Plectorhinchus makranensis (Teleostei, Haemulidae), a new species of sweetlips from the Persian Gulf and the Gulf of Oman

Abstract Plectorhinchus makranensissp. nov. is described on the basis of 16 specimens from the Persian Gulf and Gulf of Oman, in the Northwest Indian Ocean. The new species can be distinguished from congeners by a combination of dorsal fin rays XII, 18–20, pectoral-fin rays 16–17, tubed lateral-line scales 55–57, gill rakers count (10–12 on the upper limb and 16–17 on the lower limb), 17–18 scales between the lateral line and the first anal-fin spine, 30–31 circumpeduncular scale rows and color pattern. Plectorhinchus makranensissp. nov. is distinguished from P. schotaf by having the posterior margin of the opercular membrane grey (vs. red in P. schotaf), fewer circumpeduncular scale rows, and a shorter base of the soft portion of the dorsal fin, 27.6–29.4% of standard length (SL) (vs. 31–32.3% of SL in P. schotaf). The new species resembles P. sordidus but is differentiated from it by having more gill rakers, a smaller orbit diameter 27.5–32.1% of head length (HL) (vs. 35.5–37.2% of HL in P. sordidus), a longer caudal peduncle 19.2–21.3% of SL (vs. 17.1–17.9% of SL in P. sordidus), and the first to third pectoral-fin rays light gray (vs. dark gray in P. sordidus). The new species can also be distinguished from the other species, including P. schotaf and P. sordidus, based on COI and Cyt b molecular markers. The phylogenetic position of this new species indicates that it is a sister taxon of P. schotaf.


Introduction
tocol. Sequences were amplified by PCR using the following primer pairs: cytochrome oxidase subunit 1 (CO1LBC_F: 5' TCAACYAATCAYAAAGATATYGGCAC 3'; CO1HBC_R: 5' ACTTCYGGGTGRCCRAARAATCA 3') and Cytochrome b (GluF: 5'AACCACCGTTGTATTCAACTACAA3; ThrR: 5'ACCTCCGATCTTCGGATT-ACAAGACCG3), following Ward et al. (2005) and Machordom and Doadrio (2001), respectively. PCR conditions for the COI gene included: initial denaturation 94 °C, 1 min then 30 cycles at 95 °C for 30 s, 52 °C for 45 s, and 72 °C for 1 min, followed by a final extension at 72 °C for 5 min. All amplification conditions were similar the COI gene except for the annealing temperature (54 °C) and the number cycles (35) for the Cyt b gene. The quality of PCR products was determined by running them on 1.5% agarose gels in 0.5X TBE buffer. The products were sent to Microsynth Company (Switzerland) for sequencing. We analysed a total of 64 sequences of Plectorhinchus species, including 36 for COI and 28 for Cyt b. Sequences of Pomadasys maculatus (Bloch, 1793) and Haemulon aurolineatum Cuvier, 1830 were used as outgroups.

Molecular data analyses
All sequence alignments were performed using the MAFFT algorithm. The pairwise DNA sequence differences within and between species of Plectorhinchus were calculated with MEGA 7.0.9 (Kumar et al. 2016) based on the Kimura two-parameter (K2P) model. The best-fit nucleotide substitution models were determined by jMod-elTest (Posada 2008) for each gene and combination of two genes. Based on Akaike information criterion (AIC), the preferred model for the two molecular markers was TVM + I + G. Analyses of phylogenetic relationships were performed for each gene and combination of two genes (Cyt b + COI) using maximum likelihood (ML) and Bayesian inference (BI). ML analysis as implemented in RAxML 7.2.6 (Stamatakis 2014) with 10,000 bootstrap replicates. BI analysis was run for 30,000,000 generations in MrBayes 3·1·2 (Ronquist et al. 2012) with two independent runs of four Markov Chain Monte Carlo (MCMC). The first 25% of the trees were excluded as burn-in and remaining trees sampling were used to compute a 50% majority rule consensus tree. The resulting phylogenetic trees from ML and BI analyses were edited using FigTree v.1.4.4. Additionally, we used from two different approaches for species delimitation including the Automatic Barcode Gap Discovery (ABGD) (Puillandre et al. 2012) and Bayesian Poisson Tree Process (bPTP) (Zhang et al. 2013). The ABGD method based on the COI gene was performed on web http://wwwabi.snv.jussieu.fr/ html, under the Kimura (K80) model with the default parameters of Pmin = 0.001 to Pmax = 0.1, steps = 10, X (relative gap width) = 1.5, Nb bins = 20. The bPTP approach used the best ML tree, which was run on the web server (http://species.h-its.org/ptp). This analysis was processed with 500,000 MCMC generations and 25% of burn-in.

