﻿Two new records and description of a new Perinereis (Annelida, Nereididae) species for the Saudi Arabian Red Sea region

﻿Abstract Annelid biodiversity studies in the Red Sea are limited and integrative taxonomy is needed to accurately improve reference libraries in the region. As part of the bioblitz effort in Saudi Arabia to assess the invertebrate biodiversity in the northern Red Sea and Gulf of Aqaba, Perinereis specimens from intertidal marine and lagoon-like rocky environments were selected for an independent assessment, given the known taxonomic ambiguities in this genus. This study used an integrative approach, combining molecular with morphological and geographic data. Our results demonstrate that specimens found mainly in the Gulf of Aqaba are not only morphologically different from other five similar Perinereis Group I species reported in the region, but phylogenetic analysis using available COI sequences from GenBank revealed different molecular operational taxonomic units, suggesting an undescribed species, P.kaustianasp. nov. The new species is genetically close and shares a similar paragnath pattern to the Indo-Pacific distributed P.helleri, in particular in Area III and Areas VII–VIII. Therefore, we suggest it may belong to the same species complex. However, P.kaustianasp. nov. differs from the latter mainly in the shorter length of the postero-dorsal tentacular cirri, median parapodia with much longer dorsal Tentacular cirri, posteriormost parapodia with much wider and greatly expanded dorsal ligules. Additionally, two new records are reported for the Saudi Neom area belonging to P.damietta and P.suezensis, previously described only for the Egyptian coast (Suez Canal) and are distributed sympatrically with the new species, but apparently not sympatric with each other.


Introduction
Based on genetic databases (i.e., BOLD and GenBank), and despite the recent advances in integrative studies focused on polychaetes (i.e., Nygren et al. 2010;Villalobos-Guerrero et al. 2021;Teixeira et al. 2023), there are still many taxonomic ambiguities and unidentified annelid species in some groups of Nereididae (i.e., Martin et al. 2021;Elgetany et al. 2022).Perinereis Kinberg, 1865 is one of the most diverse genera in this family, currently including between 97 (Wilson et al. 2023) to 106 (WoRMS Editorial Board 2024) valid species distributed worldwide.From these, approximately 16 species are reported for the Arabian Peninsula (Ocean Biodiversity Information System, OBIS ;Mohammad 1971;Wehe and Fiege 2002).Due to apparent similar paragnath patterns, overall body features and lack of detailed systematic studies, Perinereis species are often problematic to identify to the species level (Bakken and Wilson 2005;Yousefi et al. 2011).This has led to informal denomination of species complexes and recognition of geographic morphs and varieties such as P. cultrifera (Grube, 1840) species group (type locality: Naples, Italy; Scaps et al. 2000) and the P. nuntia (Lamarck, 1818) species (type locality: Gulf of Suez, Egypt) group (Wilson and Glasby 1993;Glasby and Hsieh 2006;Sampértegui et al. 2013), both reported for the Red Sea (OBIS).Thanks to molecular data, it is now easier to screen for potential new species with apparent similar morphotypes.Recent evidence comparing populations from different regions has shown that when specimens differ genetically, further analysis of the diagnostic morphological features often leads to the recognition of distinct features that were previously overlooked (i.e., Sampértegui et al. 2013;Teixeira et al. 2022b).A recent review on meiofauna (Cerca et al. 2018) and recent polychaete studies (i.e., Abe et al. 2019;Tilic et al. 2019;Martin et al. 2020), including from Nereididae (Glasby et al. 2013;Sampieri et al. 2021;Teixeira et al. 2022a, b) also demonstrate that cryptic and pseudo-cryptic species often have geographically restricted distributions, with the range of cryptic species being smaller than the parent morphospecies.
The Egyptian side of the Red Sea has been the focus of an increasing amount of polychaete studies either reviewing existing species groups (i.e., Villalobos-Guerrero 2019) or describing new species that were previously considered cryptic (i.e., Elgetany et al. 2022).The northern Saudi Arabian Red Sea and Gulf of Aqaba, despite being expected to host a large biodiversity (Roberts et al. 2002;DiBattista et al. 2016), has seen comparatively few biodiversity studies involving molecular techniques, particularly for polychaetes.To address this gap, and document the invertebrate biodiversity of the region, a bioblitz was conducted in the Neom region (northern Saudi Arabian Red Sea and Gulf of Aqaba) to document the local biodiversity, with emphasis on mobile invertebrates and cryptobenthic fish.As part of this effort, this study used a molecular approach, combined with morphological and geographic data, to investigate Perinereis samples collected from marine intertidal and lagoon-like rocky environments of the northern Red Sea.In particular, we aimed to assess species distributions and to investigate whether specimens collected belonged to existing P. cultrifera group, P. nuntia group, to other similar Perinereis species reported for the region, or if new species were undescribed.

