Origin of the pointed snout in Scolelepis and the mouth region in spionid polychaetes (Annelida: Spionidae)

Abstract

Adult Scolelepis are unique among spionids in having an elongated, pointed snout, used for digging in sediment. Earlier studies have suggested that the pointed part of the larval head was a peristomial extension. Despite this, modern authors characterize larvae and adults of Scolelepis as having an elongated, pointed prostomium. Based on new observations, as well as on literature data, I show that the pointed medioventral process of the head in larvae is formed by lateral peristomial lips, elongated and fused anteriorly, each supported internally by conspicuous fibres. During settlement and metamorphosis, the anterolateral parts of the prostomium extend ventrally and overgrow the basal part of the peristomial process in a tube-like manner, forming a typical adult snout. Thus, in adult Scolelepis, only the wide basal part of the snout is formed by the prostomium, while the anterior, pointed part of the snout (the rostrum) is of peristomial origin. The ventral peristomial lip does not develop in Scolelepis larvae, and the ventral part of the mouth in adults is formed by the anterior extensions of the first and second segments. The diverse composition of the mouth region of different spionids is illustrated, as well as other details of the morphology of Scolelepis.

INTRODUCTION

Spionidae Grube 1850 is one of the largest polychaete families (Annelida), comprising about 600 described species worldwide (Blake et al. 2020; Rouse et al. 2022). Adult spionids live in a wide variety of habitats in benthic communities from the intertidal to the deep sea, while their larvae are extremely common in coastal waters. Most spionids move freely in soft sediments or dwell in temporary burrows or permanent tubes, while others bore into mollusc shells, barnacle tests, corals, sponges, brachiopods, tunicates, coralline algae, and rocks. The fascinating diversity of habitats and reproductive characteristics of spionids, along with their ubiquity, have made them objects of numerous studies since the 18^th century (Blake et al. 2020).

Adults of the spionid genus ScolelepisBlainville 1828, which includes more than 80 known species, are common inhabitants of exposed sandy beaches in some temperate, tropical, and subtropical regions (Blake et al. 2020). In the north-east Atlantic they constitute an important food resource for juvenile flatfish and wading birds, thus playing an important role in sandy beach ecosystems (Speybroeck et al. 2007). They are easily distinguished from other spionids by their elongated, pointed head. When washed out by strong waves, the worms quickly return to the sediment, wriggling their body and digging with their head. Scolelepis is a genus of mainly broadcast spawners, and most have planktotrophic larval development. Elongation of the initially short, rounded head and its transformation into a pointed head occurs in the larval development of these worms. Thus, the settling and metamorphosing larvae can burrow into the sediment to start a new benthic life. All the authors who described Scolelepis larvae in the 19th and in the first half of the 20th centuries noted that the pointed frontal part of the larval head is formed by a medioventral peristomial process (e.g. Hannerz 1956, as NerineJohnston 1838 and NerinidesMesnil 1896). However, all subsequent authors have characterized the larvae and adults of Scolelepis as having an elongated, pointed prostomium (e.g. Blake et al. 2020).

The origin of the pointed head in Scolelepis, which I prefer to call by the neutral term ‘snout’, is irrelevant for identification purposes or ecological considerations. However, it becomes important when it comes to phylogenetic consideration and character coding for phylogenetic analysis. For example, two analyses of the morphological characters of Spionidae and closely related taxa performed to date by Sigvaldadóttir et al. (1997) and Blake and Arnofsky (1999) scored Scolelepis as having a pointed prostomium, in contrast to the majority of other spionids having other kinds of prostomia (specified differently in these two analyses).

In the most comprehensive review of the larval morphology of Spionidae members, Hannerz (1956: 186, fig. 57) described and illustrated the development of the mouth region during the metamorphosis of Spio martinensisMesnil 1896. Based on these observations, he generalized for all spionids that ‘In conjunction with the metamorphosis …. The ventral [peristomial] lip grows forward over the lateral ones, which thus will be drawn into the pharynx’ (Hannerz 1956: 157). My preliminary study of various adult spionids showed the diversity of the composition of their mouth region (Radashevsky 2012). This diversity suggests that the idea of Hannerz (1956) about the transformation of the larval peristomium into the adult mouth region may be correct only for some species or genera of Spionidae.

The main purpose of the present study was to examine the development and transformation of the head in the early ontogenesis of Scolelepis in order to elucidate the origin of the pointed snout in their adults, as well as to describe other diagnostic morphological features of these worms. My further goal was to elucidate the composition of the mouth region in different spionids.

