they are described as being 20-28 µm. This could either be due to a high degree of variation in this trait or that we are comparing non-homologous cell types. The smaller subneural gland in XV described by Stephenson could not be distinguished, either because of its small size or because it was absent. Despite a few discrepancies from the original description the small body size of the worms together with the shape of the spermathecae and other reproductive organs supports that the sampled specimens belong to L. pumilio.
Spermathecae (Fig. 12C, D) in V, spindle-shaped, without distinct ampulla. Ectal duct short, about 1/5 of total length of spermatheca, rapidly widening into ampulla. Conspicuous muscle cells encircling duct and connecting it to epidermis. Ampulla making sharp bend inwards and entally connecting with oesophagus. Sperm tightly packed in ectal duct, possibly covered by thin layer of secretion, spermatozoan tails occupying outer part of ampulla, heads aggregating into distinct clusters in inner part, and partly embedded in wall, of ampulla. Spermathecae 120-275 µm long, 60-110 µm wide at widest part of ampulla. Gland cells surrounding ectal pore, forming com- pact body with few marginal lobes, glandular body 60-135 µm in diameter at its widest part. One midventral subneural gland in XIV, 60-150 µm long.
Geographical distribution. Described from Russia and Germany, now genetically identified from Sweden.
Remarks. The specimens in this study match the original description of L. rubidus by Finogenova and Streltsov (1978) in most characters such as length, number of chaetae and sperm funnel ratio. It seems that our specimens in general possessed larger internal organs, such as sperm funnels, penial bulbs and spermathecae. Nevertheless, the strong musculature around the ectal pore of the spermathecae, originally described as a muscular bulb, was found with clear resemblance in our specimens. This muscular sleeve covering the ectal duct of the spermathecae is conspicuous in all specimens. Several separate muscle bundles radiate around the base of the duct connecting to the body wall and may have a function in widening the pore in conjunction with copulation or fertilization of eggs. In one specimen, where one spermatheca was seen from above, the layers of musculature created a circle seemingly dividing the ectal gland in two. A closer examination revealed that the musculature more probably is tightly encircling the gland cells of the ectal gland without dividing them. The muscular bulb, as originally termed by Finogenova and Streltsov, is probably the same ectal part of the ectal gland, separated by the encircling musculature, rather than a compact mass of muscles. Similar muscle structures encircling the ectal pore of the spermathecae have been seen in most species of Lumbricillus during this study but they never appeared so conspicuous as in L. rubidus.
In 1959, Nielsen and Christensen classified the species L. enteromorphae von Bülow, 1957 as a hunger form of L. rivalis. Kossmagk-Stephan (1985) rejected this idea and instead synonymized L. rubidus with von Bülow's species based on the fact that they both have muscular bulbs. We agree that L. enteromorphae should be considered a separate species from L. rivalis, because of the asymmetrical ectal glands of the spermathecae and the atrium-like part where the vasa deferentia meets the penial bulbs, observed only in the former species. However, we do not agree that L. rubidus is a synonym of L. enteromorphae, as the former has sperm funnels that are much shorter in relation to their width (2-4 compared to 8 times longer than wide) and lacks any atrium-like part of the vasa deferentia. It is true that von Bülow described a funnel-like thickening of the spermathecal ectal duct which could be a structure similar to the muscular bulb seen in L. rubidus. However, we have observed varied extents of muscular coverings of the ectal ducts in most species belonging to the lineatus group, and a true comparison of this character can only be made once we have specimens of von Bülow's species.
Male genitalia paired (Fig. 13D). Testes originating in XI, extending forwards into X, sometimes IX, with testis sacs forming regular club-shaped lobes. Sperm funnels in XI, 100-125 µm long, 95-110 µm wide making them about 1-1.5 times longer than wide. Funnels lobed rather than cylindrical, and abruptly tapering towards vasa deferentia. Vasa with few irregular coils around ovaries in XII, and about 10 µm wide. Penial bulbs round/pear-shaped 45-50 µm in diameter. Three to four mature eggs present at a time.
Remarks. The original description of L. fennicus matches the specimens of this study in most characters, but there are a few differences. Our specimens measured 2-3.5 mm in length after fixation, but considering that some had been cut directly posterior to the clitellum the length of the complete worms probably was 3-5 mm. This is smaller than the 8 mm reported by Nurminen (1964), but he based his description on living worms. Nurminen described the clitellum as covering 1/2XI-XII while in our specimens, the clitellum extends over XII-1/2XIII. The extension of the clitellum may vary as it develops, but the whole structures does not generally shift in position, and this suggests the possibility of a printing or observation error in the original description.
The lobed, as opposed to cylindrical, sperm funnels are so far (in European species) only reported for L. fennicus, and this, together with the matching shape of the spermathecae, allowed confident allocation of the specimens to this species despite some incongruence among the characters mentioned above. The interpretation of the lobes of the sperm funnels probably also differs between living and fixed specimens. Our Swedish specimens were collected in freshwater habitats, but the sites are possibly subjected to brackish water at times, making the range of salinity similar to the original records from the Gulf of Finland. Most other records in Europe are from coastal oligohaline or inland freshwater habitats.
Lumbricillus fennicus is both genetically (Fig. 1) and morphologically placed within the L. lineatus group.

The pagenstecheri group
Characteristics: Testes with testis sacs regularly lobed in bunch-like arrangement. Spermathecae with distinct ampulla and glands both surrounding the ectal pore and distributed along the duct. Chaetae 3-6 per bundle, or more; upper bundles dorsolateral. Penial bulbs round. Sperm funnels about twice as long as wide.
Type material. Typus amissus (Nomenclatura Oligochaetologica). Type locality: The original material was collected in Rhine River near Karlsruhe, and in ponds around Heidelberg, Germany (Ratzel 1869), but none of these places has yet been specifically designated as the type locality. We did not designate a neotype as we do not have material from any of the original localities, nor do we know which one, if any, of our cryptic species that represent the true nominate species.