Sampling effort
The NEOM bioblitz sampling campaign surveyed 38 shallow and coral reef sites up to 25 meters depth and some intertidal habitats, along the northern region of the Saudi Arabian Red Sea and Gulf of Aqaba (Neom area).This initiative aims to initiate a biodiversity inventory of marine benthic invertebrates (mainly mobile) and cryptobenthic fish in the Red Sea using DNA barcoding and metabarcoding.Only intertidal marine and lagoon-like rocky environments were considered for the purpose of this study, in order to perform an independent assessment within Perinereis, given the known taxonomic ambiguities in several species within the genus from this particular habitat.
Table 1 details the number of original specimens collected for each sampling location, which correspond to the same number of COI sequences analysed.The number of COI sequences from Perinereis species publicly available in GenBank, respective sampling area and references are also detailed in Table 1 and were used for comparison purposes.The collected Red Sea Perinereis specimens were deposited at NTNU University Museum, Trondheim, Norway (NTNU-VM, Bakken et al. 2024; vouchers: NTNU-VM-86010-NTNU-VM-86044).Perinereis oliveirae specimens are deposited at Biological Research Collection of the Department of Biology of the University of Aveiro (CoBI at DBUA; curated by Ascensão Ravara: aravara@ua.pt;vouchers: DBUA0002494.02.v01 and DBUA0002494.02.v02),Portugal.Specimens that were exhausted in the DNA analysis were assigned only with the Process ID from the BOLD systems (http://v4.boldsystems.org/),corresponding to MTPNO009-23 (Gulf of Aqaba, Magna).Some specimens were preserved in 96% ethanol and others in formalin with a respective sample tissue preserved in ethanol for molecular work (detailed in Suppl.material 1).

DNA extraction, PCR amplification, and alignments
DNA sequences of the 5' end of the mitochondrial cytochrome oxidase subunit I (mtCOI-5P) were obtained for all the collected Perinereis specimens and used for the main analysis.A representative number of specimens per location for the new species were also sequenced using the mitochondrial 16S rRNA and D2 region of nuclear 28S rRNA, for future reference purposes.
DNA extraction was performed using QuickExtract DNA Extraction Solution (Lucigen) with 50 µl of the reagent per Eppendorf.The tubes were then transferred to a heat block at 65 °C for 30 min and then an additional 2 min at 98 °C.Depending on the specimen size, only a small amount of tissue (i.e., a single parapodium) or the posterior end of the worm was used.
PCR reactions were performed using a premade PCR mix from VWR containing 10 µl per tube of Red Taq DNA polymerase Master Kit (2 mM, 1.1×), 0.5 µl of each primer (10 mM) and 1 µl of DNA template in a total 12 µl volume reaction.Table 2 displays the PCR conditions, primers and sequence lengths for the different markers.Amplification success was screened in a 1% agarose gel, using 1 μl of PCR product.Successful PCR products were then purified using the Exonuclease I and Shrimp Alkaline Phosphatase (ExoSAP-IT, Applied biosystems) protocol, according to manufacturer instructions.Cleaned up amplicons were sent to KAUST Sanger sequencing service for forward sequencing.

