Introduction

Besides being notoriously hard to find, low in abundance, and easily overlooked due to their minute sizes, loriciferans are known for their extremely complex life cycles that often make it difficult to link different developmental stages to the same species (Bang-Berthelsen et al., 2013). Some species of the genera Rugiloricus, Pliciloricus, and Titaniloricus have life cycles that allow them to switch between sexual and paedogenetic reproduction with multiple larvae developing inside an asexual stage (Gad, 2005a, c; Heiner, 2008; Kristensen & Brooke, 2002). Others, such as species of Urnaloricus, seem to have lost their adult stages all together (Heiner & Kristensen, 2009). However, despite this complexity, one stage appears to be common for all loriciferans, namely the Higgins larva. Higgins larvae are known from nearly 75% of the described species, but it is generally assumed that this larva occurs at some point in the life cycles of all loriciferans (Bang-Berthelsen et al., 2013; Fujimoto & Kristensen, 2020). When sexual reproduction occurs, it will always be a Higgins larva that hatches from the egg. However, Higgins larvae may also develop asexually from paedogenetic Higgins larvae, ghost larvae, or, in the case of Urnaloricus, from cyst-like mega-larvae.

The Higgins larva is easily distinguished from other stages in the loriciferan life cycle. The typical Higgins larva is composed of a relatively small head, a slightly broader thorax region with transverse rows of more or less quadratic plates, and an even broader abdominal part with plicae or plates. In addition, Higgins larvae always come with several appendages, such as its anterior and posterior setae, and highly characteristic toes (see, e.g. Kristensen & Brooke, 2002; Neves et al., 2016). However, about one decade ago, a Higgins larva with a considerably different morphology was discovered (Neves & Kristensen, 2014). In this larva, the typical body region dimensions had been switched, so the head appeared large and globular, and the trunk was slender, and gradually tapering towards the abdominal region that represented the narrowest part of the larva. Furthermore, rather than having well-arranged thoracic plates, the cuticle of the thorax region has numerous longitudinal folds with a conspicuous zigzag pattern. In order to stress the unique morphology of this larva, it was named “the Shira larva”, and considered a special, and highly aberrant subcategory of Higgins larvae. This first representative of the Shira larvae was described as a new genus and species, Tenuiloricus shirayamai (Neves & Kristensen, 2014), and treated as incertae sedis.

More recently, a second Shira larva was discovered and described as Patuloricus tangaroa (Sørensen et al., 2022), and also this larva turned out to be an orphan, without any known adult stage. It had the same aberrant appearance as T. shirayamai, including the large globular head, slender trunk and thoracic zigzag folds, which suggested a closer relationship between the two species (Sørensen et al., 2022). But the new species did not provide any novel clues about the phylogenetic affinities of the Shira larvae, and without further information about their adult stages, the Shira larvae are still left as orphans with unresolved systematic positions.

However, we might get a hint about their affinities from the unpublished Loricifera literature. In his “Diplomarbeit”, or master’s thesis, Dr Gunnar Gad (Gad, 2000) presented a series of preliminary descriptions of several putatively new species of Loricifera, collected along a huge transect stretching from Punta Arenas in Chile to Bremerhaven in Germany, around the Great Meteor Seamount in the Northeast Atlantic, and from Papua New Guinea. In this work, he drafted the description of a new genus, and assigned no less than six yet undescribed species to this genus. The diagnostic characters for the new proposed genus were based exclusively on Higgins larval morphology, because adults were known from only one of the six species. Interestingly, the larvae of this new genus show a general morphology that is somehow in between the typical Higgins larvae and the aberrant Shira larvae. The head of this new larva is bulbous, but still relatively small, which resembles the condition of the typical Higgins larva, but the trunk is clearly more slender and elongate, which points towards the Shira larvae. Furthermore, the thoracic plates are more irregular, with a tendency towards forming longitudinal zigzag folds, rather than actual plates. However, this work from Dr Gad never got to a stage where it could be considered for publication, and it did not make it through the peer review that was clearly needed to address a number of obvious shortcomings in the drafted descriptions. For instance, all descriptions suffered from a shortage of photo documentation of morphological structures, and there also appears to be some systematic errors in the provided locality information. In addition, Fujimoto et al. (2020) questioned the monophyly of this tentative genus. However, despite these potential flaws, and the fact that this work remained as an unpublished draft, it provides clear indications of the presence of a loriciferan larval morphotype that could link the typical Higgins larvae with the Shira larvae.

The same kind of Higgins larva was also very recently reported from the southern Gulf of Mexico (Neves et al., 2022). In a study of loriciferan diversity in this region, the authors found three Higgins larval specimens with the same peculiar appearance and those reported by Gad (2000). No information about identity or adult stages were available, but their finding stresses that this larval type is relatively widespread, and it is also the first published reporting of this larva.

Recently, numerous sediment samples were collected on the continental shelf and slope in the Santos Basin, Brazil, as part of the Santos Basin Regional Environmental Characterization (PCR-BS), coordinated by PETROBRAS (Moreira et al., 2023). Among the collected meiofauna three specimens of loriciferan Higgins larvae were recorded, and the initial examinations soon revealed that these larvae showed great resemblance to those proposed for the new genus in Gunnar Gad’s thesis. Thus, based on this newly surfaced material, we describe the Higgins larvae as a new loriciferan genus and species. In order to understand the phylogenetic position of the new genus, and as an attempt to link the Shira larvae with other Higgins larvae, we also set up a morphological matrix based on Higgins larval characters, and used it to carry out a phylogenetic analysis of loriciferan relationships.

Materials and methods

Sampling, preparation and examination

Samples were collected by the research vessel R/V Ocean Stalwart through 2019 as part of the SANSED Benthic Surveys (Moreira et al., 2023). Sampling was carried out along eight northwest-to-southeast directed transects (Transect A to H) perpendicular to the continental shelf until the 2.400 m isobath, in the Santos Basin, Brazil, off the coastline between São Sebastião and Rio de Janeiro. Sediment was collected using a GOMEX-type box corer (0.25 m2) at eleven stations in each transect and 12 additional stations at Sao Paulo Plateau. Out of these 100 stations, three yielded the loriciferan larvae for the present study (Table 1; Fig. 1).

Table 1 Summary of data on stations with Scaberiloricus samba gen. et sp. nov., putative larval instars, type status, and catalogue numbers (NHMD-)
Full size table
Fig. 1
figure 1

Map showing the three stations on the continental shelf and slope off Rio de Janeiro marked with flags. Inset shows South America with the sampling region marked

Full size image

Sediment samples were fixed in formalin, and meiofauna was subsequently extracted by LUDOX centrifugation (Vincx, 1996) and sorted to main group. Three specimens appeared to be loriciferan Higgins larvae representing an unknown genus and species, and were isolated for the present study. Before further processing, each specimen was photographed with a Zeiss Axio Imager M2 light microscope equipped with differential interference contrast lenses (DIC), and an AxioCam MRC5 digital video camera. Images and measurements were obtained using the ZEN lite 2.5 2018 image software. Subsequently, one specimen (future holotype, catalogue number NHMD-1184551) was mounted for scanning electron microscopy (SEM), whereas the two additional specimens were mounted for light microscopy (LM). The specimen for SEM was dehydrated through a graded water-alcohol-acetone series, critical point dried, and mounted on an aluminium stub. The mounted specimen was sputter coated with a platinum/palladium mix, and examined with a JEOL JSM-6335F Field Emission scanning electron microscope. Specimens for LM were dehydrated through a graded series of glycerine, mounted in Fluoromount-G between two cover slips, and attached to a plastic H–S slide. The specimens were examined with an Olympus BX51 light microscope with differential interference contrast, and photographed with an Olympus DP27 camera. After examination with conventional light microscopy, the specimens were studied further with a Nikon Eclipse Ti2 confocal laser scanning microscope (CLSM). The autofluorescent signal from the specimens was detected with a laser excitation wavelength of 488 nm, and optical sections were obtained through a 60 × oil objective. CLSM Z-projections were compiled in Fiji ImageJ software v2.0.0. Three-dimensional reconstructions based on the acquired Z-stacks were made with Imaris 9.5.1 software.