Remarks. The molecular studies by Klinth et al. (2017) supported the delimitation of four different species with the morphology of L. pagenstecheri, here denoted as cryptic species A-D. Particularly the morphology of the spermathecae characterizes this group. There are two groups of gland cells, one creating the typical mass of glands surrounding the ectal pore, as seen in the other species of Lumbricillus, and the other group composed of numerous, rather long, gland cells covering the ectal duct. These two groups of gland cells can be difficult to distinguish from each other, depending on the orientation of the mounted specimens, but they create the impression of a very narrow duct followed by a distinct, almost spherical, thin-walled ampulla. While there seems to be some morphological differences between the four species in this study, such as size and number of chaetae, there are too few sampled specimens to verify that these characters do not overlap.
Lumbricillus pagenstecheri was originally described by Ratzel (1869) from the Rhine River in Germany and has later been re-described by Nielsen and Christensen (1959) as well as others and today includes five synonymized species (listed above). Such synonymies may need reappraisal as there are some differences between the original descriptions concerning size, number of segments and number of chaetae and there is a possibility that some of the synonymized species are present in our material. Moreover, about thirteen described species from the Northwestern Pacific and eight from the Northeastern Pacific have a morphology similar to that of L. pagenstecheri (Timm 2005), and a more extensive phylogenetic study focused on this part of the genus will be necessary, to resolve the taxonomy of this complex group.
For this study, we chose to present the morphological measurements only for our cryptic species A, which is the only one with a sufficient sample size, and provide a comparison of some characters with the other three cryptic species in Table 2. In general, species B and D where the largest, species D possessed fewer chaetae per bundle than the others, and for species C we unfortunately had no fully mature specimens. Full information on collection localities and accession numbers of all four species are given in Appendix 1. Fig. 14, Table 2 Material examined. SMNH 152766 (CE1896), SMNH 152767 (CE1897), SMNH 152768 (CE1899), SMNH 152769 (CE2497), SMNH 152770 (CE2498) & SMNH 152771 (CE2500), six mature specimens from Sweden. For information on specimen collection localities and GenBank accession numbers see Appendix 1.
Male genitalia paired (Fig. 14D). Testes originating in XI, extending forwards into X, sometimes IX, with testis sacs forming regular club-shaped lobes. Sperm funnels in XI, 210-300 µm long, 145-225 µm wide, making them about as long as wide or twice as long as wide, funnels gradually tapering towards vasa deferentia. Most of vasa irregularly coiled in XII, 10-15 µm wide. Penial bulbs round, 135-185 µm in diameter. Ovaries in XII. About two to eight mature eggs present at a time.
Spermathecae (Fig. 14C) in V, club-shaped, with distinct ampulla. Ectal duct narrow, about as long as ampulla, abruptly widening into ampulla. Ampulla round, entally connecting with oesophagus. Sperm arranged in circular masses in ampulla. Spermathecae 140-215 µm long, 75-110 µm wide at widest part of ampulla. Two groups of gland cells, one covering ectal duct, the other surrounding ectal pore. Gland cells surrounding ectal pore forming compact mass, slightly lobed, whole glandular  body 80-195 µm in diameter at its widest part. Two midventral subneural glands in XIII-XIV, 130 µm and 95 µm long, respectively. Geographical distribution including BOLD data. Genetically identified from Sweden; also recognized from Canada and Denmark (BIN-number: BOLD:AAF9627). Stephenson, 1911 Fig. 15 Lumbricillus viridis Stephenson, 1911 (CE23255), three mature and one half-mature specimens from Norway. For information on specimen collection localities and GenBank accession numbers see Appendix 1.
Spermathecae (Fig. 15C) in V, club-shaped, with distinct ampulla. Ectal duct narrow, shorter than ampulla, abruptly widening into ampulla. Ampulla round. Sperm arranged in a compact central sphere in the ampulla as well as embedded in the wall of ampulla, creating a circle around the sphere. Spermathecae 265-320 µm long, 270-310 µm wide at widest part of ampulla. Gland cells surrounding ectal pore, forming compact mass, slightly lobed, whole glandular body 310-325 µm in diameter at its widest part. Gland cells also along the ectal duct. Up to four midventral subneural glands in XIV-XVII, 240-270 µm, 215-245 µm, 190-215 µm and 130 µm long, respectively; glands in XVII not observed in all specimens.
Geographical distribution including BOLD data. Genetically identified from France and Norway. Previously described from Denmark, Norway (Nurminen 1965b) the United Kingdom and Sweden (Erséus 1977). BIN-number: BOLD:AAU1636.
Remarks. Specimens in this study are smaller and possess somewhat fewer chaetae, than the ones from the original description by Stephenson and the later re-description by Nielsen and Christensen. Furthermore, the observed proportions of the sperm funnels (twice longer than wide) differ greatly from those (7-10:1, or 6-8:1) described by Stephenson and Nielsen and Christensen, respectively. However, folding of these organs may have caused us to underestimate their true length. Nevertheless, the distinct greenish colour of the sampled specimens and the resemblance between their spermathecae and particularly the one described by Nielsen and Christensen confirm these specimens as Lumbricillus viridis.
According to our knowledge, the presence of gland cells along the spermathecal ectal duct has not been reported for L. viridis before, possibly because of the difficulty of distinguishing these gland cells from the large ones surrounding the ectal pore. In this study, similar duct glands have only been observed in L. pagenstecheri sensu lato.
Lumbricillus viridis is genetically most closely related to the L. pagenstecheri species complex ( Fig.1: L. pagenstecheri A-D).

The "tuba" group
Characteristics: Testes with testis sacs regularly lobed in bunch-shaped arrangement. Spermathecae with ampulla distinctly set off from the duct and glands surrounding the ectal pore. Chaetae usually 3-6 per bundle; upper bundles dorsolateral. Penial bulbs round. Sperm funnels about as long as wide.
Note that this group, containing also L. scandicus sp. n., is not monophyletic (Fig. 1); it is based on morphological similarity only. Type material. Typus amissus (Nomenclatura Oligochaetologica). Type locality: Firth of Clyde, Millport, Island of Cumbrae, United Kingdom (Stephenson 1911). We did not designate a neotype as we do not have material from the type locality. Material examined. ZMBN 107916 (CE22614), one mature specimen from Norway. For information on specimen collection locality and GenBank accession number see Appendix 1. Note that this specimen is the only sexually mature available for this study, but additional (immature) worms were studied genetically by Klinth et al. (2017).