Phylogenetic analysis and MOTU clustering
For comparison purposes, GenBank COI sequence data from P. marionii (Audouin & Milne Edwards, 1833); P. vallata (Grube, 1857); P. helleri (Grube, 1878) and the outgroup Alitta virens (M.Sars, 1835) completed the final dataset (Table 1, Suppl.material 1).The phylogenetic analysis was performed through maximum likelihood (ML) for the entire dataset.Best-fit models were selected using the Akaike Information Criterion in MEGA.The phylogenetic relationship analysis was executed with 500 bootstrap runs using the General Time-Reversible model with gamma distributed rates and a portion of the sites invariable (GTR+G+I).The final version of the tree was edited with the software Inkscape v. 1.2 (https://www.inkscape.org).
Three delimitation methods were applied to obtain Molecular Operational Taxonomic Units (MOTUs): The Barcode Index Number (BIN), which makes use of the Refined Single Linkage (RESL) algorithm available only in BOLD (Ratnasingham and Hebert 2013); the Assemble Species by Automatic Partitioning (ASAP, Puillandre et al. 2021), implemented in a web interface (https://bioinfo.mnhn.fr/abi/public/asap/asapweb.html) with default settings using the Kimura-2-Parameter (K2P) distance matrix; lastly, the Poisson Tree Processes (bPTP; Zhang et al. 2013) performed in a dedicated web interface (https:// species.h-its.org/),using the ML phylogeny obtained above, for 500000 MCMC generations and twenty-five percent of the samples discarded as burn-in.
The mean genetic distances for mtCOI (K2P; Kimura 1980) within and between MOTUs were calculated in MEGA.

Morphological analysis
Specimens were studied using a Leica stereo microscope (model M205 C).Stereo microscope images were taken with a Flexacam C3 camera.Compound microscope images of parapodia and chaetae were obtained with a Leica DM2000 LED imaging light microscope, equipped with a Flexacam C3 camera, after mounting the parapodia on a slide preparation using Aqueous Permanent Mounting Medium (Supermount).Parapodial and chaetal terminology in the taxonomic section follows Bakken and Wilson (2005) with the modifications made by Villalobos-Guerrero and Bakken (2018).The final figure plates were edited with the software Inkscape v. 1.2.
For measuring length of dorsal ligules, not only the lengths of the tips were considered, but the proximal part of the ligules was also included (e.g., Conde-Vela and Salazar-Vallejo 2015; Villalobos-Guerrero and Carrera-Parra 2015; Teixeira et al. 2022b).Like Hutchings et al. (1991), a specimen is described as having a greatly expanded dorsal notopodial ligule posteriorly only if the dorsal ligule is more than two times as long as the ventral ligule.For analysis of variation, only complete specimens were considered; total length (TL), length up to chaetiger 15 (L15), width at chaetiger 15 (W15) were measured with a millimetre rule under the stereomicroscope.Number of chaetigers (NC) were also taken into consideration.TL was measured from anterior margin of prostomium to the end of the pygidium, and W15 were measured excluding parapodia.Measurements of the length of the antennae (AL), palps (PL), dorsal cirri (DCL), dorsal ligule (DLL), ventral cirri (VCL), ventral ligule (VLL), median ligule, the length and width of the head (HL and HW, respectively), and the length of all four tentacular cirri, including the longest one (postero-dorsal cirri, DPCL), were also retrieved.Heterogomph falciger blade size comparison (short, long, and extra-long) based on Wilson et al (2023).Spiniger serration based on the comparison between P. cultrifera (lightly serrated) and P. rullieri (coarsely serrated) from Pilato (1974).
Paragnath counts were performed to compare patterns with other morphologically similar Group I Perinereis species (Hutchings et al. 1991).Pharynx paragnath terminology follows Bakken et al. (2009) and paragnath description of areas VII and VIII follow Conde-Vela (2018).
Terminology for molecular vouchers follows Pleijel et al. (2008) and Astrin et al. (2013).Overall description follows a similar structure to those of Villalobos-Guerrero (2019).Dates of sample collection follow the DD/MM/YY format.

Phylogenetic analyses
The phylogenetic reconstruction recovered ten MOTUs of Perinereis (Fig. 1A), the delimitation of which are cohesively supported by the three species-delimitation tests applied, except for MOTU 1 and GB1, which are clustered together with the ASAP method.Sequences from P. fayedensis and P. anderssoni are not present in BOLD and have no associated BIN.