MATERIAL AND METHODS

Collections were made in the Caribbean Sea, Belize, in 1997; Iceland, in 1998; Brazil, in the states of Paraná, São Paulo and Rio de Janeiro in 1998, 2001–10; and in the Sea of Japan and off the Kurile Islands, Russia, in 1980–2010. Plankton tows were made in coastal waters with a 100-µm mesh plankton net. Sediments were collected from intertidal habitats and also from the shallow subtidal using SCUBA equipment. Sediments were sieved in the field on a 300-μm mesh sieve, and the residue was taken for further examination in the laboratory. Live larvae, juvenile, and adult spionids were removed from the residue, relaxed in an isotonic solution of magnesium chloride, and examined and photographed using compound light microscopes equipped with digital cameras. Images of multiple focal layers were stacked using Zerene Stacker 1.04 software. The successive stages of development and metamorphosis described here are not a series of growth of a single specimen, but selected individuals of various sizes from natural habitats. After examination, adult worms were fixed in 10% formalin solution, then rinsed in fresh water and transferred to 70% ethanol. Preserved specimens were deposited in the polychaete collections of the Museum of the A.V. Zhirmunsky National Scientific Center of Marine Biology (MIMB), Vladivostok, Russia; the National Museum of Natural History, Smithsonian Institution (USNM), Washington, DC, USA; Museu de Zoologia da Universidade de São Paulo (MZUSP), São Paulo, Brazil; and Senckenberg Museum, Frankfurt am Main (SMF), Germany. Additional specimens in the USNM collection were also examined (see Table A1 in the Appendix).

Formalin-fixed specimens of Scolelepis species and other spionids were critical point-dried in carbon dioxide, coated with gold palladium, and viewed with a LEO 440 scanning electron microscope equipped with a digital camera at the USNM. Final plates were prepared using CorelDRAW®2021 software.

RESULTS

Head formation in Scolelepis

Scolelepis larvae are easily recognized in plankton samples by their prostomium, which extends into an anteriorly directed conical tip, and has a ventral groove that encases an acutely pointed medioventral process (Figs 1A–C, 2A–F, 3A–E). This process appears at about the 2-chaetiger stage and elongates in larval development. Larvae of nine Scolelepis species, including one species from Belize, six species from Brazil, and two species from Russia (Sea of Japan), were examined in this study. Larvae of only two species from Brazil were identified: S. quinquedentata (Hartmann-Schröder 1965) and S. goodbodyi (Jones 1962). These larvae, as well as those described in the literature, showed similar processes of development and head transformation. These processes are summarized and illustrated below.

The medioventral process characteristic of Scolelepis larvae is formed by lateral peristomial lips, elongated in the longitudinal direction, narrowed and fused anteriorly along the midline (Figs 2E, F, 3C). Most of the lateral peristomial lips lie in front of the mouth, not lateral to it as in other spionid larvae. Both the anterior process and the posterior part of the lips are well demarcated from the prostomium (Fig. 2F) and are internally supported by conspicuous fibres. These fibres start from the posterior, outward sides of the peristomial lips and pass anteriorly and towards the medial part of the medioventral process. They converge along the anterior-medial part of the process and on their distal end support a small tuft of short, non-motile, probably sensory cilia (Fig. 3B, C). The fibres are able to retract the tuft of cilia into the medioventral process. It is not known, however, if they do the contracting or if they are connected to muscles that do so. The retraction is slight, and possibly serves only to hide and protect the cilia, with no effect on the shape and size of the process. When the tuft of cilia is pulled back, a small hole can be seen at the distal end of the process (Figs 2B, E, F, 3B). At the end of larval development, the peristomial process exceeds the length of the prostomium and extends beyond it (Fig. 3D, E).

In the course of settlement and metamorphosis, the medioventral peristomial process further elongates into a distally pointed solid rostrum, at which point the inner fibres are no longer visible (Fig. 3D–F). The posterior parts of the peristomial lips, on the contrary, become shorter, semi-oval, slightly delimited from the rostrum, and cover the rostrum’s most posterior part from the sides (Fig. 3D). In juveniles, the anterior conical tip of the prostomium retracts backwards, so the frontal edge of the truncate prostomium becomes rounded to blunt. The anterolateral parts of the prostomium extend ventrally and overgrow the basal part of the peristomial rostrum laterally (Fig. 3D). Eventually, the two prostomial extensions meet and fuse midventrally, forming a tube-like envelope around the base of the peristomial process (Fig. 4A). The two structures remain well distinguished in recently settled juveniles (Fig. 3D, E) and even in adults of some species (Figs 3F, 4B–D, 5D), but they often fuse completely in adults, obscuring their true origin (Figs 4E, F, 5F, G).