Geographical distribution including BOLD data. Genetically identified from Norway and Sweden. Also described from Denmark and the United Kingdom. BINnumber: BOLD:ACQ1913. Remarks. Our specimen matches the original description by Stephenson (1911) well in most characters, including the shape of the spermathecae and the proportions of the sperm funnels. However, its body size was smaller and the chaetae per bundle slightly fewer than in the original description.
Lumbricillus tuba is genetically most closely related to the L. pagenstecheri group (including L. viridis), but it is morphologically most similar to L. scandicus sp. n. described below.
Coelomocytes numerous, 15-20 µm long, round or oval. Paired pharyngeal glands present in IV, V and VI; each pair converging dorsally, with large ventral lobes (Fig.  17B). Posteriormost pair sometimes extending into VII. Dorsal vessel originating in either XII or XIII, difficult to distinguish due to presence of mature eggs. One nephridium observed in XIV about 85 µm long, anteseptale consisting of funnel only, duct originating posteroventrally. Brain widening posteriorly, possibly with posterior incision.
Geographical distribution. Genetically identified from Norway and Sweden. Also reported from Denmark and Russia (White Sea).

Remarks.
The new species corresponds well to the description of Lumbricillus cf. helgolandicus (Michaelsen, 1927) by Nielsen and Christensen (1959), which is why this name was used in the molecular study by Klinth et al. (2017). However, Nielsen and Christensen noted several differences in the morphology of their specimens in comparison to the extended description of L. helgolandicus given later by Michaelsen (1934), the most important being the morphology of the spermathecae and the sperm funnels. The spermathecal ampulla was interpreted by Michaelsen as being filled with an irregular mass of spermatozoa. Von Bülow (1957) instead redrew the spermathecae as having a distinct circle of spermatozoa which also corresponds to the interpretation by Nielsen and Christensen and what we observed in this study. Furthermore, Michaelsen originally described the sperm funnels as 12 times longer than wide. Nielsen and Christensen, on the other hand, found the funnels to be only 2-3 times longer than wide, which corresponds better to the ratio measured in our material. von Bülow, 1957, unfortunately did not comment on the length/width ratio of the funnels.
These circumstances prompted re-examination of the last remaining syntype of Pachydrilus helgolandicus from Michaelsen's collection in the Zoological Museum in Hamburg (see description of that material below). We found that the sperm funnels were more than 4 times longer than wide, compared to his reported 12 times. This difference could be explained by Michaelsen having examined live material, whereas the syntype that we studied had been fixed (contracted) in formalin or alcohol, shortening the sperm funnels. Furthermore, we might have underestimated the true length of the sperm funnels due to the difficulties with measuring folded organs in mounted material. Regardless, compared to our material of "L. cf. helgolandicus", here described as L. scandicus sp. n., the sperm funnels of L. helgolandicus sensu stricto clearly have a higher length/width ratio.
The spermathecae of L. helgolandicus are similar to those of L. scandicus in having a distinct ampulla and a very large ectal gland. However, in L. helgolandicus, the spermatheca contains sperm that are arranged in an irregular mass, and it has a very distinct musculature covering the ectal duct (possibly made more apparent by the aging of the material), whereas the spermatheca of L. scandicus has sperm arranged in a more circular manner and only weakly defined musculature covering the ectal duct.
Lumbricillus helgolandicus is larger than L. scandicus and has generally larger internal organs. It also has fewer chaetae per bundle, no more than 5 in preclitellar, and 2-3 in postclitellar bundles, whereas L. scandicus has up to 7 chaetae in preclitellar, and up to 6 in postclitellar bundles. Nielsen and Christensen (1959) examined specimens that seem to have been larger than ours and closer to L. helgolandicus in size. However, like our material, they had sperm funnels that were not much longer than wide, more chaetae per bundle and spermathecae with sperm arranged in a circular manner. Similarly, the material that von Bülow (1957) referred to as L. helgolandicus also had more chaetae per bundle than Michaelsen's worm and had spermathecae with sperm arranged in a circular manner.
Based on our assessment of the syntype from Helgoland, we conclude that our Scandinavian material is not conspecific with L. helgolandicus (Michaelsen, 1927), and instead deserves to be treated as a new species (L. scandicus). Furthermore, we conclude that L. helgolandicus sensu von Bülow 1957 and L. cf. helgolandicus sensu Nielsen and Christensen 1959 are identical to L. scandicus.

Remarks.
We have mounted and re-examined the only remaining syntype of L. helgolandicus from the Zoological Museum in Hamburg and found the worm to correspond well to the descriptions of L. helgolandicus by Michaelsen (1927Michaelsen ( , 1934 and no doubt to represent the attached species name. However, we found some discrepancies with the published measurements of the sperm funnels (also discussed above). As the slide of the mounted specimen is of good quality and since this is the only remaining syntype of the species we designate ZMH V5786 as the lectotype of L. helgolandicus.
Based on morphology, L. helgolandicus is similar to L. scandicus and is probably closely related to this species. In the DNA-based phylogeny, L. scandicus is placed close to the lineatus group (Fig. 1), and it is likely that L. helgolandicus phylogenetically belongs there too. However, for convenience of morphological identification, both taxa are referred to the paraphyletic "tuba" group.

Lumbricillus buelowi Nielsen & Christensen, 1959
Type material. Typus amissus (Nomenclatura Oligochaetologica). Type locality not precisely defined; the species was originally described from four different sites (Kalø, Femmøller, Ebeltoft and Avedøre) in Denmark (Nielsen and Christensen 1959). We did not designate a neotype as we do not have material from any of the type localities. Geographical distribution including BOLD data. Genetically identified from Norway and Sweden. Also known from Denmark and Germany. BIN-number: BOLD:ACQ3084.
