Crossing the borders: First record of Bleeding Wrasse, Polylepion cruentum Gomon, 1977 (Labriformes, Labridae), in the Northern Peru

The Labridae is a species-rich family of colorful fishes distributed globally in different habitats. Polylepion cruentum was described based on type material collected from the Gulf of California, and Quepos, Pacific coast of Costa Rica. It has since been reported from several other localities, including Mexico to Nicaragua and the Cocos Island. Recent collecting efforts in Northern Peru yielded an unusual species of labrid with uncertain taxonomic identity. Measurements, counts, coloration (in life and preserved) and DNA barcoding of the specimen agrees well with Polylepion cruentum. This represents the first record of the species from Northern Peru (Acapulco, Tumbes Department) and South America. The expansion of the distribution range of this species is perhaps due to the input of warm waters to Northern Peru, increasing especially during El Niño events. This new record is important to include in future marine checklists and consequently evaluate the conservation status of this species in Peru.


INTRODUCTION
The order Labriformes presents the families Labridae, Odacidae and Scaridae. Labridae is a species-rich family with 68 genera and 562 valid species (Fricke et al., 2021;Nelson et al., 2016). The family comprise colorful fishes characterized by a continuous dorsal fin, thick lips, jaws with well-developed dentition to capture a wide range of prey, and a recognizable labriform locomotion through propulsion of pectoral fins (Bray, 2017;Westneat & Alfaro, 2005). Some species are protogynous hermaphrodites, changing sex from female to male with maturity. The change is usually accompanied by stark sexual dimorphism (Bray, 2017). Species of this family are found in the Atlantic, Indian, and Pacific Oceans in different habitats, such as tidal pools, grassy, rocky or coral reefs, open sand bottoms, with a wide range of water temperatures (Hanel et al., 2002). The genus Polylepion Gomon, 1977comprise two valid species (Fricke et al., 2021, Polylepion russelli (Gomon & Randall, 1975) and Polylepion cruentum Gomon, 1977. Polylepion russelli its distributed from across the Indo-Pacific: Ryukyu Islands, Mariana Islands, Johnston Atoll, Hawaiian Islands, French Polynesia and Reunion Island (Béarez et al., 2013); and P. cruentum from the Tropical Eastern Pacific, ranging from Mexico (Baja California) to Nicaragua, and Cocos Island.
Polylepion cruentum is known as the "Bleeding Wrasse", inhabiting depths ranging from 150 to 200 m and measure up to 25 cm (Gomon, 1997). An unidentified labrid was collected during fieldwork in 2016, and was identified as Polylepion cruentum based on morphological and molecular (DNA barcoding) methods. The purpose of this study is to report the occurrence of P. cruentum from northern Peru (Acapulco, Tumbes Department), extending the distribution range of this species to include South America.

Taxon sampling and morphological analysis
In July 2016 an unknown labrid was collected from a bottom curtain net in front of 14-16 nautical miles (NM) from Acapulpo (Plateros), Tumbes in extreme Northern Peru (03°34′06.08″, ISSN On-Line: 1807-0205 ISSN Printed: 0031-1049 ISNI: 0000-0004-0384-1825 80°56′06.10″). Measurements were taken point-to-point under a stereomicroscope with a digital caliper to the nearest 0.1 mm on the left side of the specimen following Gomon (1977). Measurements are expressed as percent of standard length (SL), except head measurements, which are recorded as percent of head (Table 1).
A piece of muscle was collected from the specimen and preserved in 95% ethanol for molecular analysis. The voucher specimen was fixed in 10% formalin and deposited in the Ichthyology Collection of the Laboratorio Costero de Tumbes of Instituto del Mar de Peru (LCT-IMARPE).

Molecular analysis
DNA sequencing: Total DNA was extracted from ethanol-preserved muscle using the Wizard Genomic DNA Purification Kit (Promega, Madison, Wisconsin, U.S.A.) following the manufacturer's recommendations. Partial sequence of mitochondrial Cytochrome C Oxidase subunit I (COI) was amplified using polymerase chain reaction (PCR) using the primers Fish F1 and Fish R1 (Ward et al., 2005). All PCR products were sequenced using the BigDye sequencing kit cycle terminators (Applied Biosystems, California, U.S.A.) according to the manufacturer's instructions, and finally the sequences were analyzed using ABI PRISM 3130 Genetic Analyzer (Applied Biosystems).

Sequence alignment and phylogenetic analyses:
Contigs were assembled and edited in Geneious v5.5.6 (Biomatters, Auckland, New Zealand), and aligned using the Muscle algorithm under default parameters (Edgar, 2004). The sequence was aligned and translat-ed using Geneious v5.5.6 (Biomatters, Auckland, New Zealand) to check for the presence of stop codons. The COI gene sequence of P. cruentum from Peru was compared with other sequences of Polylepion cruentum, as well as those of P. russelli. Species related to Polylepion as proposed by Gomon (1977Gomon ( , 1997 were also included. These were Bodianus bimaculatus, B. neopercularis, Decodon melasma, and Semicossyphus pulcher. Abudefduf troschelii (Pomacentridae) was used as outgroup. All sequences were retrieved from Genbank and Bold system ( Table 2). Analysis of genetic distance according to the Kimura-2-parameter model, including bootstrap analysis (Felsenstein, 1985) with 1,000 replications was performed using MEGA v. X to provide a graphic representation of the pairwise distance among labrid species (Kumar et al., 2018).
Maximum-likelihood (ML) analysis was performed using RAxML Web-Servers Black-Box (Stamatakis et al., 2008) with GTR + G + I as the model. Random starting trees were used for each independent ML tree search, and all other parameters were set at default values.  Gomon (1977) introduced for comparison. Standard length (SL) is expressed in mm; all other measurements are expressed as percentages of SL, except for subunits of head that are expressed as percentages of head length (HL).