Taxonomic account
Distribution and habitat.Confined to the northeastern Red Sea (Duba, Shushah Island) and Gulf of Aqaba (Magna) so far.Type locality: Saudi Arabia, Gulf of Aqaba: Magna region (marine site), 28°26'57.3"N,34°45'35.4"E.Specimens collected both in lagoon-like environments and fully marine sites in rocky areas, usually among coarse-grained sand under rocks.Apparently more abundant and easier to find in marine sites from the Gulf of Aqaba.Can be found in sympatry with P. damietta (Fig. 1B, C) and P. suezensis (Fig. 1B, D).The latter two species as described by Elgetany et al. (2022).
Etymology.The species designation pays tribute to the King Abdullah University of Science and Technology (KAUST) in Saudi Arabia, a globally recognized graduate-level research institution.This naming honours KAUST's substantial and enduring contributions to marine science, particularly in advancing our understanding of the Red Sea over the course of more than a decade.Through its dedicated research efforts, KAUST has significantly enriched the scientific community's knowledge of this unique marine environment.Description.Specimens used: NTNU-VM-86011 (holotype) and NT-NU-VM-86015 (paratype), both preserved in ethanol 96%, stored at NTNU University Museum (Norway, NTNU-VM).
Head (Fig. 2A, B, E, J): Prostomium pyriform, 1.2× wider than long; 2.5× longer than antennae.Palps with a round or conical palpostyle (Fig. 2A); palpophore longer than wide, subequal to the entire length of prostomium.Antennae separated, gap half of antennal diameter (Fig. 2E); tapered, less than half the length of the palpophore.Eyes black, anterior and posterior pairs well separated (Fig. 2J).Anterior pair of eyes oval shaped, as wide as antennal diameter; posterior pair of eyes round or oval shaped, subequal width to anterior pair.Distance between the anterior eyes 1.25× longer than posterior ones.Nuchal organs covered by the tentacular belt.
Pharynx: Pair of dark brown curved jaws with 7-8 denticles; two longitudinal canals emerging from the pulp cavity, both in the mid-section of the jaw (Fig. 2C).Pharynx consisting of maxillary and oral rings with conical shaped paragnaths (Fig. 2A, B).Maxillary ring: Area I = two small paragnaths arranged in a longitudinal line (Fig. 2F).Area II = Cluster of 5-7 small paragnaths (Fig. 2F).Area III = central patch of nine small paragnaths, lateral patches with two small paragnaths each (Fig. 2D).Area IV = 13 small paragnaths arranged in wedge shape without any bars (Fig. 2D).Oral ring: Area V = a triangle of three large paragnaths (Fig. 2E).Area VI (a+b) = two narrow bar-shaped paragnaths, one on each side, displayed as a straight line (Fig. 2E).Areas VII-VIII = 20-24 small paragnaths in total; Area VII, ridge region with two transverse paragnaths, furrow regions with two longitudinal paragnaths each (Fig. 2G); Area VIII, ridge regions with one paragnath each, furrow regions with two longitudinal paragnaths each (Fig. 2G).
Remarks.Some nereidid species groups can have similar morphological features, including paragnath patterns, that may cause misidentifications.The new species COI clade revealed no GenBank match based on the BLAST tool.Perinereis kaustiana sp.nov.and a sequence belonging to a specimen from Malaysia identified as P. helleri (type locality: Bohol, Philippines) not only are sister to each other and phylogenetically close (Fig. 1A; 19.9 ± 2.4% K2P COI distance), but they also seem to share the same paragnath sizes, shapes and patterns (Park and Kim 2017: 255, fig. 4e; sampled in South Korea; no molecular data available), including in Area III, with the presence of lateral patches with two paragnaths each (Fig. 2D) and the same paragnath arrangements in the furrow and ridge regions of Areas VII-VIII (Fig. 2G).This makes them morphologically very similar and possibly belonging to the same cryptic complex, which could range from the Red Sea to the Indo-Pacific based on the available COI data.However, P. kaustiana sp.nov.seems to differ from P. helleri in some key features: shorter postero-dorsal tentacular cirri, reaching up to chaetiger 9, instead of the reported chaetiger 16 for P. helleri; median parapodia with much longer dorsal cirri (3×) compared to ventral one; posteriormost parapodia with much wider dorsal ligule (2.5-3.0×)than the median ligule (Fig. 3C, I) and dorsal ligule greatly expanded (3× longer than ventral ligule).Based on parapodia drawings from Hutchings et al. (1991: 255, fig. 9; Syntype ZMB Q3464), the ratio between dorsal and ventral cirri in P. helleri is subequal to slightly longer than ventral cirri throughout the body and posteriormost dorsal ligules with double the width of median ones and slightly expanded (up to 2× the length of Table 4. Comparison between selected characters in the most morphologically similar species to P. kaustiana sp.nov., reported for the Arabian Peninsula and Mediterranean Sea and lacking DNA data.The Indo-Pacific P. helleri is also included.Morphological details of paragnath patterns for P. cultrifera and P. rullieri species complexes also includes partial data from topotypical specimens belonging to the private collection of the first author, to be published in the forthcoming future.Mohammad 1971;Hutchings et al. 1991Hutchings et al. 1991;Pilato 1974Pilato 1974 the ventral ligules; Table 4).Furthermore, P. helleri from Hutchings et al. (1991) does not seem to possess ligules with finger-like ending tips.