Morphological analysis
We used Johnson and Wilmer (2015) for morphometric and meristic characteristics which consisted of 23 morphometric and seven meristic features. The univariate and multivariate analysis were run in SPSS v.16 (SPSS Inc., Chicago IL) and PAST v. 4.03 (Hammer 2020). We assessed the normally distributed parametric data using the Shapiro-Wilk test. The morphometric characters were divided by standard length (SL) and head length (HL) to remove the size effect from the dataset. The univariate Analysis of Variance (ANOVA) was performed for morphometric characters to evaluate the significance of phenotypic differences between species. The principal components analyses (PCA) was used for multivariate analyses to characterize the morphological variation among species.
Body elongate, moderate deep, its depth 2.8-3.4 in SL, compressed laterally and covered with ctenoid scales; scales on the middle of the body largest; lateral line extends slightly as smaller scales onto the caudal-fin base; scales present on suborbital; snout and chin without scales; predorsal scales extending through interorbital. Head moderately large, head length 3.4-3.7 in SL, upper profile convex; mouth moderately small and terminal, lips fleshy, upper jaw protruding slightly beyond the lower jaw; nostrils small, posterior nostril half diameter of anterior nostril, anterior nostril on horizontal line through the lower margin of eye; orbit diameter 3.3-3.9 in HL; three pores on each side of the chin, but no pit; teeth cardiform, approximately 2 rows laterally and 5 rows anteriorly in the upper jaw, approximately 2 rows laterally and 6 rows anteriorly in lower jaw, approximately 20-24 teeth in the upper jaw on each side and approximately 16-18 in the lower jaw on each side, palatine and vomer without teeth. Opercle with a single, exposed, short and weak spine; preopercle slightly concave and serrate, including few serrae on the posteroventral margin.
Origin of dorsal fin above the pectoral-fin base, first spine shortest, fifth spine longest, first dorsal-fin spine about 1.2 length of fifth, first spine 6.4 (6.1-6.5) in HL, fifth spine 2.6 (2.2-2.7) in HL, 6 th and 7 th soft dorsal-fin ray longest, 6 th and 7 th 3.6 in HL, 18 th to 20 th soft dorsal-fin ray shortest, its length 9.6-9.8 in HL, base of soft portion of dorsal fin 1.1 in base of the spinous portion; anal fin short, with somewhat rounded posterior margin, origin below base of 7 th soft dorsal-fin ray, second spine longest, first ray is the longest, anal-fin length 2.5 (2.3-2.6) in HL; posterior margin of caudal fin slightly emarginate, caudal-fin length 1.7-1.8 in HL; pectoral fin reaching vertical between bases of seventh and eighth dorsal-fin spines, pectoral-fin length 1.4-1.5 in HL. Origin of pelvic fins behind pectoral-fin base, its tip reaching vertical at ninth dorsal-fin spine, second ray longest, pelvic-fin length 1.4-1.5 in HL.
Color pattern in preservative. (Holotype: Fig. 3). Head and body steel grey; head and edge of fins slightly darker than the rest of the body; posterior part of opercular membrane dark grey; lips grey; ventral part of body including underside of head and belly to lower part of caudal peduncle white; iris yellow.
Color in fresh. (paratypes: Fig. 4). Body silver-grey; all fins dark grey; pectoral-fin base light grey; uppermost first to third pectoral-fin rays light gray; orbital margin orange; iris grey; ventral part of body including subopercle, chest and pectoral-fin margin opaque white; lips and chin pink-grey; posterior part of opercular membrane grey.
Distribution and habitat. The new species has been observed at six localities along the coast of the Gulf of Oman and the Persian Gulf in the Northwest Indian Ocean. Abundance was greater in the Gulf of Oman compared to the Persian Gulf. All specimens have been collected from shallow rocky and coral areas. Other species of this family which occur sympatrically at the type locality (Beris coast) with Plectorhinchus makranensis sp. nov. include: Diagramma pictum, Plectorhinchus pictus, Pomadasys kaakan, P. maculatus and P. stridens.
Etymology. The species name is derived from the Makran coast and refers to the coastal land in southeastern Iran and southwestern Pakistan, north of the Gulf of Oman. Multivariate analysis. The first two Principal Components (PCs) of the meristic and morphological characters accounted for 78.3% and 60% of the variation, respectively (Fig. 5A, B). In the meristic PCA, the number of total gill rakers, circumpeduncular scales and transverse scale rows below the lateral line loaded heavily on the first PC. Both species were completely distinguished along the first axis ( Fig 5A). In the morphometric PCA, measurements including the length of the soft dorsal-fin base, body depth and depth of the caudal peduncle separated Plectorhinchus makranensis sp. nov. from P. schotaf along the first PC1 (Fig. 5B).