Genetic data:
The COI sequence alignment consists of 652 base pairs, with 197 variable nucleotides and 141 parsimony informative sites. The estimated index of substitution saturation (Iss) performed in DAMBE 5.2.31 (Xia & Xie, 2001) showed that the data was not saturated (i.e., Iss.c value greater than Iss). Genetic distances (Kimura, 1980) of the COI gene between Polylepion cruentum and P. russelli was 0.111 ± 0.015; and between Polylepion cruentum from Peru and P. cruentum from Mexico was zero (Table 3).

DISCUSSION
Results from our phylogenetic analyses shows that the clade of Polylepion genus (P. cruentum and P. russelli) are supported by a high bootstrap index and are sister group relationship of Bodianus and Semicossyphus in the Maximum Likelihood analyses (Fig. 3). The result reinforces overall similarities with Gomon's (1997) phylogeny,  but according to our results, the genus Polylepion is more closely related to genus Bodianus and Semicossyphus, than to the genus Decodon. Gomon (1997), listed several primitive morphological and osteological features found only in the genus Polylepion, when compared to others labrids fishes, such as highest numbers of unbranched pectoral-fin rays, procurrent caudal fin rays, and others; probably related to the ancestors of the family Labridae.
The intraspecific distance was zero between the only two samples of P. cruentum tested from Tumbes (Peru) and Baja California (Mexico). This indicates that the Tumbes sample is the same species reported in Baja California, thus extending its distribution range in the Eastern Pacific Ocean (Table 3).
Polylepion cruentum is found over sandy bottoms associated with rocky reefs (150-200 meters of depth). Baldwin et al. (2018) analyzed several fishes that lived in rariphotic zone (below the mesophotic but above the deep aphotic) in the Caribbean Sea: Lipogramma (Grammatidae), Haptoclinus (Labrisomidae), and Decodon plus Polylepion (Labridae). Polylepion sp. from the Caribbean Sea is found in the upper rariphotic zone with 130-189 meters of depth, corroborating the habitat of P. cruentum in the Northern Peru. The extreme depths in which P. cruentum appear to be confined is, undoubtedly, responsible for their rarity in museum collections and the absence of records from other areas throughout its distribution.
The Peruvian record of P. cruentum also may be explained by the tendency of marine species to be sensitive to changing climate and their greater capacity for colonization, mainly by due to biotic and abiotic factors, for example, the increase in the extent of marine currents, extension of the diet distribution, and others. Also in reef fishes, the early life stage serves as the primary opportunity to expand their range and is strongly influenced by warm currents (Pinsky et al., 2020).
Northern Peru is the end of an extensive coastal marine region, known as the Panamic Province (Hastings, 2000) or Guayaquil Ecoregion (Spalding et al., 2007) with predominantly tropical characteristics, where it presents large bodies of water known as Tropical Surface Waters (TSW), with temperatures close to 30℃ and salinities below 33.8 ups; and Equatorial Surface Waters (ESW), with temperatures between 25 and 30℃ and salinities between 33.8 and 34.8 ups (Zuta & Guillen, 1970;Takahashi, 2002). Generally, the TSW and ESW remain north between 04 and 06°S respectively, but when projected to the south they carry low salinity (less than 29 ups) up to approximately 06 and 08°S, especially during El Niño (EN) or the arrival of Kelvin waves (Flores et al., 2013). This might account for the expansion of the distribu- Figure 3. Maximum likelihood tree based on mitochondrial Cytochrome oxidase c subunit 1 for several species of labrids, including the new specimen of P. cruentum. The series of three numbers (e.g., 100) at each of the main nodes represents the percentage of bootstrap support obtained by Maximum Likelihood (ML) analysis, respectively (1,000 pseudoreplicates). Dashes represent values < 50%. tion of P. cruentum to Northern Peru (Acapulco, Tumbes Department).
The appearances of many fish species might be related to the El Niño event, when warm waters flow to neighboring temperate regions (Chirichigno & Velez, 1998). Studying the family Chaenopsidae, Hastings (2000) comments on barriers to the dispersal of rocky reef fishes due to habitat gaps, like wide expanses of pelagic habitat isolating oceanic islands from the mainland, and long stretches of sandy, muddy, often mangrove lined bottoms, separating coastal hard-bottom reef regions. For some species of fishes, these barriers are ineffective, such as Mugil setosus (Britzke et al., 2019) and possibly also for Polylepion cruentum, which inhabits greater depths (150-200 m).
This distribution of P. cruentum in Pacific Ocean is similar to the biogeographic region described by Briggs (1974Briggs ( , 1995, named Tropical Eastern Pacific, from southern Baja California to Northern Peru, including the Galapagos, the Revillagigedo archipelago, Clipperton, Cocos, and Malpelo islands. This region presents many highly endemic species and general species richness, which has been associated with the Gulf of California environment; rise of Panama Isthmus; immigration of species for the north, south and west; and presence of non-effective geographical barriers (Robertson & Cramer, 2009). This distribution area of P. cruentum is limited in the south by the cold Humboldt Current that flows along to coast of Peru towards the south Galapagos Islands (Penven et al., 2005).
Finally, we suggest that the expansion of the distribution range of P. cruetum is due to the input of warm waters in the Northern Peru, especially during El Niño events. Its incidental capture is due to the depths it inhabits which are difficult to access.

CONCLUSION
The registration of P. cruentum in the Northern Peru (Acapulco, Tumbes Department) is important for future checklists of marine fishes found in deep rocky reefs, and consequently to evaluate the conservation status of this species in Peru. Our results also suggest that COI barcoding is useful in corroborating initial identifications based on morphological data.