Line art was constructed in Adobe Illustrator CS6, and was based exclusively on information from the holotype, obtained from SEM imaging. Same software was used for all other figure plates. All examined specimens are stored in the collection of the Natural History Museum of Denmark (NHMD).

Morphological phylogenetic analysis

In order to understand the phylogenetic relationships of the new Higgins larva, a morphological character matrix was compiled and analysed. Selected taxa for the analysis included the new larva, and all loriciferan species for which early or intermediate Higgins larval instars are known. This included Higgins larvae of 25 additional species, inclusive Armorloricus sp. I in Kristensen and Gad (2004)—the putative Higgins larva of Armorloricus elegans (Kristensen & Gad, 2004) (see also Neves et al., 2021). Species excluded from the analysis even though Higgins larvae were reported are Nanaloricus khaitatus (Todaro & Kristensen, 1998) (morphological information from Higgins larva too limited), Rugiloricus cauliculus (Higgins & Kristensen, 1986) (the only reported Higgins larva is more likely a postlarva), and Pliciloricus leocaudatus (Heiner & Kristensen, 2005) (the only described Higgins larva is a penultimate stage).

For the construction of characters, we attempted to select characters with putative phylogenetic significance. A priori assumptions regarding character state weights and ordered character state transformations were avoided. Likewise, we chose to exclude phylogenetically non-informative characters, e.g., characters where all taxa, or all but one taxon, would have identical coding. Coding is following the morphology of the earliest Higgins larval stage available. Based on this, 24 characters (Supplementary Table S1), representing 56 unweighted and unordered character states, were coded for the 26 selected taxa (Supplementary Table S2). The coding was mainly based on information from the original descriptions, and in a few cases when needed, confirmed by examination of types stored at the Natural History Museum of Denmark. The Nexus file is available through MorphoBank (O’Leary & Kaufman, 2012) as Project 4698.

The data set was analysed with PAUP* version 4.0 (Swofford, 2003) using branch-and-bound search algorithms. The analysis was carried out without an appointed outgroup, since no taxa outside Loricifera possess characters traits that could be homologised or coded in a meaningful and informative way. Thus, the resulting trees are unrooted, and provide only the tree topology, but no direct character polarisation.

Literature sources for the character coding include the following: Kristensen (1983), Higgins and Kristensen (1986), Kristensen and Shirayama (1988), Heiner (2004, 2008), Kristensen and Gad (2004), Gad (2004, 2005a, b, c, 2009), Gad and Martínez Arbizu (2005), Heiner and Kristensen (2005, 2009), Heiner and Neuhaus (2007), Kristensen et al. (2007), Heiner et al. (2009), Pardos and Kristensen (2013), Neves and Kristensen (2014), Fujimoto et al. (2020) and Sørensen et al. (2022).

Results


Phylum Loricifera Kristensen, 1983


Family Pliciloricidae Higgins & Kristensen, 1986


Genus Scaberiloricus gen. nov.

ZooBank registration: urn:lsid:zoobank.org:act:0735EA86-4EBA-4EEC-AD0A-49FEB2D78C5A


Scaberiloricus gen. nov. diagnosis

Higgins larva with small bulbous head, constriction behind head, and elongate trunk, divided into thorax and abdomen; head and trunk width similar in early instars. Anterior part of head with smooth surface; posterior part with numerous longitudinal folds, only interrupted in positions with scalids. Mouth cone with eight, radially arranged oral striae and oral teeth. Introvert with eight scalid rows. Introvert Row 6 with pen nib scalids; Row 7 with bifurcated scalids; Row 8 with alternating trifurcated and kite-shaped scalids. Thorax region accordion-like, with numerous (+40) longitudinal zigzag folds. Abdomen with fewer (ca. 20) more straight plicae, with only one zigzag fold. Thorax region slightly, but not conspicuously, longer than abdomen. Anterior setae include one lateral, and one ventral pair; posterior setae include one dorsal and one lateral pair. Toes long and slender, with distal ¼ forming offset tips; proximal parts of toes supported by toe lamellae on dorsal side.

Adult morphology unknown.

Etymology

The new genus is named Scaberiloricus. The prefix, Scaber-, is Latin for “rough”, and postfix, -loricus, refers to the lorica, thus the name “rough lorica” refers to the numerous zigzag folds in the thorax part of the trunk. The name was originally proposed by Gunnar Gad in his unpublished Master’s Thesis, for a series of loriciferan larvae from the Atlantic Ocean. Thus, the name is also proposed in recognition of Gad’s great effort with a study that unfortunately never got published.


Scaberiloricus samba sp. nov.

(Figs. 2, 3, 4, 5, 6, 7, 8, and 9)

Fig. 2
figure 2

Line art illustrations of early instar Higgins larva of Scaberiloricus samba gen. et sp. nov. A Lateral view. B Dorsal view. C Ventral view. Abbreviations: ab, abdomen; als, anterolateral seta; avs, anteroventral seta; cl, clavoscalid followed by row number; ost, oral striae; pds, posterodorsal seta; pls, posterolateral seta; sc, scalid followed by row number; th, thorax; tl, toe lamella; to, toe

Full size image
Fig. 3
figure 3

Diagram of larval mouth cone and introvert in Scaberiloricus samba gen. et sp. nov., showing distribution of scalids

Full size image
Fig. 4
figure 4

Light micrographs showing focus series from dorsal to ventral through holotypic, early instar Higgins larva of Scaberiloricus samba gen. et sp. nov., NHMD-1184551, before mounting for SEM. Scale in A applies to all images. A Overview, focused on dorsal cuticle. B Overview, focused internally in larva, and on posterolateral setae. C. Overview, focused internally in larva, and on anteroventral setae. D Overview, focused on ventral cuticle. Abbreviations: ab, abdomen; avs, anteroventral seta; cl, clavoscalid followed by row number; pls, posterolateral seta; th, thorax; to, toe. Dashed circles mark the distal parts of the clavoscalids, composed of a lobed unit and a hook

Full size image
Fig. 5
figure 5

Scanning electron micrographs showing head morphology details of holotypic, early instar Higgins larva of Scaberiloricus samba gen. et sp. nov., NHMD-1184551. A Mouth cone, lateral view. B Head, frontal view, dorsal is left. C Detail showing clavoscalids of introvert Row 1. D–I Rotating series showing introvert morphology and scalid arrangements from dorsal to ventral. D Dorsal view. E Subdorsal view. F Laterodorsal view. G Midlateral view. H Sublateral view. I Paraventral view. Abbreviations: cl, clavoscalid followed by row number; mc, mouth cone; md, middorsal line; mv, midventral line; ost, oral striae; ot, oral tooth; sc, scalid followed by row number

Full size image
Fig. 6
figure 6

Scanning electron micrographs showing overviews and details of holotypic, early instar Higgins larva of Scaberiloricus samba gen. et sp. nov., NHMD-1184551. A Dorsal overview. B Left side lateral overview. C Ventral overview. D Lorica, dorsocaudal view. E Lorica, caudal view, dorsal is up. F Posterior part of lorica, ventral view. Abbreviations: ab, abdomen; als, anterolateral seta; avs, anteroventral seta; he, head; mc, mouth cone; pds, posterodorsal seta; pls, posterolateral seta; pvp, posteroventral plica; th, thorax; tl, toe lamella; to, toe

