Redescription of two overlooked species of the Perinereis nuntia complex and morphological delimitation of P. nuntia (Savigny in Lamarck, 1818) from the Red Sea (Annelida, Nereididae)

ABSTRACT Two poorly-known species are systematically studied in detail: Nereis (Nereis) latipalpa Schmarda, 1861 from Cape Town, South Africa, and Nereis (Neanthes) larentukana Grube in Peters, 1881 from Larantuka, Flores, Indonesia. Both are raised from synonymy and transferred to the Perinereis nuntia species complex based upon the revision of type material using characters already known, novel, or even forgotten in the literature. Perinereis latipalpa n. comb. is different from its previous senior synonym P. vallata (Grube & Kröyer in Grube, 1858) and related species. The type or topotype specimens of three other nereidid species share the same relevant features as P. latipalpa n. comb.; these are herein regarded as junior synonyms: Neanthes latipalpa Kinberg, 1865, Neanthes latipalpa typica Willey, 1904, and Perinereis namibia Wilson & Glasby, 1993. Perinereis larentukana n. comb. is redescribed and compared to specimens of the morphologically similar species P. nuntia (Savigny in Lamarck, 1818). An improved characterization of P. nuntia using Red Sea specimens similar to the type description was also performed in order to clarify its morphology and restrict the species-complex name content. Other synonymized species of the P. nuntia complex were recognized as distinct: Perinereis brevicirris Grube, 1866, P. quatrefagesi Grube, 1878, and P. weijhouensis Wu, Sun & Yang, 1981. This was based upon features not considered in the previous revisions but pointed out in the original descriptions. An identification key to the 20 currently valid species within the Perinereis nuntia complex is also provided.


INTRODUCTION
Perinereis Kinberg, 1865 is the second most species-rich nereidid genus that includes more than 80 valid species (Read & Fauchald 2018). Although a review (Horst 1889) or regional studies (e.g. Rozbaczylo & Castilla 1973;Hylleberg et al. 1986;Hutchings et al. 1991) have been undertaken, a complete revision of the genus is still imperative to delimit the morphology of all its species, and to obtain an accurate definition of Perinereis since it is polyphyletic (Bakken & Wilson 2005).
Perinereis has been divided into different groupings based on a combination of the number of bars in pharyngeal Area VI and the expansion of dorsal ligule throughout the body (Hutchings et al. 1991). Among them, the group known informally as the Perinereis nuntia species complex Wilson & Glasby 1993) has been traditionally considered as a set of several varieties of P. nuntia (Savigny in Lamarck, 1818) (e.g. Augener 1913;Fauvel 1919Fauvel , 1921Fauvel , 1932Fauvel , 1953. Nevertheless, several of them are currently regarded as valid species or synonymized with others (e.g. Bertrán 1980; Hylleberg et al. 1986;Hutchings et al. 1991;Wilson & Glasby 1993). Wilson & Glasby (1993) revised the accepted species of the P. nuntia complex, and the species have received particular attention ever since Glasby & Hsieh 2006;Park & Kim 2007;Yousefi et al. 2011;Sampértegui et al. 2013).
In the present study, the type materials of these two species were examined. Perinereis latipalpa n. comb. is herein reinstated, whereas three other species are regarded as junior synonyms: Neanthes latipalpa Kinberg, 1865, Neanthes lati palpa typica Willey, 1904, andPerinereis namibia Wilson &Glasby, 1993. Likewise, the poorly known Indonesian species P. larentukana n. comb. is redescribed and transferred to Perinereis after a detailed comparison of the original specimens with some individuals of the similar species P. nuntia (Savigny in Lamarck, 1818). Also, the latter species was characterized based on specimens from the Red Sea to restrict its morphology. An identification key to all species within the P. nuntia complex is provided.

MATERIAL AND METHODS
Type and non-type materials were examined in this study from several collections (see Abbreviations). Measurements of the specimens include total length, length to chaetiger 15, body width at chaetiger 15 excluding parapodia, and the total number of chaetigers. Paragnaths of each single area and both sides of paired areas were counted, as well as the number of teeth on the jaws; the range is also given (minmax). Paired paragnath areas were differentiated and reported as a for left side and b for right side of the specimens. The observation of features on non-everted pharynx required a longitudinal dissection in the mid-ventral oral region. Parapodia were dissected and mounted on glass slides to examine parapodial features.
Light microscopy observations were made using a stereomicroscope and a compound microscope. Specimens were photographed with a digital camera mounted on both the stereo-and compound microscopes, around 10-12 photos were assembled to improve the depth of field in the final figures using Helicon Focus® 6 (method C) or by hand using Adobe Photoshop® CS6. Parapodia were illustrated in anterior views unless otherwise stated.
In the descriptions, the character data for the described specimen are given first, followed by values for the remaining examined material in brackets. The standardized terminology of nereidid parapodial features by Hylleberg et al. (1986) and modified by Bakken & Wilson (2005), the nomenclatorial terminology for jaws by Jansonius & Craig (1971), and for the pharynx and paragnaths by Bakken et al. (2009) were used. Likewise, the term "merged paragnaths" sensu Conde-Vela & Salazar-Vallejo (2015) is applied to paragnaths fused at the base on area AIV, previously regarded as "smooth, parallel bars" sensu Wilson & Glasby (1993) or "smooth bars" sensu Bakken et al. (2009), since a series of individual cones seems to be fused to shape the bars on area AIV of some Nereis (Conde-Vela & Salazar-Vallejo 2015) and Neanthes (pers. obs.), but still further studies on the origin and shape of such paragnaths are needed.