Geographical distribution including BOLD data. Genetically identified from Norway and Sweden. Also known from Denmark and Germany. BIN-number: BOLD:ACQ3084.
Remarks. It is clear that the two species here identified as L. buelowi and L. knoellneri Nielsen & Christensen, 1959 are closely related (Klinth et al. 2017) (see also Fig. 1), and most morphological characters such as the spermathecae, sperm funnels and penial bulbs are virtually identical between them. However, there are some general differences in our studied material: L. buelowi is on average larger than L. knoellneri, as originally noted by Nielsen and Christensen, although there is overlap between the two. Lumbricillus buelowi possesses 2-3 chaetae in the lateral bundles anterior of the clitellum while L. knoellneri possesses only 2. In their original description, Nielsen and Christensen also differentiate the two species by colour and size of coelomocytes where L. buelowi is red with two types of coelomocytes, one being larger than the chaetae, compared with L. knoellneri which is white and have only one type of coelomocytes, shorter than the chaetae. We observed only the smaller coelomocytes in both species, and as we stained the material we only have observations of the live animals; some specimens of L. buelowi were noted as being pinkish.
There are a number of species with descriptions similar to the ones of L. buelowi, and therefore also of L. knoellneri, such as L. eltoni (Stephenson, 1924), L. muscicolus (Stephenson, 1924) and L. nielseni Nurminen, 1965. All these latter three were described from Svalbard where we found specimens of L. knoellneri but not L. buelowi. Unfortunately, the two species described by Stephenson (1924) were not illustrated and the descriptions are not extensive enough for us to synonymize either of them with L. knoellneri (Stephenson's species would in that case hold seniority in name). Also L. nielseni was described too briefly (Nurminen 1965a) and the illustrated spermathecae seems slightly different from those of L. buelowi and L. knoellneri, causing us once again to avoid synonymization.
Male genitalia paired (Fig. 20D). Testes originating in XI, extending forwards into X, with testis sacs covering compact mass, slightly lobed but not regularly arranged. Sperm funnels in XI, 100-150 µm long, 70-155 µm wide, making them about as long as wide or 1.5 times longer than wide, funnels tapering towards vasa deferentia. Most of vasa irregularly coiled in XII, 5-10 µm wide. Penial bulbs round, 70-115 µm in diameter. Ovaries in XII. One to two mature eggs present at a time.
Geographical distribution including BOLD data. Genetically identified from Norway (mainland and Svalbard) and Sweden. Also known from Denmark and Germany. BIN-number: BOLD:ACM5261.
Remarks. Lumbricillus knoellneri is described as having only 2 chaetae throughout the body but the newly studied material suggests that the preclitellar ventral bundles possess 2-3 chaetae. In fact, all of the eight studied specimens had 3 chaetae in at least 2 of the preclitellar ventral bundles, some in as many as 9. We found no pattern of this distribution and for each preclitellar ventral segment bearing chaetae (II-XI), we found representatives with either 2 or 3 chaetae. This shows how variable this trait is and could explain the difference to the description by Nielsen and Christensen. However, it could also mean that our "L. knoellneri" is in fact another species. Many of the internal organs of L. knoellneri were as long as or even slightly longer than the ones in L. buelowi. This in combination with a generally smaller size caused the segments of L. knoellneri to appear more contracted. For a further discussion see the Remarks for L. buelowi above.
In 1985, Kossmagk-Stephan synonymized L. cervisiae (which he himself had described as a new species two years earlier) and L. christenseni Tynen, 1966 with L. knoellneri. All three species are small, have only two chaetae per bundle (at least according to the original descriptions) and similarly shaped spermathecae. However, L. christenseni has a sperm funnel that is 7-8 times longer than wide which is significantly longer than the 1.5 times measured in L. knoellneri. The sperm funnel of L. cervisiae is 3-4 times longer than wide which is also more than that of L. knoellneri. Furthermore, the testis sacs of L. cervisiae cover several small scattered lobes and the vasa deferentia extends backwards into XIII. Finally, L. cervisiae appears to be more slender than L. knoellneri and has significantly smaller internal organs, which we were able to discern by examining the mounted original material of Kossmagk-Stephan. Therefore, we reject the idea of L. cervisiae and L. christenseni being synonyms of L. knoellneri and treat them as separate species.
Male genitalia paired (Fig. 21D). Testes originating in XI, extending forwards into X, with testis sacs covering mass of rather large irregularly arranged lobes. Sperm funnels in XI, in some specimens extending back into XII, 375-975 µm long, 55-103 µm wide, making them 6-13 times longer than wide, funnels tapering towards vasa deferentia. Most of vasa irregularly coiled in XII, in one specimen extending back into XIV, 5-10 µm wide. Penial bulbs round, 110-140 µm in diameter. Ovaries in XII. One to six mature eggs present at a time.
Geographical distribution including BOLD data. Genetically identified from Norway (mainland and Svalbard) and Sweden. Also reported from Denmark, Canada, Germany, Greenland, Iceland, Ireland, Wales and North-Western Australia . BIN-number: BOLD:AAT8953.
Remarks. The original description by Michaelsen (1889) was later amended by Knöllner (1935) who redrew the shape of the nephridia and spermathecae, also confirmed by Nielsen and Christensen (1959). The newly examined material in this study resembles the original description in most characters but the spermathecae and nephridia are in agreement with the amended descriptions. Coelomic corpuscles were found with hooked ends which seemed to bind to the internal tissue in a way that is described by Michaelsen. The testes seemed to be either an irregular compact mass or divided into separate lobes, encased in testis sacs, but these lobes were not arranged in the bunch-shape seen in the lineatus, pagenstecheri and "tuba" groups.
