A new species of Hyphessobrycon Durbin from northeastern Brazil: evidence from morphological data and DNA barcoding (Characiformes, Characidae)

Abstract A new species of Hyphessobrycon is described for the upper Munim and Preguiças river basins, northeastern Brazil, supported by morphological and molecular species delimitation methods. This new species belongs to the Hyphessobrycon sensu stricto group, as it has the three main diagnostic character states of this assemblage: presence of a dark brown or black blotch on the dorsal fin, absence of a black midlateral stripe on its flank and the position of Weberian apparatus upward horizontal through dorsal margin of operculum. Our phylogenetic analysis also supported the allocation of the new species in this group; however, it was not possible to recover the species sister-group. Pristella maxillaris and Moenkhausia hemigrammoides were recovered as the sister-clade of the Hyphessobrycon sensu stricto group.


Introduction
Hyphessobrycon Durbin, 1908 is one of the most species-rich genera of Characidae, currently comprising approximately 150 valid species (Ohara et al. 2017). It is widely distributed along the river basins of the Neotropical region, from southern Mexico to the La Plata River basin in northeastern Argentina (Carvalho and Malabarba 2015, Teixeira et al. 2015, Garcia-Alzate et al. 2017, with highest diversity in the Amazon basin (Miquelarena and López 2006, Lima et al. 2013, Bragança et al. 2015, Carvalho and Malabarba 2015, Marinho et al. 2016, Carvalho et al. 2017, Moreira and Lima 2017.
Extensive data show that Hyphessobrycon is not a monophyletic group (Carvalho et al. 2017, Moreira and. It was diagnosed by an artificial combination of character states proposed by Eigenmann (1917), such as: the presence of an adipose fin; maxillary with few teeth or none; lateral line incomplete; third infraorbital bone not in contact with the sensory canal of the preopercle; premaxillary with two series of teeth; and caudal-fin lobes without scales at the base. Nevertheless, new species descriptions continue to follow this artificial combination (e.g., Ohara et al. 2017, Garcia-Alzate et al. 2017, Carvalho et al. 2017, Moreira and Lima 2017. In addition, some artificial species groups of Hyphessobrycon were proposed based on the combination of character states (e.g., Weitzman and Palmer 1997, Zarske 2014, García-Alzate et al. 2008a, b, 2013, Carvalho and Malabarba 2015 relying mainly on coloration patterns. However, in many cases, it is not possible to assign without reasonable doubt to which group a particular species belongs (Bragança et al. 2015).
The key point is that the remaining species of Hyphessobrycon included in the other groups will probably need to be assigned to other genera or new genera (Hyphessobrycon sensu lato) (Carvalho and Malabarba 2015).
One way to overcome the confusing taxonomy of problematic groups, to have accurate species identifications and species diversity estimates of groups is to use different operational criteria for species delimitation (Goldstein andDesalle 2010, Padial et al. 2010). Any operational criteria (species delimitation methods) may separately provide evidence about the species limits and identity independently from other criteria (de Queiroz 2005(de Queiroz , 2007, but evidence corroborated from multiple operational criteria is considered to produce stronger hypotheses of lineage divergence (de Queiroz 2007, Goldstein andDesalle 2010), converging to the proposal for an integrative taxonomy (Goldstein andDesalle 2010, Padial et al. 2010). Gathering morphological and molecular data has become a common practice to identify and delimit species of fish (Teletchea 2009), mainly in groups including cryptic or morphologically similar species. The most widespread molecular method used in taxonomy has been the DNA barcoding, which consists on the use of a single gene from mitochondrial DNA (cytochrome oxidase subunit I -COI) as a proxy for species differentiation (Hebert et al. 2003). In fact, several studies have been carried out using molecular markers and new species have been delimited and/or described, in most cases, based both on molecular and morphological evidence (e.g., Costa and Amorim 2011, Costa et al. 2012, Roxo et al 2012, Villa-Verde et al. 2012, Castro-Paz et al. 2014, Costa et al. 2014, Benzaquem et al 2015, Mattos et al. 2015, Costa et al. 2017.
A new species of Hyphessobrycon, member of the Hyphessobrycon sensu stricto Carvalho and Malabarba, 2015 is herein described from the Munim and Preguiças river basins, two coastal river basins of the Maranhão State, northeastern Brazil, based on both morphology and molecular data.