Particular patterns on the shape of ridges and furrows of areas were recognized. Shape of ridges and furrows of AV depends entirely on the shape of those of AVI. For instance, the shape of the ridge and furrows of AV is altered by the shape holding furrows and both ridges of AVI. This areas AVI-V-VI pattern may form a λ-shaped furrow, i.e., the lateral dorsal ridges of the oral ring (AVIa and AVIb) are proximally separated from each other but coalescing submedially and distally forming a mid-dorsal inverted Y-shaped furrow (Area V) . This areas AVI-V-VI pattern may also be π-shaped, where the lateral ridges (AVIa and AVIb) are proximally and distally separated from each other but ridges splits sub-medially in different degrees see below). In descriptions, the above mentioned arrangements of ridges and furrow are referred as "AVI-V-VI". They are diagnostic and used mainly to distinguish closely related species of P. nuntia complex. Areas pattern recognized in the species herein studied and related ones can be summarized as follows: -λ-shaped: ridges of AVI distally and sub-medially coalesced, inverted 'Y' furrow ( Fig. 1A-C) π-shaped: ridges of AVI distally separated from each other ( Fig. 1D-L); subdivided as: χ-shaped: ridges of AVI sub-medially (nearly) coalesced, concave furrows ( Fig. 1D-F) υ-shaped: ridges of AVI sub-medially markedly separated, parallel furrows ( Fig. 1G-I) ɔc-shaped: ridges of AVI sub-medially markedly separated, concave furrows ( Fig. 1J-L) The relative extension of parapodial structures and the difference between the width and thickness of ligules and lobes follow that used in Villalobos-Guerrero & Carrera-Parra (2015). The relative metric length in features concerning to dorsal cirrus and dorsal ligule is only used to measure the entire length of dorsal cirrus with the entire length of the distal and proximal lobes of dorsal ligule, which is not related to the relative length between the dorsal cirrus and the dorsal ligule. The distal one refers to the lobe of dorsal ligule projected distally from the base of dorsal cirri (Fig. 4E, black arrowhead), whereas the proximal refers to the upper edge of dorsal ligule placed between the base of parapodia and that of dorsal cirri (Fig. 4E, white arrowhead).
The length of some structures of the heterogomph falcigers were measured to relate proportions. Following Bakken & Wilson (2005), the length of the blade at the proximal end (a) (Fig. 4K, black arrowheads) was compared with the length of the blade on the distal end (b) (Fig. 4K, white arrowheads). The species herein studied usually have medium heterogomph falcigers, i.e., the length of the distal end of the blade (b) is longer than the length of the basal end of the same (a) but shorter than twice its size (a < b < 2a); and rarely short hetero gomph falcigers (a ≈ b). This proportion is represented in descriptions as b /a = 00-00 times. According to Glasby & Hsieh (2006), the length of the serrated region (Fig. 4L, black arrowheads) was compared with entire length of the blade (Fig. 4L, (1874) developed an accurate analysis of nereidids morphology, and despite the fact that he did not recognize several of the Kinberg's genera, including Neanthes, he suggested that Kinberg's classification of pharyngeal armament was useful for practical purposes. Thus, Grube (1874) proposed several subgenera of Nereis based on the occurrence of papillae and/or paragnaths on a single or both rings of pharynx; among them, N. (Lycoris) that was characterized by having only conical paragnaths on both rings. This subgenus was simultaneously divided into three groups mainly by the number and arrangement of paragnaths on AVI: 1) four or five paragnaths in quadrangle or cross, or more in a circular group (e.g. N. pelagica Linnaeus, 1758 and related species); 2) several paragnaths in a long single arched row, running towards the middle of AV (e.g. N. nuntia species complex); and 3) paragnaths in an oval or rounded patch of transverse rows, or a small linear transverse row of three cones (e.g. N. zonata Malmgren, 1867 and related species). Likewise, Grube (1874) also proposed N. (Perinereis) mainly by having AVI with 1-2 paragnaths transversely stretched. Later, von Marenzeller (1879) described a new species (Nereis mic todonta) related to N. nuntia-like species, and he stated that this group could not be included in the genus Perinereis nor Neanthes by sometimes having either conical and transverse paragnaths on AVI.
Langerhans (1880) recognized the generic level of Perinereis with no further details. Horst (1889), based mainly in Grube (1874), extended the N. (Perinereis) definition by including those species in which the AVI have either only transverse paragnaths, or both conical and transverse paragnaths. He included P. marionii (Audouin & Milne-Edwards, 1833) and P. mictodonta, but N. nuntia was not mentioned. Later, de Saint-Joseph (1898) recognized as valid several of the Kinberg's names, either as genus or subgenus, based only on the pharyngeal characters, excluding all parapodial features. As a consequence, Perinereis was considered as valid only by the presence of conical and transverse paragnaths, whereas Neanthes was regarded as a subgenus of Nereis by having only conical paragnaths; nevertheless, it is noteworthy that N. nuntia was not considered in his proposal. Afterward, Gravier (1902) slightly improved the de Saint-Joseph's classification but recognized that it is still necessary a greater understanding of subgenera delimitation. Gravier (1899Gravier ( , 1902 proposed the inclusion of N. nuntia into N. (Neanthes) by having apparently only conical paragnaths, particularly on AVI; likewise, he supported Horst definition of Perinereis and described a new species (P. heterodonta) with conical and transverse paragnaths on such pharyngeal area. The generic level and the expanded definition of Perinereis prevailed mainly within French annelidologists (de Saint-Joseph 1906;Fauvel 1911Fauvel , 1914Fauvel , 1918Fauvel , 1919Fauvel , 1921Fauvel , 1932, whereas the subgeneric level and the older definition by Grube prevailed within German, Dutch and English annelidologists (Ehlers 1897(Ehlers , 1905(Ehlers , 1920Augener 1913Augener , 1918McIntosh 1910;Horst 1924). Augener (1913) suggested N. nuntia into N. (Perinereis), as for N. vallata Grube & Kröyer in Grube, 1858 and his new subspecies N. (Perinereis) heterodonta var. mictodon toides. Afterward, Fauvel (1919) did not recognize the species closer to P. nuntia (known and valid at the moment), referring them only as varieties of P. nuntia. He regarded it as a polymorphic and widespread species distributed in the Red Sea, Persian Gulf, Indian Ocean, South Africa, Chile, Indo-Pacific, and Japan. Therefore, a general diagnosis of the species complex was provided for the first time, many species were synonymized, and several varieties arose from these assumptions: Perinereis nuntia var. brevicirris, P. nuntia var. djiboutiensis, P. nuntia var. heterodonta, P. nuntia var. vallata, and P. nuntia var. typica. The earlier species variety P. heterodonta var. mictodontoides Augener, 1913 and P. mic todonta were regarded as synonym of the brevicirris type (Fauvel 1919). Later, Fauvel (1921 proposed P. nuntia var. majungaensis from Madagascar, and kept recognizing the six varieties in further works. Perinereis nuntia species complex has been retained in this genus ever since (Fauvel 1932(Fauvel , 1953Day 1967;Paik 1975;Wu et al. 1985;Hutchings et al. 1991;. Wilson & Glasby (1993) developed a revisionary effort of the P. nuntia species complex using type and/or non-type materials of many related species. They provided a diagnosis for the complex, several species were recognized to species level, and others were regarded as synonyms; in total, 12 species were considered as valid within the complex. Later, Glasby & Hsieh (2006) emended the P. nuntia species complex diagnosis, described three new related species and redescribed another two from the East Asia seas.