Geographical distribution. Genetically identified from Norway. Remarks. Initial comparisons found similarities between this species and Lumbricillus westheidei Kossmagk-Stephan, 1983, such as similar shape of spermathecae and slightly bilobed penial bulbs. However, having re-examined Kossmagk-Stephan's type material we found some important differences compared to our specimens. First, L. westheidei has only two chaetae per bundle, whereas our specimens have up to three chaetae in the preclitellar segments (the position of the upper bundles is identical in the two species). Second, the three pairs of pharyngeal glands are clearly separated in L. westheidei but in our specimens at least the first two pairs appear to have a dorsal connection. Third, the testis sacs of L. westheidei are much smaller than the ones we observed in our specimens. Fourth, the vasa deferentia appear to be much longer and form many more coils in segment XII in L. westheidei compared to our L. sp. H. Finally, the sperm funnels of L. westheidei are about 10 times longer than wide, against the 4:1 length:width ratio observed in our specimens. Unfortunately, none of our examined specimens appeared to be fully mature, as sperm were not observed either at the sperm funnels or in the spermathecae, and there were no mature eggs present. This suggests that the sperm funnels and spermathecae were not fully developed and could at maturity resemble those of L. westheidei more. Due to this uncertainty we cannot completely rule out that our specimens are of the same species as L. westheidei; however, for now we will continue to treat it as an unknown species, simply referred to as L. sp. H.
Since L. westheidei resembles, in its general morphology, our L. sp. H, it is important to add some notes on its generic allocation. Kossmagk-Stephan (1983) questioned the placement in Lumbricillus due to the undivided testis sacs. He had also observed this feature in some other "Lumbricillus" species, such as L. arenarius and L. semifuscus, the latter here below transferred to Claparedrilus gen. n. Furthermore, he noted a similarity in the morphology of the spermathecae between L. westheidei, L. buelowi, L. knoellneri and L. codensis Lasserre, 1971. In 1985, Coates and Erséus established the new genus Randidrilus and designated Lumbricillus codensis as its type species. Because of the resemblance to the latter in the bilobed penial bulb, the long spermathecal ectal duct, the long sperm funnels and the undivided testis sac, Kossmagk-Stephan (1985) proposed in his doctoral thesis the new combination, R. westheidei, and since then the species was regarded as another member of Randidrilus (Coates, 1989;Mackei and Erséus 1997;Schmelz and Collado 2012). However, we confirm here that, unlike other species of Randidrilus, L. westheidei has more than a single chaeta per bundle, does not lack chaetae in numerous lateral and ventral bundles and does not have an unpaired sperm sac extending backwards into postclitellar segments. Instead, L. westheidei resembles members of the arenarius group within Lumbricillus by having few chaetae, long sperm funnels, slightly bilobed penial bulbs and paired testis sacs that are not regularly lobed. Therefore, we transfer this species back from Randidrilus into Lumbricillus, making it L. westheidei once again.
Spermathecae (Fig. 23C) in V, pouch-shaped, without distinct ampulla, gradually widening, entally connecting with oesophagus. Sperm completely occupying lumen of duct and ampulla, regularly arranged with spermatozoan heads facing the wall and tails along the duct, forming denser aggregation throughout the centre of the spermathecae. Spermathecae 95-205 µm long, 40-100 µm wide at widest part of ampulla. Gland cells surrounding ectal pore, divided into few flaps, whole glandular body 70-120 µm in diameter at its widest part. Up to two midventral subneural glands in XIV-XV, 60-80 µm, 60-65 µm long, respectively; glands in XV not observed in all specimens.
Geographical distribution including BOLD data. Genetically identified from Norway, Russia (White Sea), Sweden and the United Kingdom (Klinth et al. 2017). BIN-number: BOLD:AAU0151.
Remarks. The specimens examined match the description of Lumbricillus dubius by Stephenson (1911) well, but they are smaller in body size and have testes that seem to form an irregular mass rather than being made up of branches as observed by Stephenson. The clearly divided penial bulbs and the morphology of the spermathecae bear a close resemblance between our specimens and the original description. Stephenson wrote that no sperm were observed in the spermathecae, but his illustrations depicting sections of the same clearly show the unique distribution of spermatozoa with heads regularly arranged perpendicular to the spermathecal wall. It is possible that he did not recognize them as sperm, simply because of this unusual arrangement.
Lumbricillus dubius has irregularly lobed testis sacs and spermathecae that are at least superficially similar to those of L. arenarius. The chaetae are straight to slightly sigmoid and few in number, which further supports the close relationship with L. arenarius and L. sp. H.
Claparedrilus gen. n. http://zoobank.org/E7A1215B-D41F-4721-89E9-226176994F0C Genus description/diagnosis. Prostomium hemispherical. Head pore at 0/1. Epidermis with transverse rows of gland cells. Chaetae straight to sigmoid, without nodulus, grouped into two dorsolateral and two ventrolateral bundles per segment. Oesophageal appendages absent. Pharyngeal glands in four pairs, in IV-VII, converging but not connected dorsally, some with ventral lobes, but without secondary glands. Only nucleated coelomocytes present. Dorsal vessel originating intra or in segments posterior to clitellum. Nephridia with anteseptale made up of funnel on a short stalk. Clitellum more or less covering segments XII-XIII. Testes paired, surrounded by testis sacs; the latter forming compact mass with shallow lobes irregularly arranged. Penial bulbs round and compact. Midventral subneural glands present in XIV-XV. Spermathecae in V, attached to and usually communicating with oesophagus lumen, and with crown of glands surrounding ectal part of ectal duct. Spermathecae club-shaped with ampulla distinctly set off from duct. Spermathecal diverticula absent. Marine, living in the littoral zone. Type species. Claparedrilus semifuscoides sp. n. Other species. Claparedrilus semifuscus (Claparède, 1961) comb. nov. Etymology. Clapare-from Claparède, the original author of the species C. semifuscus, a poorly defined species with which the type species for this new genus (C. semifuscoides) has been misidentified, and -drilus (latinized Greek) for worm.