Full size image
Fig. 7
figure 7

Light micrographs showing overviews and details of intermediate instar Higgins larva of Scaberiloricus samba gen. et sp. nov., containing next larval instar, NHMD-1184552. A Overview, focused on dorsal cuticle. B Overview, focused internally in larva, on subsequent larval instar. C Head, dorsal view. D Head, focused internally. E Posterior part of lorica, focused on posterodorsal seta. F Posterior part of lorica, focused on posterolateral seta. G Posterior part of lorica, focused on toes. Abbreviations: ab, abdomen; als, anterolateral seta; cl, clavoscalid followed by row number; cu, cuticle (subscript explainers: Hl, Higgins larva; HI + 1, subsequent larval instar); ia, inner armature; mc, mouth cone; pds, posterodorsal seta; pls, posterolateral seta; sc, scalid followed by row number; th, thorax; tl, toe lamella; to, toe

Full size image
Fig. 8
figure 8

Light micrographs showing overviews and details of late instar Higgins larva of Scaberiloricus samba gen. et sp. nov., containing a postlarva, NHMD-1184553. A Overview, focused on dorsal cuticle. B Overview, focused internally in larva, on subsequent larval instar. C Overview, focused on ventral cuticle. D Mouth cone. E Introvert, ventral view. F Detail of thorax showing anterolateral seta. G Detail of dorsal lorica, showing serially arranged areas of light refractive patches (marked with dashed circles); inset shows close-up of light refractive patch. H Detail of caudal part of abdomen, with posterolateral seta and proximal part of toe. Abbreviations: als, anterolateral seta; cl, clavoscalid followed by row number; cu, cuticle (subscript explainers: Hl, Higgins larva; PO, postlarva); mc, mouth cone; pls, posterolateral seta; pvp, posteroventral plica; sc, scalid followed by row number; to, toe. Dashed circles mark light refractive patches

Full size image
Fig. 9
figure 9

Confocal laser scanning 3D reconstructions of instar Higgins larva of Scaberiloricus samba gen. et sp. nov., containing a postlarva, NHMD-1184553. A Introvert, dorsal view. B Introvert, ventral view. C Overview, dorsal view. D Introvert, ventrolateral view from right angle. E Introvert, ventrolateral view from left angle. Abbreviations: cl, clavoscalid followed by row number; cu, cuticle (subscript explainers: Hl, Higgins larva; PO, postlarva); sc, scalid followed by row number. Dashed circles mark light refractive patches

Full size image

ZooBank registration: urn:lsid:zoobank.org:act:B35A80AE-C260-4EBC-855E-CD2B7BE4AB82


Scaberiloricus samba sp. nov. diagnosis

Higgins larva of Scaberiloricus gen. nov. with introvert equipped with eight rows of scalids: Row 1: eight clavoscalids, composed of long proximal unit, lobed median unit, and curved end-piece. Row 2: sixteen dimorphic scalids, alternating between eight very short spinous ones, and eight longer flattened scalids with distal hooks. Rows 3 to 5: eight, seven, and eight scalids in each row, respectively, composed of triangular bases, flattened median units, and distal hooks. Row 6: seven scalids composed of a plate-like unit extending into a shaft, holding a distal, acicular tip (= pen nib scalids). Row 7: eight bifurcated scalids. Row 8: fifteen dimorphic scalids alternating between seven trifurcated scalids and eight kite-shaped scalids. Anterolateral and anteroventral setae unbranched and flexible, without differentiated bases. Posterodorsal setae unbranched and flexible, attached to large, triangular bases in dorsocaudal part of lorica. Posterolateral setae unbranched and rigid, and with small bulbous bases. Toes slender, and with tapered distal 1/3.

Etymology

The species was collected near Rio de Janeiro in Brazil, and is named samba, because we all associate Rio de Janeiro with good vibrations and samba rhythms.

Material examined

Holotype, early instar Higgins larva, collected from mud on October 30, 2019, at SANSED Spring Shelf Benthic St. G4r1, from 100 m depth in the Santos Basin, on the continental shelf off Rio de Janeiro, Brazil (Table 1; Fig. 1), position: 23°18′32″S, 42°55′59″W, mounted for SEM, and deposited at NHMD under catalogue number NHMD-1184551.

Paratypes include one putative intermediate instar Higgins larva, and one last instar Higgins larva. Paratypic intermediate instar Higgins larva, collected from deep-sea mud on July 10, 2019, at SANSED Winter Deep-Sea Benthic St. F8r2, from 1000 m depth in the Santos Basin, on the continental slope off Rio de Janeiro, Brazil, position: 24°26′22″S, 43°29′30″W, mounted for LM in Fluoromount-G, and deposited at NHMD under catalogue number NHMD-1184552. Paratypic late instar Higgins larva, containing a postlarva, collected from deep-sea mud on August 02, 2019, at SANSED Winter Deep-Sea Benthic St. G6r3, from 700 m depth in the Santos Basin, on the continental shelf off Rio de Janeiro, Brazil, position: 23°48′57″S, 42°40′33″W, mounted for LM in Fluoromount-G, and deposited at NHMD under catalogue number NHMD-1184553.

Description of holotype NHMD-1184551

Holotypic Higgins larva measures 167 µm in length, inclusive mouth cone. The introvert is round and bulbous, measuring 40 µm in length, and 46 µm at its widest point. The trunk is 117 µm in length, slender, and maintain nearly the same width throughout thorax and abdomen: in dorsoventral orientation the trunk width is 36 µm in its anterior and posterior ends, and 52 µm at its widest point, in the posterior part of the thorax. Thorax is longer (63 µm) than the abdomen (54 µm) (Fig. 2A–C).

The mouth cone is short (10 µm) and dome-shaped, and measures 21 µm at its base. LM imaging shows indications of an inner armature, but it could not be observed in detail. Externally, it has eight, radially arranged oral striae alternating with eight oral teeth (Figs. 3, 4B and 5A, B, D–I).

The introvert has eight rows of scalids (Fig. 3). Row 1 consists of eight clavoscalids, measuring 28 µm in length, and composed of three units: a proximal smooth and flattened part (length: 20 µm) representing about 70% of the total scalid length, a short middle piece forming a conspicuous lobe (length: 5 µm), and an even shorter, slightly curved end-piece (length: 3 µm). Two rows of fine, but relatively long hairs follow the frontal margins of the proximal unit. All eight clavoscalids show a uniform morphology, and they attach bilateral symmetrically with four on each side of the introvert, but none in middorsal or midventral positions (Figs. 2A–C, 3, 4A, C, and 5A–I).

Row 2 consists of nine plus seven dimorphic scalids. The dimorphism is expressed in the size, as well as the morphology of the scalids. One type is represented by nine very short scalids (3 µm), composed of a single, rigid, and spinous unit covered with minute hairs. The second type is represented by only seven scalids, that are twice as long (6 µm), and consisting of a projecting, triangular basis, which articulates with a laterally flattened distal unit that terminates into a small hook. The arrangement of the dimorphic scalids is semi-, but not perfectly bilaterally symmetrical. Ventrally, two short scalids are arranged side by side. From this position, and towards the more lateral and dorsal parts of the row, the scalids alternate between shorter and longer ones. This alternation is interrupted subdorsally, where two short scalids appear side-by-side, which breaks the bilateral symmetry pattern (Figs. 2A–C, 3, and 5A, B, D–F).