Based on phylogenetic analysis of nereidins using morphological data, Bakken & Wilson (2005) established Perinereis as a polyphyletic group since their species were present in two distant clades. Some species, including the type species of the genus, P. novaehollandiae Kinberg, 1865, which is a junior synonym of P. amblyodonta (Schmarda, 1861), are nested with two Pseudonereis and three Neanthes species (currently regarded in Pseudonereis, sensu Bakken 2007); whereas the other three Perinereis species (P. variodentata, P. nuntia and P. vallata) were related with Neanthes/Nereis species. Despite this evidence, P. nuntia species complex is retained in Perinereis nowadays (Glasby & Hsieh 2006;Park & Kim 2007;Glasby 2015), although its morphology is notoriously different from the type species. For instance, according to Augener (1922) and my observations on some Australian specimens of P. amblyodonta (ZMB 5274), P. nuntia is mainly distinguished from P. amblyodonta by having dorsal ligule of similar size and shape or slightly uneven throughout the body, whereas in P. amblyodonta it is notoriously expanded in posterior parapodia. Likewise, the dorsal cirrus in P. nuntia is medially placed on dorsal ligule, but it is subterminal in posterior parapodia of P. amblyodonta. Finally, P. nuntia has a single transverse row of several (≥3) shield-shaped bars and/ or conical (sometimes pyramidal) paragnaths on each side of AVI, whereas P. amblyodonta has one large shield-shaped bar on the same area.
This phylogenetic and morphological evidence suggests that the P. nuntia species complex belongs to a different genus, perhaps within the closely related genus Neanthes or Nereis; nevertheless, these two genera also undergo similar taxonomic problems since both are also regarded as polyphyletic (Bakken & Wilson 2005). In order to elucidate the taxonomic problems, it is essential to restrict the definition of the genera by reviewing the species within. It would aid in delimiting their morphology, and consequently, in avoiding speculative new combinations, as could be the case of the P. nuntia species complex.
HAbitAt. -Shallow waters, mud under rocks (Ben-Eliahu 1972.  Head. Prostomium sub-pentagonal, narrow anterior end ( Fig. 2A); 1.3 times wider than long; anterolateral sides 2 times wider than antennal diameter. Palpophores oval ( Fig. 2A), thick, as long as wide, nearly equaling entire length of prostomium; one deep wrinkle placed horizontally in distal third of palpophore. Antennae nearly joined, gap one-third of antennal diameter; digitiform, thickened, extending backward to nearly two-fifths of prostomium. Eyes and lens unpigmented due to long-term preservation, trapezoidal arrangement, anterior and posterior pairs separated (gap subequal to size of posterior; Fig. 2A); lens not visible. Anterior pair of eyes rounded, two-thirds width of antennal diameter. Posterior pair of eyes rounded, similarsized to anterior pair. Nuchal organs deeply embedded, slightly oblique, short, subequal to posterior eyes ( Fig. 2A).
Neuropodia. Neuracicular ligule shorter, 1.5 times wider than ventral ligule throughout body. Superior lobe blunt, shorter than inferior lobe throughout body, projecting beyond end of neuracicular ligule only in few first parapodia. Inferior lobe rounded, thickened, projected beyond end of neuracicular ligule in anterior parapodia, leveling it from median ones. Ventral ligule one-half (one-half or two-thirds) length of median ligule in anterior parapodia, three-quarters length from median ones, except in last parapodia, subequal to it. Ventral cirri cirriform, smaller than ventral ligule throughout body except in first parapodia, as long as it or barely longer.