The larvae of S. goodbodyi are unique in having two peristomial processes situated one dorsal to the other (Fig. 3A, B). Only the ventral process is internally supported by fibres. During metamorphosis, the dorsal process is completely reduced, while the ventral process extends into a pointed rostrum (Fig. 3D–F).

With further development of juveniles, the head elongates and transforms into a typical adult snout, the wide basal part of which is formed by the prostomium, while the long, pointed rostrum is a derivative of the medioventral peristomial process (Figs 4A–F, 5D, F, G). The posterior parts of the larval lateral peristomial lips form the lateral parts of the mouth (Figs 4A–F, 5D, F, G). The ventral peristomial lip, which remained rudimentary at the larval stage, does not enlarge during metamorphosis but becomes completely reduced. Instead, the ventral parts of the first and second segments elongate anteriorly, forming the ventral part of the mouth (Figs 4D–F, 5G).

Other morphological features of Scolelepis

The larvae of Scolelepis examined in this study and described in the literature have a body wall from transparent to opaque without distinct pigmentation. However, in some species, large, distinct melanophores or yellow chromatophores are present in the prostomium, palpophores, and pygidium, and on the dorsal side of the segments. Distinct dark green, orange and blue chromatophores were observed on the dorsum of larvae of one unidentified species from Brazil. To the best of my knowledge, these pigments have never been described in any spionid larva. A remarkable feature of Scolelepis larvae is the intense colourful pigmentation of their digestive tract. Diffuse black, dark green, or brown pigments are often present in the wall of the oesophagus, midgut, and rectum (Fig. 1A–C). The presence of certain pigments in the body wall or in certain parts of the digestive tract can be used for the specific identification of larvae. Larval pigmentation in the body wall disappears during metamorphosis or shortly thereafter. New pigments develop in the adults of some species. Those may be bright white or yellow chromatophores evenly distributed inside the coelomic cavities of the palps, or small, black grains scattered on the palps, head, segments, and pygidium. However, most described Scolelepis adults have a transparent or brownish-green body wall without any particular distinct pigmentation.

All examined Scolelepis larvae have two pairs of red, cup-shaped eyes, arranged approximately in a straight line across the prostomium (Fig. 3A, B), and one pair of spherical, unpigmented ocelli situated in front of the pigmented eyes. Pigmented eyes are multicellular, comprising one pair of median eyes that develop first in the trochophore stage, and one pair of slightly larger lateral eyes that develop shortly thereafter. Unpigmented ocelli are as those described and illustrated by Radashevsky and Migotto (2006: fig. 1). As with other spionids, in Scolelepis larvae the ocelli become visible shortly after pigmented eyes develop. They increase in size as the larva grows and probably disintegrate during metamorphosis, since they have not been observed in settled juveniles and adults. In contrast, pigmented eyes persist throughout life, usually darkening with age. The aperture of the pigment cup is directed differently in different pairs of pigmented eyes. The cups of the median eyes are open upwards; the lateral ones are open to the sides (Fig. 3F).

The palps in Scolelepis larvae develop from short palpophores extending from the posterolateral parts of the lateral peristomial lips (Figs 2A–F, 6A). The prototroch is represented by two contiguous ciliary bands running along the dorsal side of the palpophores. During metamorphosis, the posterolateral corners of the lateral peristomial lips shift dorsally due to the enlargement of the anterior-ventral parts of the first and second segments (forming the ventral part of the mouth), and slightly posteriorly due to the posterior elongation of the prostomium into the caruncle [the narrow posterior part of prostomium extending onto the anterior segment(s)]. Consequently, the bases of the palps in adults are situated dorsally, at the level of the first segment, on the sides of the caruncle (Figs 4A–D, 5A–D, F, G). The palps in Scolelepis are unique among spionids in that they have no longitudinal groove on the frontal surface lined with motile cilia that transport food particles to the mouth. Instead, they have numerous short, transverse ridges arranged in two or more rows along the entire length of the palp (Fig. 5A–C, E). The height and length of the ridges and the number of their rows vary among species (Williams 2007).