Remarks. The need for this new genus arose from the difficulty of placing the type species C. semifuscoides (which we previously referred to as L. semifuscus) in the phylogeny of the Enchytraeidae. Molecular data had previously supported that this species was not a member of Lumbricillus and instead closer to, but not a member of, Globulidrilus and Bryodrilus (Klinth et al. 2017;Martinsson et al. 2017). Both these genera share some traits with Claparedrilus, such as the shape of the spermathecae and nephridial anteseptale with small part of the nephridial body, but both have only three pairs of pharyngeal glands and are aquatic or terrestrial. The phylogenetic studies lacked representatives from several potential candidate genera of marine enchytraeids. Therefore, we compared the morphology of our species with these candidates (presented in Table 3), after which we still found support for the recognition of the new genus Claparedrilus. In particular, the combination of four pairs of pharyngeal glands, nephridia with a stalked funnel in the anteseptale and the presence subneural glands distinguishes this new genus. Other material examined. ZMBN 107908 (CE23750) & ZMBN 107912 (CE24657), one mature and one half mature specimen from Norway. For information on specimen collection localities and GenBank accession numbers see Appendix 1.

Claparedrilus semifuscoides
Etymology. Named after its similarity to Claparedrilus semifuscus, which it has previously been confused with and misidentified as. Table 3. A comparison of characters distinguishing Claparedrilus gen. n. from other marine enchytraeid taxa. Traits of particular importance highlighted in boldface.
*Coded according to Rota et al (2008); see differing interpretation in Schmelz and Collado (2008). ** One species (Randidrilus quadrithecatus Coates & Erséus, 1985) with two pairs of spermathecae, and four pairs of pharyngeal glands distributed from IV-VII. Diagnosis. This species can be distinguished from C. semifuscus by the size of the penial bulbs. In C. semifuscus, the bulbs are much larger than the sperm funnels, whereas in C. semifuscoides they are of about the same size as the funnels or smaller.
Coelomocytes numerous, 10-20 µm long, spindle-shaped, oval, round, granulated with distinct nucleus. Paired pharyngeal glands 4 pairs, in IV, V, VI and VII, respectively; each pair converging but not connected dorsally (Figs 24B-C), pair in IV with dorsal lobes only, pair in V with both dorsal and ventral lobes, pairs in VI and VII large and compact, but dorsal lobes difficult to distinguish from potential ventral ones. Dorsal vessel originating in XIII. Nephridia (Figs 24F-G) observed in VI-VIII and XIV and onwards, 65-130 µm long, with various shapes, anteseptale with funnel on a thin stalk, postseptale oval, tapering into efferent duct which seems to originate either terminally or from posterior of the midventral of the postseptale (compare Figs 24F-G). Brain with posterior incision.
Male genitalia paired (Fig. 24E). Testes originating in XI, with testis sacs enclosing compact sperm mass with numerous inconspicuous, irregularly arranged lobes, extending forwards into X, in some specimens extending into IX and XII. Sperm funnels in XI, 125-160 µm long, 110-145 µm wide, making them 1-1.5 times longer than wide, funnels tapering towards vasa deferentia. Most of vasa irregularly coiled in XII, 10-15 µm wide. Penial bulbs round, 140-155 µm in diameter. Ovaries in XII. Two to five mature eggs present at a time.
Spermathecae (Fig. 24D) in V, club-shaped, with distinct ampulla. Ectal pore midlateral. Ectal duct seemingly divided into two zones by intermediary layer of musculature. The outer zone, or coelomic lining, covering muscular layer, containing large, clearly defined nuclei. The inner zone, which is the epithelium and cuticle, lining muscular layer, appearing to be made up by numerous fine lines (perpendicular to the duct axis); these lines possibly epithelial cells or their nuclei, or microvilli crossing the cuticle. Duct twice as long as ampulla, abruptly widening into ampulla. Ampulla round, entally connecting with oesophagus, and containing irregular mass of sperm in postcopulatory specimens. Spermathecae 240-270 µm long, 60-110 µm wide at widest part (the ampullae). Gland cells surrounding ectal duct near spermathecal pore, forming compact mass, 50-105 µm in diameter at its widest part. Two midventral subneural glands in XIV-XV, 40-100 µm, 70-85 µm long, respectively.
Geographical distribution. Genetically identified from Norway and the United Kingdom. Also known (by morphology) from Iceland (Erséus 1976) and Sweden (Erséus 1977).
Remarks. In 1861, Claparède described Pachydrilus semifuscus from the Hebrides in Scotland. Due to its unusual and confusing morphology, this species has been moved around among some enchytraeid genera. It was transferred to Marionina (Michaelsen 1900), then to Enchytraeoides (Michaelsen 1927), and finally to Lumbricillus (Nielsen and Christensen 1959). The original description focused almost entirely on the reproductive organs and noted sperm funnels about 1.5 times longer than wide, spermathecae with a long thin duct and clearly separated ampulla, nephridia with anteseptales made up of funnels only, and large kidney-shaped penial bulbs. Southern (1909), studying material from Dublin Bay (Ireland) and Edinburgh (Scotland), added that his specimens had 4-5 chaetae per bundle, a concave posterior of the brain, five pairs of pharyngeal glands in IV-VII (two of them in V), but with cylindrical rather than kidney shaped penial bulbs. Stephenson (1911) also recorded this species from Scotland but then increased the number of chaetae to 4 -8 per bundle. He agreed with Southern's description of the pharyngeal glands but questioned the shape of the penial bulbs, which he found to be spherical and not unusually large, compared to the descriptions by Claparède and Southern. In 1976, Erséus described the species from Iceland, also with pharyngeal glands in IV-VII, but with fewer chaetae per bundle and with anteseptales of the nephridia made up of a few coils as well as the funnels. Having studied our material, which is partly from Wales, we are confident that we have the same species as the one studied by Stephenson, Erséus, and possibly Southern, but that this (new) species is different from the original Pachydrilus semifuscus. Considering the way that it has been misidentified throughout history, we have named it semifuscoides and, at the same time, established a new genus for it called Claparedrilus. However, based on the similarities in the spermathecae and nephridia, we have decided to also transfer L. semifuscus into this genus, making it C. semifuscus (Claparède, 1861) comb. nov.