Characters
Other species with similar paragnath patterns are Perinereis anderssoni (Kinberg 1865: 167-179; Park and Kim 2017: 255, fig. 4d) and Perinereis rullieri (Pilato 1974: 25-36, figs 1-4), which share the same small sized paragnaths as P. kaustiana sp.nov., but instead the former two species possess only one paragnath in each lateral patch of Area III and paragnaths in Areas VII and VIII are usually arranged in two regular rows, without any discernible pattern in the furrow or ridge regions.Perinereis anderssoni is reported in the Atlantic region of the American continent (type locality: Rio de Janeiro, Brazil), while P. rullieri is apparently restricted to the Mediterranean Sea (type locality: between Aci Trezza and Augusta, eastern coast of Sicily, Italy).Moreover, the morphological similar lineages found within the Perinereis cultrifera (Grube 1840: 74, fig. 6;Hutchings et al. 1991: 253-254, fig.8a-c) species complex, including P. euiini (Park and Kim 2017: 252-260, figs 1, 2, 4a, b, 5, tables 1, 4, described for South Korea), are different from P. kaustiana sp.nov.due to the overall larger paragnath sizes, lack of any lateral patches in Area III, and the presence of shorter heterogomph falcigers (Park and Kim 2017: 254, fig. 2L).Specimens of Perinereis cultrifera from Lobo et al. (2016) were misidentified and are in fact P. oliveirae (Horst 1889: 38-45, plate 3;Fauvel 1923: 354, fig.138 e-k), the latter characterised by the presence of three paragnaths in lateral patches in Area III, while this feature is absent in P. cultrifera.Perinereis oliveirae is described for the northern Iberian Peninsula, having also very long bar-shaped paragnaths in Areas VI and very short tentacular cirri compared to length of the head (reaching chaetigers 1 and 2).features were confirmed based on the two P. oliveirae specimens from this study and samples from the private collection of the first author of this study.