The nuchal organs in Scolelepis larvae are represented by two almost straight, narrow ciliary bands (Fig. 2B–F) or oval to rounded ciliary patches extending along the posterior part of the prostomium from the base of narrow caruncle to the bases of the palpophores. A single boss with numerous non-motile cilia is present within each patch of motile nuchal cilia (Fig. 3C). The same bosses are also present in nuchal ciliary patches of the larvae of other spionids, e.g. Polydora rickettsiWoodwick 1961, Polydora cornutaBosc 1802, Prionospio patagonicaAugener 1923, Rhynchospio nhatrangi Radashevsky 2007, Pseudopolydora rosebelae Radashevsky and Migotto 2006, Pseudopolydora arabica Radashevsky and Al-Kandari 2020, Rhynchospio glandulosa Radashevsky and Choi 2021 (Radashevsky and Cárdenas 2004; Radashevsky 2005, 2007; Radashevsky and Migotto 2006; Radashevsky et al. 2006; Radashevsky and Al-Kandari 2020; Radashevsky and Choi 2021). During metamorphosis, the bosses with non-motile cilia are reduced and are no longer seen in adults; patches of motile nuchal cilia transform into a pair of U-shaped ciliary bands with their arms directed forward, extending over the anterior segment(s) on the sides of the caruncle (Figs 4C, 5F). The length of these organs is one of the diagnostic features of adult spionids.

Nototrochs are weakly developed or completely absent in Scolelepis larvae, especially in those with thick, strong, serrated provisional bristles (Fig. 2A–D). These provisional bristles completely cover the dorsal side of the larvae when swimming, perhaps making the larval body firm and increasing swimming efficiency. In addition, these bristles appear to serve as a defence device against small pelagic predators. In case of danger, the larvae curl up into a ball, exposing the bristles. However, the use of the bristles as an anti-predator defence has never been experimentally tested. Provisional bristles are longest on the first chaetiger, and gradually decrease in length on succeeding chaetigers. From one to several shorter and thinner provisional bristles are also present in neuropodia, beginning from chaetigers 2–3 (Figs 2A, B, 6A, B). Scolelepis larvae with short, thin provisional bristles usually have short-cilia nototrochs, beginning from chaetigers 2–3. Adult capillary chaetae and hooded hooks develop in advanced larvae, usually among the provisional bristles. Provisional bristles fall out during settlement and metamorphosis.

Unlike nototrochs, gastrotrochs and telotroch are strongly developed and serve as the main organs of locomotion in Scolelepis larvae. Gastrotrochs are arranged on a series of chaetigers, beginning from chaetigers 2–3 (Figs 1A, C, 2A, E). They are composed of single rows of large, oval cells, each bearing numerous long cilia beating in the longitudinal direction. The telotroch also consists of one row of large cells with long cilia (Fig. 2A). It is interrupted mid-dorsally; the dorsal gap is used to accommodate the long larval bristles when swimming.

Scolelepis larvae have a short neurotroch that extends posteriorly from the mouth and terminates at chaetiger 2 in a cavity in the form of a slit or keyhole, the so-called ciliated pit (Fig. 2E). The pit is posteriorly accompanied by a large, oval cell, transversally elongated, with numerous long cilia actively beating towards the pit centre. Two smaller cells with shorter cilia are situated on the anterolateral sides of the ciliated pit; the long axes of these cells are oriented obliquely, so their cilia are also directed towards the pit centre. One or two pairs of even smaller ciliated cells are situated on the sides of the neurotroch on chaetiger 1. The ciliated pit is a common opening for two complex systems of mucous glands situated internally in chaetiger 1 or in chaetigers 1 and 2 (Hannerz 1956: 161–167, figs 54, 55). This glandular complex is a larval organ that is present in some genera of Spionidae but is not known in other annelids. It reaches its fullest development immediately before the larval settlement and disappears during metamorphosis (Hannerz 1956). The function of this complex remains unknown.

Like all other spionids, Scolelepis have lateral ciliated organs (also called ciliated sensory organs or inter-ramal ciliated organs) situated midway between the noto- and neuropodial rami (Figs 4D, F, 5G). The cilia of the organs are fine, short, and non-motile. These organs first appear in early segmented larvae; in adults they appear on newly developed segments before chaetae or any other structures have developed. In most of the studied spionids, the lateral organs externally look like small, ciliated pits, each sitting in a depression of the body wall. In the only examined Scolelepis larva from Belize, the organs appeared on top of short papillae (Fig. 6A, B), although in all examined adults, including two species from Belize, they sit in the body wall in depressions. It remains unknown whether the unusual shape of the organs in the Belizean larva is a fixation-induced artefact or a normal condition that transforms into a typical pit during the metamorphosis.

The larvae of different Scolelepis species become ready for metamorphosis at different stages (see review by Blake and Arnofsky 1999: appendix 1: 91–93). Some correlations can be seen with the geographic distribution of species: worms in temperate waters are ready for metamorphosis after reaching the 15–20-chaetiger stage, while worms in tropical and subtropical waters grow in the plankton until about the 30-chaetiger stage. Competent larvae from Belize and Brazil examined in the present study had 30–32 chaetigers. Maintained in the laboratory, they continued to grow slowly, but were difficult to stimulate to settle and metamorphose.