Villalobos-Guerrero T. F. remArks The species was briefly described but illustrated in detail based on at least two specimens (one with 118 segments) from the Gulf of Suez (Red Sea) (Savigny in Lamarck 1818; Savigny 1822), very likely from shallow waters of the city of Suez, as for others Savigny's polychaetes (e.g. Álvarez-Campos et al. 2015). Up to date, these materials have not been found within the polychaete collection at the MNHN, which currently preserve the types collected by Jules César Savigny during the Napoleonic campaign to Egypt. According to Solís-Weiss et al. (2004), some types remain "hidden" in this collection and the only possibility to find them is to inspect all the tubes in the different jars of the species in question; nevertheless, my focused search of the types of P. nuntia at the MNHN was not successful. Former zoologists who worked on this polychaete collection did not even mention the original P. nuntia material in their texts as they did for other Savigny's species (see Audouin & Milne-Edwards 1833; de Quatrefages 1866; Gravier 1902). It is likely that the type material was missing ever since. I follow Wilson & Glasby (1993) in regarding the specimens as lost. The species was redescribed by Gravier (1902) using specimens from the southern Red Sea, separating it from other similar species, such as P. brevicirris (Grube, 1866) and P. quatrefagesi, based mainly on the length of dorsal and tentacular cirri, and on the number and arrangement of paragnaths on the pharynx. Fauvel (1911) recorded the species from Djibouti, the Persian Gulf, and the Reunion Island. Afterward, Fauvel (1919, 1921 introduced the six varieties within P. nuntia, regarding it as a single polymorphic species, widely distri buted. The relevant features used by Fauvel to distinguish these varieties were the number of paragnaths on AI and AV, the length of the tentacular and dorsal cirri, the shape of ligules, and the form of paragnaths on AVI. In their revisionary effort of the P. nuntia species complex, Wilson & Glasby (1993) mainly used the following features to differentiate the species: the number of paragnaths on the pharyngeal areas, the shape of paragnaths on AIV, AVI, and AVII-VIII, the number of rows on AVII-VIII, and the presence of heterogomph spinigers in anterior chaetigers. Regarding P. nuntia, Wilson & Glasby (1993) redescribed the species using some specimens from the Red Sea collected by Felix Joussaume in 1894 (MNHN A108), but no details on a particular locality were given. According to several records of this and other Red Sea polychaetes collected by Joussaume in the same year (e.g. Gravier 1902Gravier , 1906Hanley 1991;Wehe 2006), it is very likely that the material comes from the Perim Island, located at the Bab el-Mandeb Strait that connects the Red Sea and the Gulf of Aden. This place is about 2000 km of distance from the type locality, both placed in distant marine ecoregions (Spalding et al. 2007). Later, Glasby & Hsieh (2006) expanded the previous redescription of P. nuntia, regarding it as a species with a wide distribution in the tropical Indo-Pacific, northern Australia, Taiwan, the Gulf of Aden, and the Red Sea; however, their specimens should be regarded as a distinct species (see below). Yousefi et al. (2011) stated that their Iranian P. nuntia material fits well with the description of the species by Savigny (1822); however, some relevant differences are evident between them. The Iranian specimens have more paragnaths on AVI (13-20) than Savigny's (10-11; Savigny 1822: pl. 4, fig. 2.3), and the tentacular cirri are notoriously longer (extending backward to chaetigers 8-14) than those illustrated by Savigny (reaching about chaetiger 4). Indeed, the Iranian material has one of the highest number of paragnaths on AVI within the P. nuntia complex (see Wilson & Glasby 1993;Glasby & Hsieh 2006;Park & Kim 2007). Furthermore, Yousefi et al. (2011) noticed some differences between their material and the redescription of P. nuntia (Wilson & Glasby 1993). For example, the Iranian specimens had more paragnaths on the oral ring and more transverse rows on AVII-VIII. Likewise, they also recognized some distinctions in accordance to Glasby & Hsieh's P. nuntia material, such as the presence of heterogomph spinigers throughout body (absent in anterior neuropodia sensu Glasby & Hsieh 2006) and the posterodorsal tentacular cirri reaching chaetigers 8-14 (6-8 in Glasby & Hsieh's specimens).
The P. nuntia specimens herein examined from the Red Sea, which once belonged to A. E. Grube's collection and are now deposited at the ZMB, match in accuracy with the Savigny's original description and illustrations of P. nuntia. For instance, the sub-pentagonal prostomium with narrowing end; the χ-shaped pattern of AVI-V-VI; the triangular arrangement of cones on AV; the number of paragnaths on AI (1-3 in Grube's vs 1 in Savigny's), AII (9-10 in Grube's vs 6-10 in Savigny's), AIII (11-19 in Grube's vs 14 in Savigny's), AIV (14-19 in Grube's vs 14 in Savigny's), AV (2-4 in Grube's vs 3 in Savigny's), and AVI (8-10 in Grube's vs 10-11 in Savigny's); the 4-5 transverse rows of paragnaths on AVII-VIII (2-3 most proximal rows with smaller cones). Also, the slender dorsal cirri which show similar size variations, they are much longer than the dorsal ligule in posterior parapodia (3-4 times in Grube's vs 4-5 times in Savigny's); the absence of merged paragnaths on AIV; and even the unusually short tentacular cirri. These are some of the most relevant features to distinguish P. nuntia sensu stricto from the species within the complex, as well as the two canals in the jaws, and the presence of heterogomph spinigers in all the neuropodial subacicular fascicles. Therefore, these specimens are also different from those previously referred as P. nuntia by Gravier (1902) and Wilson & Glasby (1993)  On the other hand, collecting data are scarce in the catalog and labels for the Grube's specimens of P. nuntia. Most of the material from the Red Sea revised by Grube was collected by Christian Gottfried Ehrenberg and Georg Ritter von Frauenfeld, and many of them resulted in previously undescribed taxa (Grube 1868, 1869; see also Wiktor 1980 andHartwich 1993). The marine invertebrates collected by Ehrenberg and von Frauenfeld were mainly obtained from the sediments or coral reefs of the Northern and Central Red Sea (Ehrenberg Villalobos-Guerrero T. F. 1828; von Frauenfeld 1856). However, despite the similar morphology to those specimens originally described from the Northern Red Sea (Savigny in Lamarck 1818), there is no certainty of the precise Red Sea location nor the collector. Therefore, some of the required conditions to establish a neotype could not be met (ICZN 1999, Art. 75.3).  diAgnosis. -Specimens with antennae separated, nuchal organs subequal to posterior eyes, postero-dorsal tentacular cirri reaching chaetigers 3-6. Jaws with 5-6 canals. Maxillary ring: AI = 1-2; AII = 2-10; AIII = 8-20, 1-2 lateral cones; AIV = 14-28, merged paragnaths absent. Oral ring: AVI-V-VI pattern, λ-shaped; AV = 0-2, paragnaths slightly behind AVI; AVI = 9-12, arc long, oblique, bars short, even; AVII-VIII = 48-66, 2-3 transverse rows, proximal with large cones. Gap between AVI and AVII-VIII narrow. Dorsal cirri 1.5-2 times longer than dorsal ligule, inserted posteriorly on three-quarters of it. Dorsal ligule barely uneven throughout body, subequal to median ligule; distal lobe bluntly conical posteriorly. Ventral cirri cirriform. Homogomph spinigers with proximal teeth notoriously thickened, separated; absent in subacicular neurochaeta. Heterogomph spinigers with proximal teeth barely thickened, evenly spaced; present throughout. Heterogomph falcigers with medium blades.