Claparedrilus semifuscoides can be separated from C. semifuscus by the size of the penial bulbs, where the former species have bulbs about the same size as the sperm funnels (about 150 µm in diameter) whereas the latter have bulbs larger than the funnels; they are 400-500 µm long. The nephridium illustrated by Claparède is reminiscent of what we observed (Fig. 24F-G), although our specimens seem to have the septa further back in relation to the funnel, making the funnel appear with a thin stalk, and with the efferent duct originating much further back on the postseptale. Unfortunately, Claparède did not mention the number of chaetae, subneural glands, or pharyngeal glands for his species, which makes its placement into the new genus Claparedrilus a bit tentative. This, and the fact that we have no genetic information for C. semifuscus, are the reasons why we designated the new taxon, C. semifuscoides, as the type species of the new genus.
Our specimens of C. semifuscoides are smaller than the ones described by Stephenson as M. semifusca, and they possess fewer chaetae, but we still believe that they belong to the same species. Stephenson remarks that (1) the nephridia can be found from V, (2) the anteseptale is made up of funnel only, and (3) the efferent duct extends backwards towards the pore, not forwards as illustrated by Claparède for C. semifuscus. We found nephridia from VI (possibly not finding any in V because they were obscured by the pharyngeal glands) and observed that the anteseptale consists of a funnel on a thin stalk. As this character was difficult to see and because there is no true nephridial tissue anterior to the septa this could still have been interpreted by Stephenson as a funnel only. We found that the efferent duct extended forward towards the pore which is more in agreement with Claparède's illustration of C. semifuscus than what Stephenson noted, but the interpretation of this character may differ as the animal extends or contracts. Finally, Stephenson stated that the efferent duct originates well in front of the middle of the postseptale, whereas we observed it originating behind the middle or even from the posterior end. However, Stephenson also noted that this was not apparent in living specimens and only became clear from sections, which we have not studied.
Our specimens of C. semifuscoides also largely agree with Southern's account of M. semifusca except for his description of the cylindrical penial bulbs. It is possible that the bulbs he studied were everted (as illustrated for L. pagenstecheri A in the present study; Fig. 14), which may have given the impression of them being cylindrical rather than spherical. Unfortunately, Southern did not mention the exact size of the bulbs or their size in relation to that of the sperm funnels, which makes us unable to confidently conclude that his species is the same as ours.
The species reported as L. semifuscus from Iceland (Erséus 1976) is probably the same as our C. semifuscoides, as Erséus noted the four pairs of pharyngeal glands and an anteseptale with more than just a funnel. From his sectioned material he stated that the anteseptal portion was made up of a few coils of the nephridial canal in addition to the funnel, something we could not make out in our whole-mounted material. Another illustrated nephridium that resembles the one in our species was provided by Finogenova and Timm (1988) who reported L. semifuscus from the White Sea (Russia). Their description only differs from that of C. semifuscoides in reporting three pairs of pharyngeal glands, with the third pair extending into VII, but this could be a misinterpretation of a fourth pair. As we have not examined their material, we cannot be certain that this is the case, but we find it highly probable that they actually were describing C. semifuscoides. Finally, we are not certain about the identity of the species "Lumbricillus semifuscus" that Nurminen (1965a) reported from Spitsbergen (Svalbard), as he mentions sperm funnels "considerably longer than 1.5 times the width".
Compared to the species of Lumbricillus, i.e., the genus in which we previously placed this species (and erroneously referred to it as L. semifuscus), C. semifuscoides can be distinguished mainly by its four pairs of pharyngeal glands, the stalked nephridial funnel, and the irregularly lobed testes.

General comments on Lumbricillus taxonomy
Klinth et al.'s (2017) molecular assessment of Lumbricillus was a starting point for the present taxonomic study of the genus, which also has taken the species morphology into account. The lineatus, pagenstecheri, buelowi and arenarius species groups, which are monophyletic according to Klinth et al. are also supported by consistencies in the morphological characters (Fig. 1, Table 1). The fifth, but non-monophyletic, "tuba" group appears morphologically coherent too. In Table 1, 65 additional species of Lumbricillus not studied herein, but with some exceptions regarded as valid by Schmelz and Collado (2012), are tentatively classified into these five groups after considering their original descriptions. It is likely that the placement of many of them will be challenged in the future after molecular and morphological examination, and that some of them will even prove to represent lineages of Lumbricillus not covered by our sample of care-fully studied taxa. Finally, in Table 1, yet five other species with uncertain affinities are listed. The long needed re-assessment of this large genus has just begun.
Diagnosing and delimiting Lumbricillus are problematic, as one of its most striking features, the bunch-like arrangement of the lobed testis sacs, is not shared by all species in the genus. This character appears to be a synapomorphy of the lineatus, pagenstecheri and "tuba" groups, which together make up a monophyletic clade, but excludes the buelowi and arenarius groups, which have unlobed or irregularly lobed testis sacs (Fig. 1). The buelowi group has been supported as the sister to the lineatus, pagenstecheri and "tuba" groups, sharing most characters except for the bunch-like arrangement of the testis sacs. The two species in this group studied (L. buelowi and L. knoellneri) are very small, and it seems that each of their testis sacs comprises a single large lobe, or possibly numerous shallow lobes. However, body size and number of testis lobes do not seem to be functionally correlated because there are other small Lumbricillus species, such as L. pumilio within the lineatus group, which despite their small size have testes with bunch-like arrangement. Klinth et al. (2017) discussed the possibility of splitting off the arenarius group as a separate genus, especially as this group earlier had been suggested as the sister to Grania (Erséus et al. 2010). However, the most reliable phylogenetic topology recovered by Klinth et al. (2017) still favored the arenarius group as sister to the remaining Lumbricillus. Having studied the morphology of the species included in the arenarius group we conclude that they resemble Grania in some morphological characters, such as nonsigmoid chaetae and few chaetae per bundle. However, in most other characters, first of all the organization of sperm in the spermathecae, Grania is clearly different from the arenarius group. Furthermore, the arenarius group has testes encased in testis sacs, which is only found in Lumbricillus and Enchytraeus (Stephenson 1930, Tynen et al. 1991, and now also in Claparedrilus (present study). Therefore, we propose to keep the three species of the arenarius group within Lumbricillus, thereby retaining the broad morphological definition of the genus. The different location of the upper chaetal bundles in the arenarius and buelowi lineages as compared to the other Lumbricillus species may simply reflect their adaptation to a mineral, sandy bottom rather than to a loose, water-saturated, organic-rich substrate (for a discussion on this topic see Rota 2001).