Of nine studied species of Scolelepis, competent larvae of only one species had a straight, slender body similar to the larvae of other spionids. In other larvae, after reaching a certain stage, the newly developing segments were swollen ventrally and usually differed in chaetal composition from the anterior segments. Larval bristles usually did not develop in these segments, but hooded hooks began to develop in these segments. Remarkably, this new posterior body was tucked under the ventral side of the anterior segments (Figs 1C, 2A). Larvae with the posterior part of the body bent forward swim in circles.

Composition of the mouth region in other, non-Scolelepis spionids

Transformation of the larval head during metamorphosis has been described in some other spionids, species of PolydoraBosc 1802 and PseudopolydoraCzerniavsky 1881 (e.g. Radashevsky and Cárdenas 2004; Radashevsky 2005; Radashevsky and Migotto 2006; Radashevsky et al. 2006; Radashevsky and Al-Kandari 2020). It is similar to the general scheme proposed by Hannerz (1956: 186, fig. 57) based on his study of Spio martinensis, and differs from that described above for Scolelepis. In the course of this study, I examined the metamorphosis of various other spionids, but below I show only the final composition of the mouth region of their adults.

The larvae of Spionidae are characterized by a pair of well-developed lateral peristomial lips and a weakly developed ventral lip (Fig. 7F, G; see also: Hannerz 1956; Blake 2006). Rapid reduction of their lateral lips and enlargement of the ventral lip during metamorphosis, as described by Hannerz (1956) for Spio martinensis, occurs in many but not all spionids. The degree of reduction of the lateral lips varies greatly. As in Spio martinensis, the larval lateral peristomial lips transform into prominent, narrow structures on the dorsolateral sides of the mouth in adults of DispioHartman 1951, MalacocerosQuatrefages 1843, MicrospioMesnil 1896, and BoccardiaCarazzi 1893 (Fig. 7H). These structures are clearly visible from the ventral side and are usually also from the lateral side of the head (Figs 8A–D, 9A–D). The lateral peristomial lips of the larva become large parts of the dorsolateral parts of the mouth in adult Scolelepis (Figs 4A–F, 5D, F, G, 7E). In adult AonidesClaparède 1864, these become small structures at the dorsolateral corners of the mouth (Figs 7J, 10C, D). In Polydora and DipolydoraVerrill 1881, the larval lateral peristomial lips fully retract during metamorphosis and form longitudinally elongated dorsolateral ciliary folds in the adult foregut. These folds are visible in adults only from below (Figs 7I, 9E) (see also Purschke and Tzetlin 1996).

During the metamorphosis of ParaprionospioCaullery 1914 larvae, their lateral lips fully retract and become part of a complex bifurcated pharynx used to sort food particles and transport them into the gut (Fig. 10A; see Dauer 1985: figs 4, 6, 7 for details of feeding behaviour). The ventral peristomial lip greatly enlarges and fuses with the first body segment, forming a large hood covering the prostomium from the sides and, in some species, dorsally (Fig. 10B; Dauer 1985: figs 4, 6; Radashevsky 2012: fig. 3E). In larvae of Paraprionospio, the first segment bears long, provisional bristles in notopodia (Yokoyama 1981, 1996); in adults this segment is devoid of chaetae and is fused with the ventral peristomial lip but can still be recognized by the presence of lateral ciliated organs (Fig. 10B). The functional significance of the hood formed by the fusion of the ventral peristomial lip and the first segment has never been discussed. It may serve as a strong base for the palps and their large basal sheath, the function of which is ‘to hold the flexible end of the burrow open during feeding’ (Dauer 1985: 144, fig. 7).

In LaoniceMalmgren 1867, the lateral lips of the larvae appear to be completely reduced during metamorphosis; the mouth region in adults is completely formed by a large, ventral lip (Fig. 8E, F). Complete reduction of both the lateral lips and the ventral lip of the larvae occurs during the metamorphosis of PoecilochaetusClaparède 1863 [this genus is treated here as a member of the family Spionidae, as suggested earlier by Radashevsky (2012)]. The entire mouth region in these worms is formed by the first segment. Anteriorly, the mouth is bounded by the base of the median facial tubercle, which apparently arises from the ventral side of the prostomium (Fig. 10E, F).

In the larvae of Spionidae, the roof of the mouth (the space between the lateral peristomial lips) is flat and lined with short cilia actively beating towards the gut. In adults of some species, this space remains flat, but strongly reduced, squeezed between the lateral parts of the mouth (Figs 8A, C, 10A, E). In others, it appears as a dorsal ciliary fold (Figs 8E, 9A, C–E, 10C) or ‘zungenförmiges Organ’ (tongue-shaped organ) (Hempel 1957: 102, figs 2, 3), which may be used as part of the mouth-sorting apparatus.