Discussion
Our molecular data provides compelling evidence for the existence of a new, deeply divergent, and completely sorted species within the Perinereis species Group I in the Red Sea.At first glance, P. kaustiana sp.nov.can be easily misidentified as the well-known and allegedly cosmopolitan P. cultrifera, due to the classic two bar shaped paragnaths in Areas VI and proximity with the Mediterranean Sea.This might be the reason the latter is usually reported for the Red Sea (Wehe and Fiege 2002;Bonyadi-Naeini et al. 2018;OBIS), but a greater sampling effort in the central and southern Red Sea regions are needed to confirm this.Morphological features, such as the paragnath arrangement, as well as the length of tentacular cirri and ratios within the parapodia also allowed the distinction of P. kaustiana sp.nov.from other similar species (see taxonomic key and Tables 4, 5).Upon careful morphological examination, P. kaustiana sp.nov. is morphologically closer to the Indo-Pacific P. helleri, than it is to the European P. cultrifera, based mainly on paragnath patterns, particularly in Areas III (Fig. 2D) and VII and VIII (Fig. 2G), and similar length of the falciger blades.Paragnath features in Areas VII and VIII lends support to the taxonomic importance of highlighting faint ridges and furrows in the ventral oral ring for certain Perinereis species (Conde-Vela 2018), which usually are not accounted in species descriptions due to no apparent pattern being found (i.e., Teixeira et al. 2022a).Perinereis kaustiana sp.nov.and P. helleri are also phylogenetically closely related (Fig. 1A), despite being divergent lineages, with genetic distances that are in the range used for delimitating polychaete species (i.e., Kvist 2016;Lobo et al. 2016;Nygren et al. 2018).This situ-* No available chaetae data for P. striolata.
ation, together with the absence or subtle morphological differences previously overlooked, resembles cryptic lineages within a species complex (Teixeira et al. 2022b(Teixeira et al. , 2023)), and further sampling efforts between the Red Sea to the Indo-Pacific region are needed to assess this.
The new species is so far unique to the northern Red Sea and apparently easy to find in the rocky beaches of the Gulf of Aqaba.Considering the high rate of endemism in the Red Sea (DiBattista et al. 2016), this species may indeed be endemic to this Sea, although further sampling across this region and the Indo-Pacific area might prove it to be more widespread.In the remaining sampling sites further south, along the northern Saudi coast, P. kaustiana sp.nov. is outcompeted by the sympatric distributed Perinereis nuntia species group, which seems to be the dominant coastal annelid in the region (Fig. 1B).The latter is also a species complex with several different species recently revised by Villalobos-Guerrero (2019).Our specimens initially identified as belonging to the P. nuntia complex revealed at least two different morphotypes, which after further morphological (mainly based on paragnath patterns, Fig. 1C, D) and molecular review corresponded to the new species recently described by Elgetany et al. (2022) for the neighbouring Egyptian coast (Suez Canal), namely P. damietta (Fig. 1C) and P. suezensis (Fig. 1D).These species are sympatric with P. kaustiana sp.nov., but apparently not sympatric with each other in the studied region (Fig. 1B).Perinereis damietta (which is morphologically more similar to P. heterodonta Gravier, 1899 than to P. nuntia according to Elgetany et al. (2022)), was found mainly in lagoon-like environments, whereas P. suezensis only in fully marine areas.Perinereis kaustiana sp.nov.shared both marine and lagoon-like habitats, with all the three sampled species found in intertidal coarse-grained sand, under rocks or cobles.As speculated by Elgetany et al. (2022), P. damietta seems to have a slightly wider habitat preference, since some of our specimens (from Al Muwaileh lagoon) also occurred sub-tidally, attached to small rocks at approximately 1 meter depth.

Figure 1 .
Figure 1.Phylogenetic tree and MOTU distribution for the three sampled Red Sea Perinereis species A maximum likelihood phylogeny based on COI sequences, with information regarding the different MOTU delineation methods.Numbered MOTUs (1-4) contain original sequences from Perinereis specimens analysed in this study; MOTUs "GB" are based on Perinereis sequences mined from GenBank; MOTU "OUTG" correspond to the rooted outgroup, Alitta virens.Bootstrap values lower than 80% not displayed B Red Sea MOTU distribution; each coloured pie corresponds to a unique species and respective abundance proportion; larger pie charts indicate higher number of sympatric species.Species from the Suez Canal based on mined GenBank sequences from Elgetany et al. (2022); abundance proportion based on type material C Perinereis damietta, focus on prostomium and pharynx, dorsal view, specimen NTNU-VM-86031 D Perinereis suezensis, focus on prostomium and pharynx, dorsal view, specimen NTNU-VM-86032 E Perinereis kaustiana sp.nov., focus on prostomium and pharynx, dorsal view, specimen NTNU-VM-86011.Scale bars: 500 μm (C-E).

Table 1 .
Species, number of sequences (n), geographic location, and their respective GenBank COI accession numbers for the original material and sequence data used from other studies.

Table 2 .
Primers and PCR conditions used in this study.