Rows 3 to 5 are equipped with quite uniform scalids that, except for being slightly longer (lengths: 6–7 µm), resemble the long scalids of Row 2. Rows 3 and 5 have eight scalids that are longitudinally aligned with each other, and with the clavoscalids. Row 4 has seven scalids that are longitudinally aligned with the longer scalids of Row 2, except for the dorsal one, which is perfectly middorsal (Figs. 2A–C, 3, and 5B, D–I).

Row 6 consists of seven scalids with a highly characteristic appearance (length: 7 µm). They are formed by a basal triangular, plate-like unit that narrows distally, and forms a shaft with an elongate opening. A short, distal, straight, acicular tip emerges from the shaft, like a pen nib fitting into a pen holder. The lateral edges of the basal unit, the shaft, and the distal tip are densely populated with minute hairs. In the following comparison, we will refer to this scalid type as a “pen nib scalid” (Figs. 2A–C, 3, and 5D–I).

Row 7 consists of seven scalids (length: 9 µm), each consisting of a single unit with a broad quadratic plate that distally bifurcate into two angular extensions. Most of the quadratic basis, and in particular the inferior surfaces of the bifurcating extensions are covered with minute hairs. In the following comparison, we will refer to this scalid type as a “bifurcated scalid” (Figs. 2A–C, 3, and 5D–I).

Row 8 consists of seven plus eight dimorphic scalids that appear alternatingly around the introvert. Seven scalids (length: 8 µm) resemble the deeply bifurcating ones in Row 7. However, an additional component emerges as a triangular structure that attaches to the introvert in between the lateral extensions. Together, the three extensions give the scalid a trifurcated appearance, and will be referred to as “trifurcated scalids”. Minute hairs are present along the inner and outer margins of the three extensions. The remaining eight scalids (length: 11 µm) are kite-shaped, with short lateral tips, and a long, slender distal tip. This scalid type will be referred to as “kite-shaped scalid”. The posterior part of the bases and the distal tips are covered with hairs that are longer than those on the other scalids (Figs. 2A–C, 3, and 5D–I).

Overall, scalids of the different rows follow some more or less strict longitudinal arrangement patterns (Fig. 3). Thus, clavoscalids are longitudinally aligned with scalids in Rows 3, 5, and 7, as well as the kite-shaped scalids of Row 8. In other words, all scalids that appear in a number of eight are longitudinally aligned. Likewise, the longer scalids of Row 2 (except in middorsal position) are aligned with scalids of Rows 4, 6, and trifurcated scalids of Row 8. Thus, all scalids that appear in a number of seven are also longitudinally aligned. Middorsal scalids appear in Rows 4, 6, and 8. This is the only position where scalids of these rows are not aligned with the long Row 2 scalids. Scalids never appear in midventral position (Fig. 5I).

A distinct neck region or collar is not present, but there is a clear constriction at the transition from head to trunk. The trunk is divided into a thorax and an abdomen region, each defined by the cuticular plication. The thorax region has numerous (+40) zigzagged folds, whereas the abdomen has fewer (ca 20) plicae that only makes a single zigzag twist. The abdomen can be subdivided into three parts: an anterior and a median part, divided by the single transverse zigzag twist in the plicae, and a posteriormost part where plicae on the ventral side fuse into larger posteroventral plicae (Figs. 2A–C, 4A–D, and 6A–C, F).

Anterior setae attach in the posteriormost part of the thorax near the second zigzag twist of the plicae. Both anterolateral (length: 25 µm) and anteroventral (length: 15 µm) setae are thin, delicate, flexible and unbranched. They are slightly swollen proximally, but do not attach via actual bases (Figs. 2A–C, 4C, and 6A–C). Posterior setae include a posterodorsal and posterolateral pair. The posterodorsal setae (length: 45 µm) are also thin, flexible, and unbranched, but attach to two strong, triangular bases with collars around the setae’s attachment points. The bases are located dorsally, on the caudal surface of the larva. The posterolateral setae (length: 9 µm) are more swollen and rigid, but still unbranched. They also attach on the caudal surface, but more lateral, via small, ball-shaped bases (Figs. 2A–C, 4B, and 6A–F).

The toes (total length: 46 µm) are slender and divided into a thicker proximal part (length: 28 µm) and a thinner, distal end-piece (length: 18 µm). The proximal, thicker part of the toe is supported dorsally, by a lamellar structure that proximally is as broad as the toe, but narrows to a pointed tip about 2/3 along the length of the thicker toe part (Fig. 6E). Near the toe tips, these thinner distal parts get slightly thicker, until they again taper abruptly, forming short pointed toe tips (Figs. 2A–C, 4D, and 6A–F).

Paratypes

Two additional larvae of the similar kind as the holotype were collected in the area. They are both larger than the holotype, and one is even containing a postlarva. The two larvae are considered as later, but yet conspecific instars of the new genus and species, and are therefore included as paratypes. It should, however, be stressed that the two larvae could potentially represent different species. Thus, it is urgent to remind ourselves that the only name bearing specimen in any taxonomic act is the holotype (see ICZN Article 73.1).

Description of intermediate instar Higgins larval paratype NHMD-1184552

Higgins larva putatively representing a later larval instar than the holotype, and containing a subsequent larval instar. Total trunk length, inclusive mouth cone, is 178 µm. The introvert measures 44 µm in length, and 59 µm at it widest point. The trunk is 121 µm in length, and appears ovoid in shape, and thus plumper than the trunk of the holotype; in dorsoventral orientation, the trunk width is 76 µm anteriorly, 31 µm posteriorly, and 93 µm at its widest point, in the posterior part of the thorax. Thorax is slightly longer (64 µm) than the abdomen (57 µm) (Fig. 7A, B).

The mouth cone is conical, and measures 17 µm in length, and 22 µm at its base. An inner armature is present (Fig. 7C, D).

The introvert has eight rows of scalids. Row 1 consists of eight clavoscalids, measuring 37 µm in length, and composed of three units: a proximal unit (length: 21 µm), a flattened and only indistinctly lobed mid-piece (length: 8 µm), and an acicular, hooked end-piece (length: 8 µm). The exact number of scalids in Rows 2 to 5 could not be determined, but the scalids appeared similar across all rows, and resembled those of Rows 3 to 5 in the holotype. Row 6 with uncertain number of pen nib scalids. Scalids are present in Rows 7 and 8, though their appearance was obscured by the cuticle of the internal following instar (Fig. 7C, D).

There is no distinct neck constriction between the head and the ovoid trunk, but this could be a result of the trunk cuticle getting looser in preparation for the upcoming moult. The thorax has numerous zigzagged folds, whereas the abdomen has fewer and more straight plicae, of which some are fused onto larger posteroventral plicae as described for the holotype.

Anterolateral setae are simple, flexible, and unbranched, and measure 24 µm in length. Simple anteroventral setae also present, but could not be measured. Posterodorsal setae (length: 45 µm) are slightly thicker than the anterior ones, flexible and unbranched, and attach to two strong, triangular bases. Posterolateral setae (length: 10 µm) are rigid and unbranched, and attach through ball-shaped bases (Fig. 7A, E–G).

The toes (total length: 39 µm) are slender, and divided into a thicker proximal part (length: 25 µm) and a thinner, distal end-piece (length: 14 µm). A dorsal lamella supports the thicker proximal part of each toe (Fig. 7G).

Description of last instar Higgins larval paratype NHMD-1184553

Last instar Higgins larva containing a postlarva. Total trunk length, inclusive mouth cone, is 234 µm. The introvert measures 39 µm in length, and 85 µm at its widest point, at the transition to the trunk. The trunk is bulbous, measuring 182 µm in length, and 142 µm at its widest point. There is no constriction at the transition between head and trunk, and together with the bulbous trunk this gives the outline of the larva a conspicuous avocado-shape (Figs. 8A–C, and 9C). The thorax is slightly shorter (84 µm) than the abdomen (98 µm).