Neuropodia. Neuracicular ligule slightly longer than ventral ligule (Fig. 7D). Superior lobe digitiform, projecting beyond end of neuracicular ligule in first 26 parapodia, reduced from parapodia 56. Ventral cirri smaller than ventral ligule.   N. (Nereis) latipalpa using specimens collected by himself in Table Bay during his visit to Cape Town (South Africa) in February-June 1854 (see details in Appendix). The type material of N. (Nereis) latipalpa is still deposited at the NHMW. It consists of nine syntypes, which match the original description (Schmarda 1861); the best specimen preserved is here designated as lectotype (NMHW 769a), whereas the others as paralectotypes (NMHW 769b,770,771). The size and number of chaetigers of the lectotype do not fit the specimen described by Schmarda; nevertheless, it was the only specimen dissected among the type material, agreeing with dissected parts mentioned and illustrated by Schmarda: one jaw, and one anterior (chaet. 13) and posterior (chaet. 78) parapodia, which are lost.
De Quatrefages (1866) recognized N. latipalpa as a valid species. Kinberg (1865), based on a specimen collected in Cape Town and on the description of N. latipalpa by Schmarda, described a new species, Neanthes latipalpa, but he was not entirely sure whether it was the same species as Schmarda's since the latter did not consider the paragnaths arrangement, a relevant feature used by Kinberg to distinguish nereidid genera and species.
Von Marenzeller (1888), after an examination of some polychaetes from Angra Pequena Bay (now as Lüderitz Bay), noticed the morphological similarities between the Nami bian and Cape of Good Hope specimens. Further, he realized that N. (Nereis) latipalpa by Schmarda and Neanthes latipalpa by Kinberg are the same species because of the shape of the parapodia. Thus, he synonymized N. latipalpa Kinberg but transferred N. (Nereis) latipalpa Schmarda to Neanthes Kinberg, 1865, the latter following the Kinberg's classification of genera by using the pharyngeal arrangement of paragnaths.
Ehlers (1901), who certainly did not recognize as valid Ne anthes and other Kinberg's nereidid genera (see Ehlers 1868), examined the type material and some specimens of N. (Nereis) latipalpa Schmarda from Angra Pequena, regarding the species as a junior synonym of Nereis vallata (currently in Perinereis sensu Hartman 1959). Later, Willey (1904) examined the type specimen of Neanthes latipalpa and confirmed the synonymy proposed by von Marenzeller (1888), but established a trinominal name, N. latipalpa typica, to emphasize that it is barely different to N. latipalpa brevicirris (currently as Perinereis brevicirris sensu Hartman 1959). Afterward, Augener (1918) reinforced Ehlers' assumptions and considered N. latipalpa typica as identical to N. vallata, but placed in the subgenus Perinereis (sensu Augener 1913) by having, among other characters, a continuous transverse line of paragnaths on AVI.
Fauvel (1919,1921,1932,1953) grouped the species morphologically similar to Perinereis nuntia, mainly characterized by having an arc of paragnaths on AVI. Within the subspecies of this group, P. nuntia var. vallata was different from the others by having AI = 1-3, AV = 1, most extended tentacular cirri reaching chaetigers 7-8, and dorsal cirri subequal to dorsal ligule or slightly smaller. Fauvel considered the species as having a wide distribution (the Red Sea, India, Madagascar, South Africa, Philippine Islands, New Zealand, Australia, and Chile) based on the records of the species under several names. Also, Fauvel (1919, 1921, 1932, 1953 still retained Schmarda, Kinberg and Willey's N. latipalpa as a synonym of P. nuntia var. vallata; and since then, the three latipalpa's names have not been recognized (e.g. Wesenberg-Lund 1962), except by Hartman (1949,1959), who still retained the Kinberg's species as valid in Neanthes, but details were not provided. Hutchings et al. (1991) established three groups within Perinereis according to the number of bars on AVI; those valid species more related to P. nuntia were placed in the third group (AVI with numerous bars in an arc). A few years later, Wilson & Glasby (1993) carried out a relevant study of 20 species within the P. nuntia species complex. They restricted the morphology of the P. nuntia complex, redescribed some of the species using type, topotype or non-type materials, such as P. nuntia and P. vallata; also, they recognized as valid some previously synonymized species, and described two new ones: Perinereis akuna (New South Wales, Australia) and P. namibia (Lüderitz Bay, Namibia). Nevertheless, the authors overlooked N. latipalpa since it was not included in the source of synonyms of any species, particularly P. vallata, neither in the appendix at the end of the document.
After the comprehensive revision of the type material of N. Nereis) latipalpa Schmarda, Neanthes latipalpa Kinberg and Neanthes latipalpa typica Willey, herein is confirmed that they belong to the same species. Furthermore, this species is here considered as valid since it is different the other species within the P. nuntia complex, particularly P. vallata. Therefore, it is reinstated as P. latipalpa (Schmarda, 1861) n. comb. Besides, P. namibia is regarded as a junior synonym of P. latipalpa n. comb., based on the original description and the revision of some topotypes formerly used by Wilson & Glasby (1993).
Perinereis latipalpa n. comb. and P. vallata are indeed closely related species since they have a similar parapodial pattern, as well as a similar number and arrangement of paragnaths on the ventral areas of the pharynx. Nevertheless, according to my revision on the type material of P. latipalpa n. comb. and a specimen from Chile (Fig. 7F-K), the description of P. val lata and its accurate posterior characterizations (Ehlers 1901;Wesenberg-Lund 1962;Rozbaczylo & Castilla 1973;Bertrán 1980;Wilson & Glasby 1993;Sampértegui et al. 2013), the differences between these two species remain in the shape of the proximal teeth of homogomph spinigers and the AVI-V-VI pattern. In P. latipalpa n. comb., the first proximal teeth of homogomph spinigers are notoriously thick and separated (Figs 4G, H; 5J; 6K; 7E), whereas in P. vallata these are of similar size and equidistant from one another (Fig. 7F). Also, in P. latipalpa n. comb., the AVI-V-VI pattern is λ-shaped (Figs 3D, F; 5B; 6E; 7B), with the AVI overlapping AV proximally further behind the arc of paragnaths (Figs 3F; 5B; 7B); whereas in P. vallata, it is χ-shaped (Fig. 7G, I), with the AVI overlapping AV proximally at the level of the arc of paragnaths (Fig. 7I).