Discussion
The present study adds four species (P. flavomaculatus, P. makranensis, P. caeruleonothus and P. unicolor) to the previous molecular reconstructions (Sanciangco et al. 2011;Liang et al. 2014;Tavera et al. 2018). Our phylogenetic analysis is consistent with previous morphological and molecular studies. Based on ecological, morphological, color characteristics and biogeography, Plectorhinchus is divided into two clades: clade A includes species with deeper body and uniformly dull color compared to clade B (Johnson et al. 2001;McKay 2001;Tavera et al. 2018) (Fig. 2). The species of clade A are usually distributed in the Western Indian Ocean, with two exceptions: P. macrolepis and P. unicolor are scattered in the Eastern Atlantic and the West Pacific, respectively (Wirtz et al. 2007;Johnson and Wilmer 2015). Species of clade B are usually distributed from the East Indian to the West Pacific Ocean with the exceptions of P. centurio and P. gaterinus, which are only found in the Western Indian Ocean (Fricke et al. 2018).
A combined morphological and molecular approach should be used to distinguish closely related species (Baldwin et al. 2011;Lavoué and Sullivan 2014;Bogorodsky et al. 2017).
Based on molecular and morphological data (Fig. 2, Suppl. material 2: Table S2), P. schotaf is the sister taxon of P. makranensis sp. nov. Genetic distance between the new species and P. schotaf based on COI and Cyt b markers is consistent with species-level divergences in other fish taxa (Johns and Avise 1998;Ward et al. 2005;Hsu et al. 2007;Ward 2009). Also, these two sympatric species in the Persian Gulf and the Gulf of Oman show higher genetic distance than other congeneric species pairs from the East Africa coast to the Red Sea based on both mtDNA markers 3.42% between P. chubbi and P. sordidus for the COI gene, and 3.50% between P. chubbi and P. sordidus for the Cyt b gene (Suppl. material 1: Table S1).
The new species is morphologically most similar to P. schotaf and P. sordidus. The coloration of the new species differs from P. schotaf by having the posterior margin part of the opercular membrane grey (Fig. 4) (vs. red in P. schotaf (Fig. 6A,  B). The two species also differ in the number circumpeduncular scales 30-31 (vs. 32-34 in P. schotaf). Additionally, there are modal differences in counts, transverse scale rows below the lateral line (17-18, modally 17, vs. 18-20, modally 19 in P. schotaf), and morphometric differences, with P. schotaf having a shorter base of the soft portion of the dorsal fin (Suppl. material 2: Table S2). Plectorhinchus makranensis sp. nov. can be distinguished from P. sordidus by the number of gill rakers (10-12, modally 12 upper rakers, 16-17, modally 17 lower rakers, 26-29, modally 28, rarely 26 total rakers, vs. 9-11 upper rakers, 15-16 lower rakers, 24-26 total rakers in P. sordidus), a longer caudal peduncle and smaller orbit diameter (Suppl. material 2: Table S2) and the first to the third pectoral-fin rays light grey (Fig. 4) (vs. dark grey in P. sordidus (Fig. 6C, D)). Additionally, these two species can be different from each other in the number of tubed lateral-line scales (55-57, modally 56, vs. 48-55, modally 54 in P. sordidus). Plectorhinchus makranensis sp. nov. is distinguished from other similar congeners as follows: from P. caeruleonothus by having 10-12 gill rakers on the upper limb (vs. 7-9 in P. caeruleonothus) and 10-11 scales above the lateral line to the base of the first dorsal-fin spine (vs. 15), from P. unicolor by having 17-18 transverse scale rows below lateral line (vs. 19-21), from P. griseus by having 18-20 dorsal-fin rays (vs. 21-23); from P. playfairi in having 55-57 lateral-line tubed scales (vs. 58-60) and 16-17 gill rakers on lower limb (vs. 21-23), from P. chubbi by XII dorsal-fin spines and 16-17 gill rakers on the lower limb (vs. XI spines and 21-23 rakers respectively). The number of dorsalfin spines is XII in new species vs. XIII in P. chrysotaenia and XIV in P. flavomaculatus, P. ceylonensis, P. gibbosus, P. macrolepis and P. plagiodesmus. Furthermore, the ANOVA analysis reveals that the numbers of dorsal-fin spines and soft rays and scales below the lateral line to the first anal-fin spine, as well as the numbers of circumpeduncular scales and total gill rakers, significantly differ from the other examined species. The molecular and morphological differences mentioned above indicate that the new species is separated from other congeners.