The mouth cone is elongate and conical, and measures 29 µm in length, and 24 µm at its base (Figs. 8D, and 9B). An inner armature is present.

The introvert has eight rows of scalids. Row 1 consists of eight clavoscalids (total length: 55 µm) composed of three units: a proximal unit (length: 33 µm), a flattened and unlobed mid-piece (length: 11 µm), and an acicular, straight end-piece (length: 11 µm) (Figs. 8E, and 9A, B, D, E).

Row 2 consists of seven scalids (length: 43 µm), each composed of two slender, articulating units. The radial arrangement of Row 2 scalids follows the arrangement of the seven longest Row 2 scalids in the holotype; thus, the nine shorter scalids of this row, present in the holotype, appear to be reduced in the late instar.

Rows 3 to 5 consist, respectively, of eight, seven, and eight scalids in each row. The scalids in each row are morphologically similar, consisting of proximal unit with broad, triangular basis and an elongate end-piece, but they gradually decrease in length from Row 3 (length: 35 µm) towards Row 5 (length: 27 µm) (Figs. 8E, and 9A, B, D, E).

Row 6 with seven pen nib scalids (length: 12 µm), and Row 7 with eight bifurcated scalids (length: 10 µm), both with same radial arrangement as in holotype (Figs. 8C, E, and 9A, B, D, E).

Row 8 as in holotype, with seven trifurcated and eight kite-shaped scalids.

Despite the lack of a neck constriction, the thorax region is easily recognised by its ca. twenty zigzagged folds. The abdomen has twenty plicae. The plicae are most well-defined in the anterior part of the abdomen. In the medial part of the abdomen, the plicae are obscured by a transverse zigzag pattern, and posteriorly some are fused into larger posterodorsal and posteroventral plicae.

An interesting cuticular structure occurs repeatedly on the trunk. Patches of light refractive bodies appear transversally arranged, with one patch in each plical fold throughout the lorica (Figs. 8F, G, and 9C). The refractive bodies are either intra- or epicuticular and could be minute pillars inside the cuticle or groups of micropapillae on the surface of the cuticle. They are easily visualised with both LM and CLSM. The thorax has six transverse rows of patches with refractive bodies, which more or less corresponds to the number of transverse zigzag folds. Likewise, the abdomen has three transverse rows, corresponding with the well-organised anterior plicae, the median zigzag folds, and the posterior, enlarged plical fields. Each transverse row has twenty patches, which fits the number of plicae. Thus, these patches appear to be strictly arranged, and following the same pattern as the plicae.

Anterolateral setae are simple, flexible, and unbranched, and measure 33 µm in length (Fig. 8F). Anteroventral setae were not observed, with neither LM nor CLSM. Posterodorsal setae where not observed, and since these setae otherwise have been the strongest and most distinct in the other larval stages, we consider it most likely that they are actually absent in this late larval instar. Posterolateral setae (length: 25 µm) are unbranched, flexible and attach through ball-shaped bases (Fig. 8H).

The toes were unfortunately damaged in the specimen. One toe was broken off at its base, whereas only the proximal part of the other toe was left (Fig. 8B, H).

Very little information could be extracted from the postlarva, which appeared to be at an early developmental stage. The anterior end does not show any indications of developing scalids, and the trunk is filled with undifferentiated, globular structures (Figs. 8B and 9C).

Phylogenetic analysis of Higgins larvae

Branch-and-bound searches in PAUP*4 produced 31 most parsimonious trees, of 41 steps, and Consistency Index 0.7561. A strict consensus tree summarising the obtained topologies is shown in Fig. 10. Selected character transformations are marked on the tree.

Fig. 10
figure 10

Unrooted strict consensus tree based on 31 equally parsimonious trees of 41 steps, obtained after branch-and-bound searches in PAUP*4. Selected character transformations are indicated. Character states are explained, or abbreviated as A (Absent) or P (Present). Asterisk (*) indicates that this character transformation appears in another taxon outside the clade; number sign (#) indicates that the character state has reversed to the original condition for a taxon within the clade; circle (¤) indicates that there are further character transformations happening for taxa within the clade. Wataloricus courtesy to S. Fujimoto

Full size image

The tree is unrooted, due to the lack of outgroup, but we have set an artificial root, based on the assumption that the family Nanaloricidae is an early branching, monophyletic clade. Thus, all proposed character polarisations in the following are likewise based on this assumption. Nanaloricidae is supported by several character transformations, inclusive [C5]: honeycomb sculpturing of the lorica, [C6]: formation of closing plates on the thorax, and [C21]: short, tube-like ventral setae. All remaining taxa are nested in a clade that makes up the sister group to Nanaloricidae. We will refer to this clade as Pliciloricidae sensu lato. It is supported by the presence of [C19]: anterolateral setae. Two additional, potential synapomorphies for either nanaloricid or pliciloricid taxa regards the composition of the lorica ([C4]: plates or plicae), and of the anteroventral setae ([C18]: locomotory or simple setae). However, without a proper outgroup comparison, it is impossible to determine objectively which character states represent the plesiomorphic conditions.

Within Pliciloricidae sensu lato, a clade with the four species of Rugiloricus branch off early, confirming Rugiloricus monophyly. Rugiloricus is supported by [C11]: the presence of a middorsal hook-shaped scalid. All taxa in the clade are furthermore characterised by [C22]: short and stout toes, but again, the plesiomorphic condition of the toe shape remains uncertain.

The remaining taxa form a clade supported by [C13]: more than ten scalids in introvert row 2 (but with uncertain plesiomorphic condition), [C15]: presence of bifurcated scalids in second posteriormost introvert row, and [C22]: abruptly tapered toes (but again with uncertain plesiomorphic condition, and also subsequent modification of the character state). This group splits into two clades, with one of them nesting Urnaloricus gadi (Heiner & Kristensen, 2009) as sister taxon to an unresolved group with Titaniloricus inexpectatovus (Gad, 2005a, b, c) and all taxa of Pliciloricus. This clade is supported by [C7]: presence of a differentiated collar region between the head and thorax, and [C24]: toes articulating with ball-and-socket joint.

The other clade is of particular interest for the present study, since it accommodates the new species and genus described in the present contribution, but also the other two genera with highly aberrant and orphan Higgins larvae. The clade includes Tenuiloricus shirayamai, Patuloricus tangaroa, Wataloricus japonicus (Fujimoto et al., 2020), and Scaberiloricus samba gen. et sp. nov., and is supported by [C3]: the presence of a thorax with zigzag wrinkles, rather than plates, [C12]: clavoscalids with distal units forming lobe with a hook, [C17]: alternating trifurcated and kite-shaped scalids in posteriormost introvert row (but with secondary loss in T. shirayamai). The four taxa branch off in a ladder-like topology, with Scaberiloricus samba gen. et sp. nov. branching out as sister to the others. The remaining three taxa are supported by [C8]: loss of oral teeth in the mouth cone. The next species that branches off is W. japonicus, leaving T. shirayamai, and P. tangaroa as sister taxa. This sister group-relationship is supported by [C1]: head being wider than trunk, and [C14]: Row 2 scalids terminating into small, pincher-shaped claw (Fig. 10).