Comments on Enchytraeoides
Before we came to the conclusion that the arenarius group should remain within Lumbricillus, we explored the possibility that it could be treated as a separate genus. Searching the taxonomic literature for possible candidate genera we found Enchytraeoides, in which L. arenarius had previously been placed by some authors (Ude 1929, Knöllner 1935, von Bülow 1957. Due to the similarity in morphology between Enchytraeoides and the arenarius group, it is important to determine the proper status of Enchytraeoides in relation to Lumbricillus. It is difficult to establish whether Pachydrilus enchytraeoides Saint-Loup, 1885 is the same species as Enchytraeoides marioni Roule, 1888, but through personal communications with Roule, Delphy (1920) was persuaded that the former had studied the same species, a species that Delphy pointed out should therefore be named Enchytraeoides enchytraeoides (more on this in Rota et al. 2008). This taxon, which was described from Marseille, resembles the arenarius group in some traits; mainly the irregularly lobed testis sacs, sperm funnels that are several times longer than wide, and spermathecae with what appears to be gradually widening ducts that are difficult to distinguish from the ampulla. However, E. enchytraeoides was described as having two to eight chaetae per bundle, which is much more variable than the two to three chaetae per bundle in the arenarius group.
Roule described the penial bulbs as situated in segment XI, a trait neither seen in any Lumbricillus nor in any other typical enchytraeid in which the bulbs are in XII. Furthermore, there were some discrepancies in Roule's description as to the location of the spermathecae, which were described as located in segment VI, but were illustrated as in segment V. This could be explained by Roule counting the prostomium as a separate segment, but this still does not explain the placement of the penial bulbs which are also illustrated as part of segment XI. If the material was studied live without proper magnification, and the description and illustration were produced later, this might have caused a misinterpretation of the position of the penial bulbs. However, given the extensive descriptions and illustrations based on such large amounts of material this seems unlikely. Therefore, the true phylogenetic placement of Enchytraeoides remains uncertain until newly sampled species have been examined and sequenced and for now it should remain as a junior synonym of Lumbricillus.

Geographical distribution and habitat
It is difficult to make any conclusions of the full geographical distribution of the species in this study, as most of our samples are from Norway and Sweden, with some also from the United Kingdom and other parts of Northern Europe. When taking into account the BOLD data and the reports by other authors it seems that some species are very common, such as L. lineatus, L. rivalis and L. pagenstecheri sensu lato (e.g. Timm 2005). These taxa have been found throughout Europe and in some cases also in North America and the northern Pacific region. Unfortunately, for most reports we do not have DNA sampled specimens to compare with and we are unable to ensure that it is indeed the same species as ours that have been found at some remote location. Fortunately, we could find matching barcoding sequences in the BOLD database for several species that were also collected in Canada, such as L. lineatus, L. rutilus, L. rivalis and L. pagenstecheri sensu lato, showing how widespread these species can be and supporting the idea of a partly Holarctic fauna. There are few reports of Lumbricillus from tropical areas (possibly due to poor sampling), but species have been reported from the Southern hemisphere, mainly from islands surrounding Antarctica. It would be interesting to compare the genetic information of these species to see if they are parts of, or lineages separate from, the species occurring in the Northern hemisphere. For the L. pagenstecheri complex, in which we found support for four different species, our material as well as the many descriptions of species that are very similar to this species seems to indicate that this group is more diverse in the Arctic than it is further south (see for example Shurova 1974Shurova , 1977Shurova , 1978Shurova , 1979. As evident from Appendix I and the taxonomic literature on Lumbricillus, this genus is mostly associated with seashores and brackish waters, but many species (such as L. arenarius, L. fennicus, L. knoellneri, L. rutilus and L. scandicus) are commonly encountered also in freshwater habitats.

Future research
Having studied only about one fourth of the 80 or so described species of Lumbricillus with a primarily molecular approach, it is clear that a lot of taxonomic and genetic work remains to be done on this genus. Ideally, we would like to be able to link each of the described species to one or more molecular barcodes of COI, and to clearly delimit the species from each other using also nuclear genetic data (as in Klinth et al. 2017). There are several undescribed species left in this genus and we have several unnamed species that could either belong to previously described species or be new to science. A majority of the previously described species identified in our study lack types or other reference material. Thus, a future challenge is to visit type localities of such species, to determine if we truly have the same species or not. Lastly, some groups within Lumbricillus, such as L. pagenstecheri sensu lato, require some extra attention. For this complex, there are several synonymized names to examine in order to see if they match our specimens. Furthermore, there are many species described from the Pacific coast of Russia (Shurova 1974;1978;1979), and they all seem to share a morphology similar to that of L. pagenstecheri, and molecular data will be needed to determine the true number of species in this complex.

Conclusions
Having studied the morphology of the Lumbricillus species included in the molecular study by Klinth et al. (2017) we have found that this genus can be divided into at least five informal subgroups with differing morphology, which may become a useful backbone when resolving the taxonomy and phylogeny of other species in Lumbricillus. We propose to keep L. arenarius, L. dubius and L. sp. H within Lumbricillus despite their somewhat differing morphology, and their uncertain phylogeny vis-à-vis Grania. This means that the morphological characters defining Lumbricillus remain quite broad. We have described two new species of Lumbricillus: L. latithecatus sp. n., which is somewhat reminiscent of L. lineatus, and L. scandicus sp. n., which was previously thought to be the same as L. helgolandicus. Lumbricillus verrucosus, which was resurrected from junior synonymy with L. lineatus by Klinth et al. (2017), has been given a proper morphological re-description. Lumbricillus lineatus remains as the type species for the genus and a neotype has been designated. The genera Pachydrilus and Enchytraeoides remain as junior synonyms to Lumbricillus. Lastly, we have established Claparedrilus gen. n., with C. semifuscoides sp. n. as the type species, and transferred Pachydrilus semifuscus Claparède, a former member of Lumbricillus, into the same genus.