DISCUSSION

The pointed snout and the mouth region in Scolelepis

None of the authors who described and illustrated the larvae of Scolelepis reported the transformation of the larval head structures into the adult head during settlement and metamorphosis. This transformation, described for the first time in the present study, clearly shows that both the pointed medioventral process of the head in larvae and the anterior, pointed part of the snout (rostrum) in adults are derivatives of the lateral peristomial lips, i.e. of peristomial origin.

Busch (1851: pl. VIII, figs 1–4), Leuckart (1855: pl. II, figs 1–5), and Claparède (1863: pl. VII, fig. 1) were the first to describe and illustrate fully developed Scolelepis larvae from European waters. Busch (1851) assigned his larva to the type of Sars-Lovén larvae; Hannerz (1956: 18) suggested that Busch (1851) might have examined Nerine cirratulus (Delle Chiaje) [=Scolelepis squamata (Müller, 1806)]. Leuckart (1855) suggested that his larvae from Mediterranean France could have belonged to either Nerine or SpioFabricius, 1785. Claparède (1863: 86) simply called his larva from Atlantic France ‘eigenthümliche Larve (Taf. VII. Fig. 1–2) mit rüssellartigern Kopflappen’ (a peculiar larva with a proboscis-like head lobe).

Although some of the details of Leuckart’s (1855) description of the larvae seem confusing for modern readers, Leuckart (1855: 65) was the first to call the ‘lip-shaped bulge in front of the mouth’ the ‘upper lip’: ‘Das vorderste Körpersegment verlängert sich auf der Bauchfläche fast bis an die Basis des Stirnfühlers und bildet hier (Fig. 2.) vor dem Munde eine lippenförmige Aufwulstung, die schon oben erwähnte Oberlippe, die zugleich die Grenze zwischen Mundsegment und Kopfhöcker bezeichnet.’ Claparède (1863: 113, pl. VII, fig. 1), in the legend for the illustration of his larva called it ‘Larve mit rüsselartiger Oberlippe’ (larva with trunk-like upper lip). Thus, both Leuckart (1855) and Claparède (1863) considered the anterior process of the Scolelepis larvae to be the elongated upper lip, which we now call part of the peristomium.

The peristomial origin of the frontal rostrum of Scolelepis larvae was also pointed out by Gravely (1909, Spionid D in his paper) and Hannerz (1956, Nerine and Nerinides in his paper). Gravely (1909: 609, pl. 14, fig. 3) perfectly illustrated the anterior-ventral side of the developed Scolelepis larva and noted: ‘from their appearance during life I am led to believe that the “snout” [the anterior extension of the lip] is free from the prostomium at the extreme anterior end’. Despite these early assumptions about the peristomial nature of the frontal medioventral rostrum in Scolelepis larvae, all subsequent authors referred to this larval rostrum as a prostomial structure (Dean and Hatfield 1963; Bhaud and Lefèvre 1986; Bhaud and Cazaux 1987; Scheltema et al. 1997; Bhaud et al. 1999; Blake and Arnofsky 1999; Pernet et al. 2002; Blake 2006; Abe and Sato-Okoshi 2021). Scheltema et al. (1997: 401) noted that ‘No functional role for this structure has ever been postulated’ and suggested that ‘the prostomial process [in Scolelepis larvae] is sensory in nature’.

Hannerz (1956), who provided the first and most comprehensive review of the larval morphology of Spionidae and closely related taxa, avoided any speculation about the transformations of the head in the early ontogenesis of Scolelepis, but simply described the initial and final conditions. For the developed pelagic larvae, Hannerz (1956: 8, 157) summarized that ‘The prostomium is drawn out into an anteriorly directed lip encasing a muscular process originating from the peristomium’ and this peristomial process ‘is supported by two longitudinal muscular bands’. For a 53-segment juvenile of Nerine species I, Hannerz (1956: 19) noted: ‘The prostomium was sharply tapered and the lateral parts of the peristomium were no longer clearly demarcated from the prostomium’. In the analytic section of his work, summarizing the external morphology of the larvae, Hannerz (1956: 155) noted that in Nerine larvae ‘a median process develops ventrally from the anterior part of the peristomium …. In connection with the growth of this process there is also an anterior extension of the prostomium. It becomes terminally tapered and acquires on the under side a groove, which will partly enclose the peristomial process’. No details of further changes in the larval head were given, but then Hannerz (1956: 157) generalized for all spionids that ‘In conjunction with the metamorphosis . . . . The ventral [peristomial] lip grows forward over the lateral ones, which thus will be drawn into the pharynx’.