Discussion

Identity of examined specimens

The collected specimens clearly represent a yet undescribed genus. Especially the dimensions and cuticular morphology of the larval lorica makes them differ from any other described genus. In all Higgins larvae of Nanaloricidae and Pliciloricidae described so far, the thorax region is usually either shorter, or in a few cases same length, as the abdomen (see, e.g. Kristensen, 1983; Higgins & Kristensen, 1986; Kristensen & Gad, 2004; Heiner, 2008; Gad, 2005a, b, c). Furthermore, the lorica of the thorax is typically composed of quadratic or rectangular fields, formed by the longitudinal plicae and transverse folds. Scaberiloricus gen. nov. differs from this common pattern by having a thorax which is longer than the abdomen and characterised by numerous zigzag folds. These zigzag folds in particular differ greatly from the well-organised thorax plates found in most other genera, and this makes it very easy to visually distinguish Scaberiloricus gen. nov.

Only three other genera have larvae that differ from this common pattern with thoracic plates and a longer abdomen. These include the two enigmatic orphan genera Tenuiloricus Neves and Kristensen (2014) and Patuloricus Sørensen et al. (2022), and the recently described pliciloricid genus Wataloricus Fujimoto et al. (2020). In the thoraxes of Tenuiloricus and Patuloricus, we see zigzag folds that are very similar to the folds in Scaberiloricus gen. nov., and both genera are also characterised by having a thorax which is longer than the abdomen (Neves & Kristensen, 2014; Sørensen et al., 2022). However, other dimensions make them differ considerably from Scaberiloricus gen. nov., as both genera are characterised by large and bulbous heads combined with rather slender trunks. On the opposite, the holotypic Higgins larva of Scaberiloricus gen. nov. has more or less similar head and trunk width, and in the putative subsequent instar, the trunk is even wider than the head. Thus, despite the similarities in thoracic morphology, the general habitus of Scaberiloricus gen. nov. differs from the two genera.

The Higgins larva of Wataloricus japonicus Fujimoto et al. (2020) is characterised by a thorax which is longer than the abdomen, and thoracic plates that are more irregular than those in most other genera. However, even though the thoracic plates are less well-defined in this genus, they are not obscured by zigzag folds. Furthermore, the W. japonicus larva has a distinct caudal area, which is not present in Scaberiloricus gen. nov.

Whereas it appears clear that the three Higgins larvae represent a new genus, it is more uncertain whether they represent the same species. The three specimens certainly share several morphological traits—especially if viewed as a developmental series. Similarities include the zigzag pattern in the thorax, which is distinct in the holotype (putative early larval instar) and one paratype (putative intermediate larval instar), but less clear in the paratype representing the last larval instar. But we also observe differences. The general habitus changes gradually from the slender outline of the holotype to the avocado-shaped last instar, and we also observe differences in the scalid morphology in the anteriormost rows. The clavoscalids of the holotype have an intermediate unit forming a small lobe, and a short, distal hook, whereas the paratypes have more elongate intermediate units, and a distal end-piece forming an acicular spine. Likewise, scalid Row 2 of the holotype has 16 scalids that alternate between regular sized ones, and very short ones. On the opposite, the two paratypic larvae have only seven, but much longer scalids. These differences could suggest that the larvae represent different species, but they could just as well be a result of polymorphism between developmental stages in a series of larval instars. It is difficult to choose between the two options, because even though much effort has been invested in understanding the different stages and larval types in the loriciferan life cycle, we actually know surprisingly little about developmental stages of Higgins larvae. In the literature, we have several examples of complete loriciferan life cycles, but in the parts of the cycles that involve the developmental series of Higgins larvae, it seems to be the assumption that the Higgins larval instars simply display a slight change in size, but otherwise maintain the same morphology (see, e.g., Kristensen, 1991, 2002; Gad, 2005c; Heiner, 2008). It is crucial to stress though, that this is nothing but an assumption, since a complete developmental series of Higgins larvae was never studied or documented.

Thus, based on this uncertainty about the morphological variation across Higgins larval instars, we find it more likely that the three larvae, which are collected in the same area, also represent the same species. Hence, for now we propose with some hesitance that the two paratypic larvae represent later Higgins larval instars of Scaberiloricus samba gen. et sp. nov.

Selected morphological traits in the larva instars

During the examination and comparison of the three larval instars, certain morphological traits of particular interest were noted. Some of these traits were unique for the specific instar, whereas others appeared in all instars, and could even be observed in other taxa as well.

Examination of the introvert in the holotype and putatively early instar Higgins larva, revealed a highly characteristic morphology in several scalids. Each clavoscalid is composed of a long proximal unit, a median unit forming a conspicuous lobe, and a short, distal hook. The lobe is not always visualised optimally with SEM, because it depends on the view angle (compare clavoscalids in different angles on Fig. 5C–I), but when observed with LM it is very distinct (Fig. 4A, C). This lobed unit is of particular interest, because its distinct shape makes it easy to recognise across species and genera. But interestingly, it only appears in three other taxa: W. japonicus, T. shirayamai, and P. tangaroa (Fujimoto et al., 2020; Neves & Kristensen, 2014; Sørensen et al., 2022), i.e., the two Shira larvae and the Wataloricus larva that according to Fujimoto et al. (2020) could ‘bridge the gap between Higgins and Shira larvae”. Strangely, these lobes are not present in the clavoscalids of the two later instar paratypes of S. samba gen. et sp. nov. (Figs. 7C and 8E), which brings us back to the question whether they in fact represent different species, or if the scalid morphology goes through a morphological transformation during Higgins larval development. Due to presence of these lobed clavoscalid components in four aberrant kinds of larvae, we would certainly consider this trait to be of phylogenetic significance, but it also stresses our need to obtain a better understanding of morphological changes across the Higgins larval instars.

Also the scalids of the posteriormost introvert rows show a morphology that is easily recognised in Higgins larvae of other genera, and again we find the greatest similarity in Wataloricus, Patuloricus, and Tenuiloricus. In S. samba gen. et sp. nov., the three posteriormost introvert rows have Pen nib scalids (Row 6), bifurcated scalids (Row 7), and alternating trifurcated scalids and kite-shaped scalids (Row 8). This arrangement is identical to the three posteriormost rows in W. japonicus, in terms of scalid numbers as well as morphology (Fujimoto et al., 2020). Also the posteriormost introvert rings in P. tangaroa shows great similarity with this pattern, but its Row 7 differs by having only four bifurcated scalids that alternate with a triangular scalids with minute, bifurcated tips (Sørensen et al., 2022). T. shirayamai differs more, by having simpler scalids in its two posteriormost rows, but its third row, counting from posterior, is also composed of seven pen nib scalids (Neves & Kristensen, 2014).

The detailed and rather well-documented descriptions of Wataloricus and the two Shira larvae facilitate a very easy comparison of scalid morphology across the genera. However, outside these genera the comparison becomes less straight forward. The scalid type from the three posteriormost rows that is most easily recognised in other genera is the bifurcated type. This kind of scalid seems to appear in Higgins larvae of most Pliciloricus and in U. gadi, and is often described as “…eight W-shaped scalids…” in U. gadi (see Heiner & Kristensen, 2009) or as “…eight double, short, triangular projections…” in P. gracilis (Higgins & Kristensen, 1986). Species of Pliciloricus and Urnaloricus also have scalids in the third row, counting from posterior, that show some resemblance with the pen nib scalids, but the similarity is less striking, and we would be more hesitant about interpreting these scalids as belonging to the pen nib-shaped kind.

The trifurcated and kite-shaped scalids of the posteriormost row appear to be restricted to S. samba gen. et sp. nov., W. japonicus, and P. tangaroa. The comparison of scalids across the genera left us with the impression that clavoscalids plus scalids of the three posteriormost rows displayed the most characteristic appearances, suitable for comparison, whereas the more spinous scalids in the rows in between did not provide much useful information.