ZOOSYSTEMA • 2019 • 41 (24)
Perinereis latipalpa n. comb. also differs from other two closely related species of P. vallata: Perinereis gualpensis Jeldes, 1963 (based on the original description and characterizations of topotype specimens, e.g. Bertrán 1980;Sampértegui et al. 2013) and P. akuna Wilson & Glasby, 1993. Perinereis gual pensis, who also shares the same type of blade of homogomph spiniger and the same pattern of AVI-V-VI as P. vallata, can be distinguished from P. latipalpa n. comb. because the latter has AIII = 8-20 and AVII-VIII = 48-66, whereas P. gualpensis has AIII = 21-55 and AVII-VIII = 27-47. Likewise, the dorsal ligule is longer than median ligule and slightly expanded in posterior parapodia of P. latipalpa n. comb.; whereas in P. gualpensis, it is subequal to median ligule and not expanded in same parapodia. Furthermore, in P. latipalpa n. comb., the dorsal cirrus extends beyond the end of dorsal ligule, and it is placed on the distal quarter of the ligule, whereas in P. gual pensis the dorsal cirrus is subequal to dorsal ligule and it is located on one-half of the ligule.
On the other hand, although P. latipalpa n. comb. and P. akuna share the same λ-shaped pattern in AIV-V-IV, they differ in several respects. Perinereis latipalpa n. comb. has the first proximal teeth of homogomph spinigers notoriously thick and separated, whereas in P. akuna these are of similar size and equidistant from one another. The dorsal ligule is longer than median ligule and slightly expanded in posterior parapodia of P. latipalpa n. comb., whereas in P. akuna the dorsal ligule is subequal to median ligule and not expanded in same parapodia. Also, in P. latipalpa n. comb. the dorsal cirri extend beyond the end of dorsal ligule and it is placed on the distal quarter of the ligule; whereas in P. akuna, the dorsal cirri is subequal to dorsal ligule and it is located on one-half of the ligule. Likewise, P. latipalpa n. comb. has a lower number of paragnaths on AVII-VIII (48-66) which are arranged in 2-3 rows of small and large cones; whereas P. akuna has a larger number of paragnaths  which are arranged in more than four rows, with tiny cones in a distinct band. Finally, P. latipalpa n. comb. always lacks merged paragnaths on AIV, which are always present in P. akuna (numbering 1-5).
Wilson & Glasby (1993) regarded the occurrence of merged paragnaths on AIV as a relevant taxonomic feature to distinguish species within the P. nuntia complex; however, this character may be present or absent in individuals of the same species (Hutchings et al. 1991;Wilson & Glasby 1993;Park & Kim 2007;Sampértegui et al. 2013). Concerning P. vallata, Sampértegui et al. (2013 considered this character as varying since it was only present in a few of their Chilean specimens. The specimens herein examined of P. latipalpa n. comb. lack this type of paragnath, but a small merged one was present in the examined individual P. vallata. It is uncertain if their presence/absence on AIV of the latter species is related to size or to a particular developmental stage. Nevertheless, it is noteworthy that any of the specimens of P. latipalpa n. comb. herein examined and many of those examined by Wilson & Glasby (1993) as P. namibia, have merged paragnaths. For this reason, this feature should be carefully used to distinguish species within the P. nuntia complex. type mAteriAl. -7 syntypes of Nereis (Neanthes) larentukana: 1 specimen, MPW 411, labelled as "holotypus lub syntypus", in poor condition; 4 specimens, ZMB 863, labeled as "Syntypen", in good conditions; two specimens (ZMB 4002-Q), labeled as "Syntypen", in poor conditions. All collected by E. K. von Martens, from Larantuka, Flores, East Nusa Tenggara, Indonesia, no further data but see remarks and Appendix.
remArks Grube in Peters (1881) described Nereis (Neanthes) larentukana using specimens from Larantuka (Indonesia) and collected by von Martens in January 1863 (see details in Appendix).
The type material of N. (Neanthes) larentukana is currently lodged at the Museum of Natural History of the Wroclaw University (MPW) and at the ZMB (Wiktor 1980;Hartwich 1993;Glasby & al Hakim 2017). It consists of seven syntypes that match accurately the original description that was based apparently in at least two specimens.
Grube (1874) proposed three groups within the subgenus Nereis (Lycoris); among them, a group lumping the N. nuntia related species, characterized by having several paragnaths in an arched row running towards the center of AV. Later, Grube in Peters (1881) considered the above concerning species in N. (Neanthes), which is also a member of the P. nuntia com- Villalobos-Guerrero T. F. plex. It is still uncertain why Grube resumed using Neanthes as subgenus; however, it is probably that based on his 1874 work, he realized lately that N. nuntia group might represent other distinct but valid subgenus with Neanthes as representative. Perhaps, he followed the second group of Neanthes proposed by Kinberg (1865) (1901), who examined the type specimen of N. variegata, and my revision on a topotype previously identified by Grube (ZMB 3671), the species has one large cone on AV and one single large compressed paragnath in each side of AVI. Intriguingly, it is noteworthy that Ehlers (1901) found, although very few, topotypes having three cones in a triangle or transverse row on AV. The type material of N. (Neanthes) larentukana has not been examined since its description. Furthermore, the species has been scarcely addressed in the literature and commonly listed as Neanthes Fauchald 1972;Wilson 1984;Salazar-Vallejo et al. 2014;Glasby & al Hakim 2017); however, no details on its validity have been provided. After the examination of the type specimens, it is unveiled that N. (Neanthes) larentukana is a member of the P. nuntia species complex, mainly by having conical paragnaths on both pharyngeal rings, an arc of several shield-shaped paragnaths on AVI, notochaeta with only homogomph spinigers, and dorsal and median ligules slightly uneven throughout the body (Wilson & Glasby 1993; Glasby & Hsieh 2006; this study). Therefore, it is regarded as a Perinereis species, P. larentukana n. comb.