Notes by Hannerz (1956) show he believed that during metamorphosis the anteriorly tapered prostomium of Scolelepis larvae forms the pointed head of adult worms, while, like in all other spionids, the lateral peristomial lips are strongly reduced and ‘drawn into the pharynx’. This opinion of Hannerz was followed by all subsequent authors who characterized adult Scolelepis as having an elongated pointed prostomium (e.g. Fauchald 1977; Blake et al. 2020).

In studying the settling larvae of nine species of Scolelepis, I found that their medioventral peristomial process is not reduced during metamorphosis, but elongates and forms an anteriorly tapering adult rostrum (Figs 3D, E, 7A–E). The posterior parts of the lateral peristomial lips first transform into small semi-oval structures, slightly delimited from the rostrum (Figs 3D, 7D), and then, in connection with the formation of the massive ventral part of the mouth, they elongate and form large lateral parts of the peristomium (Figs 3F, 4F, 5G, 7E). The conical anterior part of the larval prostomium does not elongate forward, but retracts backwards, so that the anterior margin of the adult truncated prostomium becomes rounded to blunt (Fig. 4B) or even concave (Fig. 5D, F). The anterolateral parts of the prostomium extend ventrally, overgrow the basal part of the peristomial rostrum laterally, and finally meet and fuse midventrally. In adults, the line of the junction between the rostrum and the anterior margin of the prostomium looks different in different species. In some of them, the two structures merge into a monolithic, gradually tapering snout, concealing their true origin (Figs 4E, F, 5F, G; S. lightiDelgado-Blas 2006: fig. 6A, B), while in others the prostomium is expanded anteriorly, appearing as a bulge or lateral processes on the basal part of the pointed rostrum (e.g. Scolelepis carunculataBlake and Kudenov 1978: fig. 2a; S. occipitalis Blake and Kudenov 1978: fig. 4a, b; S. blakeiHartmann-Schröder 1980: fig. 97; S. kudenoviHartmann-Schröder 1981: fig. 124; S. branchiaImajima 1992: fig. 2a; S. lingulata Imajima 1992: fig. 4a; S. sagittaria Imajima 1992: fig. 8a, b; S. planata Imajima 1992: fig. 15a; S. vossaeDelgado-Blas 2006: fig. 9A, B; S. alisonaeWilliams 2007: fig. 9A–C; S. magnicornuta Williams 2007: fig. 15A–C; and S. villosivaina Williams 2007: 16A, B, 17A–D). In some species, the epithelia of the peristomial rostrum and the prostomium have different textures, so these two structures are easy to distinguish (Fig. 4B).

The formation and composition of the mouth region of Scolelepis species seems unique among the Spionidae. In most studied spionids, the adult peristomium is formed by the enlarged ventral peristomial lip of the larva, as was described by Hannerz (1956) for Spio martinensis. In Scolelepis, the lateral part of the mouth of adults is derived from the basal parts of the lateral peristomial lips of the larvae, and the strong ventral part of the mouth is formed by the anterior extensions of the first and second segments. The ventral peristomial lip does not develop in either larvae or adults of Scolelepis. Complete reduction and retraction of peristomial structures into the foregut occurs during the metamorphosis of Poecilochaetus larvae. Like Scolelepis, Poecilochaetus does not develop a ventral peristomial lip. The lateral peristomial lips, well developed in larvae (see Hannerz 1956: fig. 49c), are reduced and completely retracted during metamorphosis; therefore, the mouth in adults of Poecilochaetus is shifted to the ventral side of the first segment (Fig. 10E, F).

Morphological characters and phylogenetic relationships of Scolelepis

In addition to the unique head formation, the larvae of at least some Scolelepis species acquire an unusual body shape towards the end of their pelagic life. Competent larvae with posterior swollen segments characteristically tucked under the ventral side of the anterior segments, as in Figures 1C and 2A, were common in the plankton in Paranaguá Bay (Paraná) and São Sebastião Channel (São Paulo) at certain periods. It remains unclear whether these larvae could have settled prior to the development of swollen segments and the development of the modified segments was associated with delayed settling in the absence of an appropriate substrate, or whether it was an advanced development in larvae of adult traits necessary for their successful settlement. It is noteworthy that the larval bristles do not develop in the posterior swollen tucked segments, but adult hooded hooks begin to develop in these segments. The development of modified segments makes the larvae poor swimmers: they begin to move in a circle and sink down when their cilia stop beating. In any case, the appearance of modified segments is probably a sign that the larval period of individual life has ended and it is time to settle down and begin a new adult life. Soon after settling, all segments become uniform, and the body takes on a straight adult shape.