A final morphological trait that should be highlighted is the numerous series of peculiar cuticular structures observed in the late instar paratype (NHMD-1184553). Throughout the trunk of the specimen we observed patches of light refractive bodies, appearing as groups of papillae or reinforced pores (Figs. 8E, G, and 9C). At first sight, the distribution of these patches appeared like they were just randomly scattered, but a closer examination revealed that they strictly followed the squares formed by plicae and transverse folds in the lorica, leaving a single patch in each square. To our knowledge, such structures have never been previously reported from any described loriciferan specimen, but they are actually illustrated in some of the unpublished descriptions in the thesis of Gad (2000). At this stage, we would refrain ourselves from making any guesses about their function or nature. We would, however, encourage researchers to keep an eye out for these structures in the future, and report them if they turn out to exist in other species or life stages.

Phylogenetic position of Scaberiloricus gen. nov.

Background for the analysis

The relationship between loriciferan species and genera has never previously been addressed in a phylogenetic analysis, thus the present attempt is the first of its kind. The main objective with the present analysis was solely to get a hint about the phylogenetic affinity of the new taxon, and we certainly do not see this as a final and conclusive solution of loriciferan phylogeny. We hope, however, that the results of this analysis can represent a first step towards understanding phylogenetic pathways within Loricifera, and we also think that the result of the analysis, despite its obvious limitations, provides some insights in the positions of certain taxa that previously have been left unclear.

The phylogenetic study comes with two clear limitations: the character matrix is restricted to Higgins larval morphology, and the analysis has been made without an appointed outgroup. The restricted focus on Higgins larva was an active choice, made because the study focuses narrowly on shedding light on the phylogenetic affinities of the new genus and species. By restricting the analysis to taxa with known and well described Higgins larvae, we enabled ourselves to make a more precise character coding and minimise the amount of blank codings for Scaberiloricus gen. nov. and the other orphan larvae. Thus, with this restricted approach in mind, we do not see the exclusion of adult morphology as a major constraint.

The shortage of a suitable outgroup represents a more significant limitation, as it disables the analysis from producing rooted trees and providing a polarisation of the characters states. Identifying a closest relative to the loriciferans is not a challenge as such, since it is generally accepted that Kinorhyncha and/or Priapulida represent their sister group, and the three taxa altogether are united in the clade Scalidophora (Edgecombe et al., 2011; Howard et al., 2022; Worsaae et al., 2023). However, despite the similarities between the three scalidophoran phyla, inclusive the presence of an introvert with scalids, most of the character codings in the morphological dataset would be inapplicable for kinorhynch or priapulid outgroup taxa, and therefore not provide desired character polarisation. We would furthermore be very hesitant about homologising specific loriciferan character traits with those observed in these potential outgroups. For instance, kinorhynchs are direct developers, and we would be cautious about coding larval loriciferan morphology together with juvenile/adult kinorhynch morphology. As for priapulids, most groups have larval stages, and the loricated Priapulus larva, for instance, shows some resemblance with loriciferans (see, e.g., Schmidt-Rhaesa, 2013; Schmidt-Rhaesa & Raeker, 2022). However, proposing a direct homology between the lorical plates in the Priapulus larva and the nanaloricid Higgins larva is highly questionable, and could lead to incorrect character optimisations. Thus, we prefer a cautious approach which requires omission of outgroups, rather than taking a risk by potentially introducing erroneous codings.

As an alternative to outgroup rooting, we found it more justified to go with the a priori assumption that Nanaloricidae—if appearing as a monophyletic group—would be branching early within Loricifera, and not as an ingroup in Pliciloricidae. Whereas monophyly of certain genera and internal relationships within Pliciloricidae always appeared slightly more controversial (see, e.g., Gad, 2004; Fujimoto et al., 2020), the monophyly of Nanaloricidae and sister-group relationship to all remaining loriciferans has never been questioned. Thus, using Nanaloricidae to set the base of the tree appears both safe and logical. Going with this a priori assumption, we propose a consensus tree, rooted at the split between Nanaloricidae and remaining taxa, and interpret our character optimisation based on this rooting (Fig. 10).

Loriciferan phylogeny—the basal splits

Based on the a priori rooting discussed above, we obtained a phylogeny with Loricifera splitting into two basal clades. One clade accommodates all taxa of Nanaloricidae, whereas all remaining terminals, inclusive species of Pliciloricidae, Urnaloricus gadi, and the orphan larvae with unassigned families, are nested in the other clade. The interrelationships between the nanaloricid species remain unresolved, but this is due to scope of the analysis. Obtaining a fully resolved tree for all taxa was never an objective. We note though, that the analysis supports monophyletic Nanaloricidae, which is well aligned with the general ideas about loriciferan relationships (e.g., Bang-Berthelsen et al., 2013; Higgins & Kristensen, 1986).

All remaining terminals are nested in the sister clade of Nanaloricidae. Within this clade, monophyletic Rugiloricus appears as sister clade to the remaining taxa. Consequently, with Urnaloricus and the orphan larvae nested among the remaining taxa, Pliciloricidae appears as paraphyletic. However, rather than breaking this family up, we would for now, prefer to include Urnaloricus and the orphan larvae in this family. Thus, in the following we will refer to the entire sister clade of Nanaloricidae as Pliciloricidae sensu lato.

Pliciloricidae sensu lato

Two synapomorphies support the taxa of Pliciloricidae sensu lato: [C18]: presence of simple anteroventral setae (plesiomorphic state uncertain), and [C19]: presence of anterolateral setae (putative plesiomorphic state: setae absent). However, [C18] is problematic, because the plesiomorphic state of simple anteroventral setae remains unresolved. If the simple pliciloricid setae are homologous with the anteroventral locomotory setae in Nanaloricidae, it is uncertain which of the two kinds of setae represents the ancestral stage. If we consider the setae to be homologous, and polarise the character based on their level of complexity, the nanaloricid seta type would obviously represent the derived state. Thus, under this assumption, the presence of simple anteroventral setae would be synapomorphic for all loriciferan taxa, and not only for Pliciloricidae sensu lato. If we consider the nanaloricid locomotory seta, and the simple pliciloricid setae as convergently evolved, the presence of simple anteroventral setae would be autapomorphic for Pliciloricidae sensu lato. The second character [C19] is more unambiguous, and it makes good sense to consider the presence of anterolateral setae as synapomorphic for taxa of Pliciloricidae sensu lato.

Pliciloricidae sensu lato splits into Rugiloricus and a clade with the remaining taxa. Rugiloricus is supported by [C11] the presence of a middorsal scalid modified into a small hook, and [C22] the presence of short and stout toes.

The clade with the remaining taxa represents the most significant result of the analysis, as it supports a clade with the three orphan larva taxa + Wataloricus as sister group to a clade with Urnaloricus + Titaniloricus + Pliciloricus spp. The clade is supported by [C13] the presence of more than 10 scalids in introvert Row 2 (but with uncertain plesiomorphic condition), [C15] the presence of bifurcated scalids in second posteriormost introvert row (but with secondary loss in Tenuiloricus), and [C22] abruptly tapered toes (but with uncertain plesiomorphic condition, and secondary modification in Urnaloricus and Tenuiloricus). These characters might appear slightly dubious, and we would not consider the two with uncertain plesiomorphic conditions as useful in supporting the clade. However, the presence of bifurcated scalids in one of the posterior rows is well-documented for most taxa within the clade. This particular and visually distinct kind of scalids are very conspicuous in taxa such as Patuloricus, Wataloricus, and Scaberiloricus gen. nov. (Fujimoto et al., 2020; Sørensen et al., 2022; Figs. 2, 3, 5D–I, 6A–C). They are also reported from Urnaloricus as “The 6th row consists of eight W-shaped scalids…” (Heiner & Kristensen, 2009), in Titaniloricus as “..type A scalids (sr5a) consisting of paired plate-like elements..” (Gad, 2005a), and from species of Pliciloricus, e.g. P. gracilis: “The sixth row of head appendages consists of eight double, short, triangular projections” (Higgins & Kristensen, 1986). Throughout the comparative study of introvert appendages, it became clear that different authors observe, interpret, and count introvert rows in different ways. Consequently, in order to homologise scalids of the different rows we needed to re-interpret the row numbering for certain descriptions. This is why we refer to the row with the bifurcated scalids as “second posteriormost”, rather than Row 5 or 6. Comparison of the second posteriormost introvert rows within this clade suggests that having eight bifurcated scalids represent the plesiomorphic condition among the taxa. This is the condition in all terminals of Pliciloricus, Titaniloricus, Urnaloricus, Wataloricus, and Scaberiloricus gen. nov. In Patuloricus the number of bifurcated scalids is reduced to four, and in Tenuiloricus the row is completely lost, or scalids are modified into simpler spinoscalid-like appendages.