Morphological analysis
Measurements and counts were made according to Fink and Weitzman (1974), with exception for the scale rows below lateral line, which were counted to the insertion of pelvic fin. Horizontal scale rows between the dorsal-fin origin and lateral line do not include the scale of the median predorsal series situated just anterior to the first dorsalfin ray. Counts of supraneurals, vertebrae, procurrent caudal-fin rays, unbranched dorsal and anal fin rays, branchiostegal rays, gill-rakers, premaxillary, maxillary, and dentary teeth were taken only from cleared and stained paratypes (C&S), prepared according to Taylor and Van Dyke (1985). The four modified vertebrae that constitute the Weberian apparatus were not included in the vertebra counts and the fused PU1 + U1 was considered as a single element. Osteological nomenclature follows Weitzman (1962). Institutional abbreviations follow Sabaj-Pérez (2016), with addition of CICCAA Coleção Ictiológica do Centro de Ciências Agrárias e Ambientais and CPUFMA Coleção de Peixes da Universidade Federal do Maranhão.

Comparative material examined
All specimens are from Brazil.

DNA extraction, amplification, and sequencing
DNA extraction was carried out with the Wizard Genomic DNA Purification kit (Promega) following manufacturer's protocol. DNA quality was evaluated by agarose gel electrophoresis stained with GelRed (Biotium) and was quantified using Nanodrop 2000 (Thermo Fisher Scientific). DNA was stored at -20 °C until further procedures. Samples (N= 4; Table 1) were amplified using standard PCR (Polymerase Chain Reaction) for partial cytochrome oxidase subunit 1 (COI) gene, with primers designed by Ward et al. (2005) (FISHF1 5´-TCAACCAACCACAAAGACATTGGCAC-3´and FISHR1 5´-TAGACTTCTGGGTGGCCAAAGAATCA-3´). Amplification reactions were performed in a total volume of 15 µl comprising 1× buffer, 1.5 mM MgCl 2 , 200 µM dNTP, 0.2 uM of each primer, 1 U of Taq Polymerase (Invitrogen), 100 ηg of DNA template, and ultrapure water. The amplification program consisted of a denaturation of 2 min at 94 °C, followed by 35 cycles of 30s at 94 °C, 30s at 54 °C, and 1 min at 72 °C, ending in an extension phase of 10 min at 72 °C. Amplicons were visualized in 1% agarose gel electrophoresis stained with GelRed (Biotium) and purified with Illustra GFX PCR DNA and Gel Purification Kit (GE Healthcare). Samples were sequenced using both forward and reverse primers and BigDye Terminator kit 3.1 Cycle Sequencing kit (Thermo Fisher Scientific) in ABI 3730 DNA Analyser (Thermo Fisher Scientific) Consensus sequences were edited in Geneious 9.0.5 (Kearse et al. 2012) and aligned using ClustalW (Thompson 1994) with those from Hyphessobrycon species available in Barcode of Life Database (BOLD) and Genbank (NCBI-National Center for Biotechonology Information) (accession numbers are in Table 1).

Species concept, species delimitation, and diagnoses
The unified species concept (de Queiroz 2005(de Queiroz , 2007 is herein adopted by expressing the conceptual definition shared by all traditional species concepts -"species are (segments of ) separately evolving metapopulation lineages" -when operational criterion elements to delimit taxa are excluded from the concepts. According to this concept, species are treated as hypothetical and could be tested by the application of distinct criteria (species delimitation methods) (de Queiroz 2005(de Queiroz , 2007. It allows for any criteria to separately provide evidence about species limits and identities, independently from other criteria (de Queiroz 2005(de Queiroz , 2007. Evidence corroborated from multiple operational criteria is considered to produce stronger hypotheses of lineage separation (de Queiroz 2007, Goldstein and Desalle 2010), a practice called "integrative taxonomy" (Dayrat 2005, Goldstein and Desalle 2010, Padial et al. 2010. Two distinct operational criteria to delimit species, based on morphological and molecular data, were implemented here. The Population Aggregation Analysis (Davis and Nixon 1992) is a character-based method (hereafter PPA), which consists of an exclusive shared combination of character states assigned to a given population or group of populations. The second method, DNA barcoding, as proposed by Hebert et al. (2003Hebert et al. ( a, b, 2004) (hereafter DBC), is a genetic distance-based cut-off method.