The development of the swollen posterior segments tucked under the ventral side of the anterior part of the body is also found in some other spionid larvae. Large, unidentified pelagic larvae with the same posterior body type as described above for Scolelepis were collected from the São Sebastião Channel in November–December 2003, March–April 2004, and August 2009. These larvae lacked the medioventral peristomial process characteristic for Scolelepis, and had a pair of dorsal pygidial cirri absent in Scolelepis (Fig. 1D). Thus, the unusual body formation towards the end of larval development appears to have evolved either independently in two different branches of the family Spionidae, or only once in a common ancestor of Scolelepis and an as yet unidentified sister-taxon whose larvae were found in the São Sebastião Channel.

The phylogenetic relationships of Scolelepis remain poorly understood. Parsimony analyses of morphological characters of Spionidae and related taxa by Blake and Arnofsky (1999: fig. 13C) and Blake (2006: fig. 13.2B) suggested a close unresolved relationship between Scolelepis, Malacoceros, and MarenzelleriaMesnil 1896. However, no one character was shown to support this relationship. The maximum likelihood analysis of sequences of nuclear 18S and mitochondrial 16S rRNA genes of spionid polychaetes by Abe and Sato-Okoshi (2021: fig. 3) placed Scolelepis in a large group of taxa referred by the authors to the subfamily Nerininae. The analysis of the nuclear 18S and 28S and mitochondrial 16S rRNA gene sequences by Abe and Kan (2022: fig. 5) suggested a sister relationship between the two species of Scolelepis and Dispio remaneiFriedrich 1956, but again no morphological character was found to support the monophyly of this clade.

Relationships between Scolelepis species also remain poorly understood. Maciolek (1987: 17) distinguished two subgenera within the genus based on hooded hook morphology in adults. The subgenus Scolelepis (Scolelepis) was defined by the presence of hooks ‘falcate with 0–2 small apical teeth and straight shaft’; the subgenus Scolelepis (Parascolelepis) was defined by the presence of hooks ‘multidentate with large main fang, several apical teeth and curved shaft’. This taxonomic division was followed by all later authors, although no analysis of the full set of morphological characters of these worms has ever been provided. Blake et al. (2020) assigned 84 species described to that time to one of these two subgenera: 14 species to Scolelepis (Parascolelepis), and 70 species to Scolelepis (Scolelepis). Molecular analyses of Scolelepis reported to date have mainly included species of the subgenus Scolelepis (Scolelepis) (Surugiu et al. 2020, 2022; Lee and Min 2022).

CONCLUSION

The transformation of the lateral peristomial lips of Scolelepis larvae into a pointed rostrum and large lateral parts of the peristomium in adults, as well as the composition of the mouth region in adults of other spionids, as described above, show that Hannerz’s (1956: 157) generalization about the fate of the larval peristomium is correct only for some groups of Spionidae. The present study shows the need for further analysis of the external morphology and internal anatomy of both larvae and adults of Scolelepis in order to provide a corroborated hypothesis about the evolution of these worms and, on its basis, establish a phylogenetic systematization of this large group of Spionidae.

ACKNOWLEDGEMENTS

My sincere eternal gratitude to the late Kristian Fauchald for inviting me to work at the Carrie Bow Cay Field Station (where this research began) of the National Museum of Natural History, Smithsonian Institution in 1997, and for his instant generous support during my visits to the Smithsonian Institution, Washington, DC, USA, from 1995 to 2007. My sincere eternal thanks also to the late Paulo da Cunha Lana for inviting me to work at the Centro de Estudos do Mar, Universidade Federal do Paraná, Pontal do Paraná, Brazil, in 1998 and 2000, and for his instant generous support during my intermittent life in Brazil from then until 2015. Fredrik Pleijel taught me scanning electron microscopy at the Smithsonian, and the basics of phylogenetic analysis since our first meeting in 1994; Álvaro Esteves Migotto provided generous support and assistance with sampling and photography during my work at the Centro de Biologia Marinha, Universidade de São Paulo, São Sebastião, São Paulo, Brazil, during various periods from 2001 to 2015; Bruno Pernet provided valuable comments and editing of the manuscript before the submission. Maarten Christenhusz, Christopher Glasby and Jason D. Williams helped to improve the submitted manuscript. To all these colleagues, I express my sincere gratitude. This study was financially supported by short-term visitor grants from the National Museum of Natural History, Smithsonian Institution, and the Visiting Scientists Program from the Brazilian Ministry of Education.

CONFLICT OF INTEREST

The author declares no conflict of interest.

DATA AVAILABILITY

The data underlying this article are available in Table A1 in the Appendix.

APPENDIX

REFERENCES