Among the species in the P. nuntia complex, P. larentukana n. comb. is similar to those species with dorsal cirri distinctly shorter than dorsal ligule. However, it is more closely related to P. heterodonta Gravier, 1899 (Djibouti) and P. quatrefagesi (Grube, 1878) (Bohol, Philippines), both previously referred as junior synonyms of P. nuntia (Wilson & Glasby 1993). Yousefi et al. (2011) recently reinstated the former species from the synonymy based on some specimens from near to the Gulf of Oman. They considered P. heterodonta as different from P. nuntia by lacking heterogomph spinigers in anterior neuropodia, by having shorter tentacular and dorsal cirri, AV = 0-1, and AVII-VIII with fewer paragnaths (further differences, see remarks of P. nuntia). However, the number of paragnaths on AIII, AIV and AVII-VIII from Iranian specimens also slightly differ from those from Djibouti. Due to these differences, the information of the original description and illustrations of P. heterodonta (Gravier 1899(Gravier , 1902 and the posterior characterizations with specimens from the type locality (Gravier 1902;Fauvel 1919) were used to show that it differs from P. larentukana n. comb.
Regarding P. quatrefagesi, Grube (1878) briefly described the species without illustrations and based on only one poorly preserved specimen from the Philippines; however, some relevant features were provided. For instance, some of the anterior end: antennae close enough to each other, shorter than one-half of prostomium; dorsal-most tentacular reaching chaetiger 5; AI with 3 paragnaths in a longitudinal row, AII with 7 in a circular patch, AIII with a transversely oval patch and 2 isolated cones, AIV with 18 paragnaths in 2-3 rows, AV with 1 cone, AVI with 8 bars in a transverse row, and AVII-VIII with 19 in 2 rows. Likewise, some features of parapodia; short, even throughout the body; the ligules are blunt, slightly tapered, with the dorsal ligule slightly broader, longer than median one, and ventral ligule small, rounded; dorsal cirrus short, almost one-half length of dorsal ligule; and among others. Fauvel (1919) synonymized P. quatrefagesi transferring it into his group P. nuntia var. vallata. Later, Augener (1933) recognized the species as valid, but using specimens from the northern Moluccas and the description of Gnatholycastis brocki Ehlers, 1920 (Ambon Island), which was regarded as a junior synonym of P. quatrefagesi ever since (e.g. Hartman 1959;Salazar-Vallejo et al. 2014). However, G. brocki is characterized by having a prostomium with a slightly incised distal end, a papilliform dorsal cirrus located distally on an expanded dorsal ligule in posterior parapodia, and an arc of paragnaths on AVI (Ehlers 1920). This description, in fact, matches P. caeruleis (Hoagland, 1921) (Sulawesi) rather than P. quatrefagesi, a species also recognized within the P. nuntia complex (Wilson & Glasby 1993). Although the revision of the type material is needed, it is possibly that G. brocki and P. caeruleis are the same. If this is the case, the former name has priority over the latter (ICZN 1999, Art. 23.1), which has been erroneously dated as 1920 because of the cover page of the paper, but the volume of Hoagland's work was in fact formally published until March 1921(True & Oesher 1947. Following the above, P. quatrefagesi was inaccurately regarded as valid. It is noteworthy that Glasby & Hsieh (2006) considered P. caeruleis as not belonging to the P. nuntia complex, such as other two species (P. akuna and P. rhombodonta Wu, Sun & Yang, 1981), by the presence of pyramidal paragnaths on AVI, and a band of several rows of small cones separated from larger ones on AVII-VIII.
Perinereis quatrefagesi remained somewhat unnoticed in the literature until Hylleberg et al. (1986) examined in detail several atoke and epitoke specimens from the Gulf of Thailand. Their specimens fitted accurately the original description. They were different to P. brevicirris and P. nuntia, but Nereis (Perinereis) rumphii Horst, 1919 (Banda Sea) and P. weijhouensis Wu, Sun & Yang, 1981 (Guangxi, China) were referred as synonyms of P. quatrefagesi. Nevertheless, they may be regarded as independent species according to the original descriptions and Hylleberg et al. (1986) characterization.
Nereis (P.) rumphii differs from P. quatrefagesi in the shape of median parapodia in the epitoke female. The former species has a median ligule with an expanded lower-basal lamella and a ventral ligule with a sharp upper-basal projection, which are lacking in both ligules of P. quatrefagesi. Horst's description of the atoke N. (P.) rumphii fits that provided above for P. laren tukana n. comb., although the syntypes of the former species should be examined before regarding it as a junior synonym. Likewise, P. larentukana n. comb. differs from P. quatrefagesi because the former lack heterogomph spinigers in the first few anterior chaetigers, whereas this kind of chaeta is present throughout the body in P. quatrefagesi. Also, P. weijhouensis is different from P. quatrefagesi because the dorsal cirrus is twice longer than dorsal ligule in posterior parapodia, the homogomph spinigers are in both neuropodial fascicles, and lacks heterogomph spinigers. Whereas in P. quat refagesi, the dorsal cirrus is shorter than or nearly subequal to the ligule in same parapodia, the homogomph spinigers are present only in the neuropodial supracicular fascicle, and the heterogomph spinigers in the neuropodia. Perinereis weijhouensis was also synonymized with P. nuntia by Wilson & Glasby (1993), but it may also be raised from this synonymy taking in account the distribution pattern of spinigers mentioned above (P. nuntia similar to that in P. quatrefagesi); also, by having the arc of paragnaths on AVI clearly separated from each other (rows adjoining mid-dorsally in P. nuntia).