The sister clade of the group nesting the orphan larva genera and Wataloricus is composed of Urnaloricus as sister taxon to an unresolved group with Titaniloricus and Pliciloricus spp. It is supported by the presence of [C7] a collar regions between head and thorax, and [C24] toes articulating through ball-and-socket joints. Both characters appear valid, and we consider the clade as well-supported. When described, U. gadi was assigned to its own, monogeneric family, Urnaloricidae, due to its numerous unique traits, inclusive its cyst-like mega-larvae larva, loss of adult stages, and uncommon larval traits, such as its long, spinous toes (Heiner & Kristensen, 2009). However, this unique morphology could just as well be highly derived character traits, and they do not in themselves support an early origin of Urnaloricus. It has already been well-documented that life cycles within Pliciloricidae are extremely flexible, and that new larval stages, as well as paedogenetic cycles, and alternative routes through the life cycle easily evolves (Bang-Berthelsen et al., 2013; Heiner, 2008; Kristensen & Brooke, 2002). In light of this, it does not appear unlikely that a genus within Pliciloricidae sensu lato could lose its adult stages all together. Thus, we do not consider the proposed position of Urnaloricus as improbable.

The polytomous clade with Titaniloricus and Pliciloricus spp. is supported by the presence of [C16] small w-shaped scalids in the posteriormost introvert row, and [C20] terminal setae, and we consider the clade as well-supported.

The position of the orphan larvae

The last, and for the present study most interesting clade, is the one composed of Wataloricus and the three orphan larva genera. The clade appears well-supported by [C3] the transition of regular thoracic plates to numerous zigzag lines, the presence of [C12] a conspicuous lobed unit with a hook in the distal end of the clavoscalids, and [C17] alternating trifurcated and kite-shaped scalids in posteriormost introvert row. An additional, potential character could be [C2] the thorax becoming longer than the abdomen. However, this trait is also present in Urnaloricus, and can therefore not be unambiguously optimised on the tree.

As for the three unambiguous synapomorphies, the zigzag pattern on the thorax is highly conspicuous in the three orphan larva genera, whereas the thorax in Wataloricus is “…densely wrinkled with numerous folds…” (Fujimoto et al., 2020). This thoracic ornamentation in Wataloricus could easily be interpreted as a transitional stage between the well-defined thoracic plates in other Pliciloricidae sensu lato, and the zigzag patterns in the orphan larva genera. Thus, if the topology within the clade was based exclusively on this character, one would expect Wataloricus to branch off earlier than Scaberiloricus gen. nov. In order to reduce the number of a priori assumptions, we chose to treat all multistate characters as unordered. However, we did a trial analysis of a dataset with the states of [C3] ordered as regular thoracic plates <-> irregular thoracic plates <-> zigzag wrinkles, and not surprisingly, this made the internal topology of the clade collapse, and formed a trichotomy consisting of Wataloricus, Scaberiloricus gen. nov., and Tenuiloricus + Patuloricus. Thus, bearing this in mind, we will not completely reject the possibility of finding Wataloricus as the earliest branching taxon within the clade.

The second, and also most consistent autapomorphy for the clade is [C12] the peculiar lobes in the clavoscalids (Fig. 4A, C). This trait might seem like a rather minor morphological detail, but this lobed unit is in fact quite conspicuous when comparing larval clavoscalids across the genera, and the similarity was also noted by Fujimoto et al. (2020). Yet, the character trait is left with some uncertainty, as it seems to change through larval development. Whereas it is very distinct in the putatively early larval instar represented by the holotype of Scaberiloricus gen. nov., it becomes less clear in the late subsequent instar (Fig. 8E). This only stresses the importance of achieving a better understanding of morphological changes through Higgins larval development. In the holotypes of P. tangaroa and T. shirayamai, as well as the paratypic Higgins larva of W. japonicus (catalogue number KUZ Z3666) the lobes are very distinct (Fujimoto et al., 2020; Neves & Kristensen, 2014; Sørensen et al., 2022).

The third autapomorphy [C17] is the alternating kite-shaped and trifurcated scalids in the posteriormost introvert row. As it was the case with the lobed clavoscalids, these scalids are so conspicuous in their appearance, and easily visualised with LM, SEM and CLSM that we believe it would have been reported if this composition existed in any taxon outside the clade. The scalids appear to be secondarily simplified in Tenuiloricus, but since this clade otherwise is very well supported, we interpret this as a derived condition.

Scaberiloricus gen. nov. is sister taxon to a clade composed of Wataloricus, Tenuiloricus, and Patuloricus. This latter clade is supported by [C8] the loss of oral teeth in the mouth cone. Opposite to the characters discussed above, we feel more reluctant about how much emphasis we should put on the phylogenetic signal from this reduction. As the character matrix is coded currently, the analysis certainly finds support for the present topology, but a slight modification of the coding actually changes the topology within the clade. If we, as described above, make the character transformation of thoracic plates to zigzag wrinkles ordered, and knock out [C8] regarding the presence of oral teeth, Wataloricus, and Scaberiloricus gen. nov. swap positions. Thus, we still see this as a quite likely alternative topology.

The two remaining sister taxa, Tenuiloricus and Patuloricus, also represent the two terminals with Shira larvae. Not surprisingly, one of the synapomorphies supporting the sister-group relationship is [C1] the large globular head, which is much wider than the trunk. This is one of the key characters defining the Shira larvae (Neves & Kristensen, 2014; Sørensen et al., 2022), thus, this character transformation makes perfect sense. The second autapomorphy, regards the less conspicuous, but yet unique composition of the distal tips of introvert Row 2 scalids [C14] that form small, pincher-shaped “lobster claws”. We consider both character traits as strong synapomorphies for the two terminals, and consider this sister-group relationship as well-supported. With this topology established, we can conclude that the Shira larva represents a highly derived version of the pliciloricid Higgins larva, and that Wataloricus and Scaberiloricus gen. nov. represent transitional stages from the typical larva towards the Shira larva.

Conclusions

Scaberiloricus samba gen. et sp. nov. represents a new genus and species of Loricifera. Morphologically, it shows greatest resemblance with Wataloricus japonicus, and a phylogenetic analysis of Higgins larval characters suggests that the Higgins larvae of the two genera morphologically represent transitional stages towards the highly derived Shira larva. The phylogenetic analysis supports a clade composed of Scaberiloricus, Wataloricus, and the two Shira larval genera, Tenuiloricus and Patuloricus, and suggests this clade to represents the sister group of a clade with species of Urnaloricus, Titaniloricus, and Pliciloricus. Future research efforts should focus on finding and describing adult stages of Scaberiloricus, but also on improving our understanding of morphological changes through Higgins larval development.