Wilson & Glasby (1993) examined the holotype of P. quatre fagesi regarding it as a junior synonym of P. nuntia. No further details on its morphology were added to those mentioned in the original description, except that only a few chaetae are intact and the body is almost completely digested, remaining only the pale empty skin. Nevertheless, P. quatrefagesi herein is regarded as valid; it differs from P. nuntia by the length of dorsal cirrus, the mutual separation of arcs on AVI, and the number and arrangement of paragnaths on AVII-VIII. Perinereis quatrefagesi has dorsal cirrus shorter than or nearly subequal to the dorsal ligule in posterior parapodia, whereas in P. nuntia it is much longer (2-3 times) than ligule in same parapodia. In P. quatrefagesi, the arcs of paragnaths on AVI are clearly separated from each other by the broad ridge of AV; whereas in P. nuntia, the arcs of paragnaths on AVI nearly adjoin mid-dorsally by the narrow ridge of AV. Furthermore, the AVII-VIII in P. quatrefagesi has 19-36 paragnaths in up to 3 rows, whereas P. nuntia has 36-50 paragnaths in 4-5 rows.
Finally, Glasby & Hsieh (2006) extended the description of P. nuntia using material from several localities of the Indo-Pacific, illustrating a specimen from Singapore. However, it is evident that at least the Singaporean material belongs to a species closer to P. larentukana n. comb. and P. quatrefagesi, rather than P. nuntia. Some features of the anterior region looks alike to P. larentukana n. comb./P. quatrefagesi (e.g. prostomium, antennae, dorsum of pharynx, among others). However, the slender dorsal cirrus, subequal to the dorsal ligule in posterior parapodia, and the absence of heterogomph spinigers in many anterior chaetigers, suggest that it is a different (possibly undescribed) species. This key includes all species regarded as members of the P. nuntia complex (Wilson & Glasby 1993;Glasby & Hsieh 2006), and other species previously regarded as not valid or even excluded. Nevertheless, some species deserve considerations since the broadly different morphological features may suggest that some species belong to a distinct group, but the phylogenetic analysis of the complex would be needed to support this idea. Glasby & Hsieh (2006) already suggested P. akuna, P. caeruleis and P. rhombodonta as possibly not members of the P. nuntia complex by having pyramidal paragnaths on AVI and several rows of small cones separated from larger ones on AVII-VIII. However, P. caeruleis is strikingly different within the P. nuntia complex by also having a slight distal incision on the prostomium, posterior parapodia with papilliform dorsal cirrus inserted distally on an expanded dorsal ligule, short heterogomph falcigers with a much reduced serrated region throughout the body, and a large number of minute paragnaths on AIII to AVII-VIII. Also, P. maindroni Fauvel, 1943 has been considered as a member of the P. nuntia complex by having AVI with barshaped paragnaths in a transverse row. However, it is the only species within the complex bearing a notoriously elongated digitiform dorsal ligule in median parapodia, well-developed notopodial prechaetal and neuropodial postchaetal lobes, and a dorsal cirrus basally attached in all parapodia (Fauvel 1943

DISCUSSION
The studies on the P. nuntia species complex held earlier by Fauvel (Fauvel 1919(Fauvel , 1921(Fauvel , 1932(Fauvel , 1953, and more recently by Wilson & Glasby (1993) and Glasby & Hsieh (2006), onset the morphological delimitations among the species in the complex. They pointed out several relevant features to distinguish closely related species, such as the occurrence of merged paragnaths on AIV, the presence of an additional band of minute paragnaths on AVII-VIII, the absence of heterogomph spinigers in anterior chaetigers, the proximity of paragnaths on AV to those on the AVI, among others. Also, in this study, some novel and additional features formerly overlooked in these works but previously used in other nereidid studies to distinguish species, were also considered. For instance, the AVI-V-VI pattern, the width of the gap between AVI and AVII-VIII, the presence of p-bars on AVII-VIII, the presence of homogomph spinigers in both neuropodial fascicles, the size and shape of proximal teeth in homogomph spinigers, the distance between antennae, the number of canals in the jaws, and the length of the nuchal organs. All these features together were used to compare the morphology of P. latipalpa n. comb., P. larentukana n. comb. and P. nuntia sensu stricto with the currently known P. nuntia-like species. As a consequence, the former two species were reinstated based on the type material, and the latter was restricted using some Red Sea specimens. Nevertheless, it is noteworthy that some other species formerly regarded as junior synonyms within the P. nuntia complex may also be treated as valid when comparing the original description, the re-descriptions, and/or the characterizations of topotype materials using the features above mentioned. That is the case for P. weijhouensis, P. quatrefagesi, and P. brevicirris, distinguishing features can be seen in the key, and in the remarks section of P. larentukana n. comb. Therefore, these species are herein raised from the synonymy, but a detailed revision of them should be carried out to confirm their status. The P. nuntia complex comprises 20 valid species (Table 1). However, similar to the hitherto of species herein reinstated, some others are still pendent of a comprehensive revision of the type material taking into consideration those features proposed above to confirm their taxonomic status. Likewise, it is not ruled out that some other species are still buried in the historical literature, especially in those former works were many species of nereidids from worldwide areas were briefly described and poorly or non-illustrated (e.g. Savigny in Lamarck 1818; Savigny 1822; Schmarda 1861; Kinberg 1865; de Quatrefages 1866; several by Grube); as was the case of P. latipalpa n. comb. and P. larentukana n. comb. Thus, before describing new species and hence increasing the list of names, it is essential to keep in mind that there are still several species remaining to be re-discovered. For instance, a shallow review of a few original descriptions suggests that Nereis ehrenbergi Grube, 1868 (Red Sea) might belong to the P. nuntia complex by having an arc of paragnaths on AVI (Grube 1868), which in fact seems closer to P. heterodonta, also from the same region. Likewise, Nereis (Nereilepas) pacifica Schmarda, 1861 (New Zealand) has been regarded as a synonym of P. vallata (Chile) since Ehlers (1904), who also synonymized Nereis latipalpa (South Africa) and Nereis brevicirris (St. Paul Island) with Nereis vallata (nowadays in Perinereis). However, as was noted above, the latter two species are different from P. vallata and recognized as valid. The same situation may occur with N. (Nereilepas) pacifica, an accurate comparison of the type material of both synonymized species would be desirable to clarify their status.