Population Aggregation Analysis (PAA)
Only morphological character states were used for this method. The morphological data was based on both examined material (see Comparative material examined) and the literature (e.g., Géry 1977, Géry and Uj 1987, Costa and Géry 1994, Plaquete et al. 1996, Weitzman and Palmer 1997, Zarske 2008, Hein 2009, Carvalho 2011, Lima et al. 2013, Zarske 2014, Carvalho and Malabarba 2015, Carvalho et al. 2017. The data obtained by this method are presented in the diagnosis section of results.

DNA barcoding (DBC)
Pairwise genetic distances between species were calculated using Kimura-2-parameters model (K2P) (Kimura 1980) on the MEGA 7 software (Tamura et al. 2011). Evolutionary relationships among sequences were reconstructed by Bayesian inference using the MrBayes (Huelsenbeck and Ronquist 2001) plugin in Geneious 9.0.5. An independent run with a chain length of 10 million, a burn-in length of 500,000 generations, and subsampling trees every 10,000 generations was carried out under the GTR (generalized time reversible) evolutionary model, which was estimated in jmodeltest (Darriba et al. 2012). Hyphessobrycon flammeus Myers, 1924 and H. anisitsi (Eigenmann, 1907) were used as outgroup. The ingroup was composed by the remaining terminals.  The new species differs from H. bentosi by the absence of an extended and pointed dorsal and anal-fin tips (Figures 1, 2) [vs. extended and pointed dorsal and anal-fin tips]; and from H. hasemani by the dorsal-fin black spot shape, which is located approximately at the middle of the fin's depth, not reaching its tip [vs. extended along all the fin, reaching its tip in adults] and by presenting tri to unicuspid teeth in the inner row of premaxillary and dentary [vs. pentacuspid teeth].

Color in alcohol
Dorsal fin ground coloration hyaline, with a conspicuous black or dark brown spot located on anterior portion of fin, reaching about sixth ray, approximately between half to two thirds of fin depth. Anal and caudal fins hyaline. Caudal fin with a darker, usually dark brown, posterior margin and on its base. Adipose fin hyaline to light brown, with dark brown or black chromatophores more concentrated on its dorsal portion, depending on the state of preservation of the specimen. Pectoral and pelvic fins hyaline; pelvic fin with variable amounts dark brown pigmentation remaining depending on the state of preservation of the specimen.
Color in life ( Figure 2). Pattern similar to coloration of preserved specimens. Ground coloration light yellowish brown to grey, with a reddish-brown pigmentation on vertebrae region, and usually with red chromatophores. Ventral region anterior to anal-fin origin lighter. Humeral spot inconspicuously dark brown or black. Head with same coloration as body, and ventrally lighter.
Conspicuous black spot on dorsal-fin, with yellow or orange pigmentation on dorsal and ventral margins of spot; yellow or orange pigmentation lighter and less evident on dorsal margin, reaching half to two thirds of the spot length and extending to the tip of fin; yellow or orange pigmentation darker and more developed at ventral margin of the spot, reaching entire spot base length, not extended to dorsal fin-base. Rest of dorsal fin hyaline. Anal-fin base with red pigmentation, with different degrees of intensity, with milk white pigmentation on anterior tip of anal fin, which could be extended through entire anterior margin, reaching between second to fourth rays. Posterior margin of anal fin with an inconspicuous dark brown pigmentation. Adipose fin light brown to hyaline at base, with red to black pigmentation at tip. Pectoral and pelvic fins hyaline, with some sparser dark brown chromatophores, more concentrated at pelvic fin base. First ray of pelvic fin with a white pigmentation. Caudal fin with red pigmentation on almost fin, with an inconspicuous light brown, reddish brown or dark brown margin.
Sexual dimorphism. Mature males have hooks on anal-fin and pelvic-fin rays. Hooks absent on females. Anal-fin presenting hooks from 3 rd , 4 th or 5 th rays through last ray. Number of hooks variable, increasing from the first ones to the last rays. Pelvic fin presenting 3 rd and 4 th rays with 5 smaller hooks (Figures 4, 5).
DNA-based identification. After trimming sequence ends with poor base call quality, the final alignment yielded 446 base pairs with 154 variable sites, and 22 haplotypes. The magnitude of sequence divergence clearly demonstrates the exist-ence of a new species of Hyphessobrycon inhabiting the Munim and Preguiças river basins in Maranhão State. Average genetic distances were 14.2%, with the highest values between H. pyrrhonotus and H. epicharis (19.2%), while the lowest value (2.7%) was between H. epicharis and H. sweglesi (Table 1). Hyphessobrycon piorskii sp. n. has 17% sequence divergent, on average, from the other taxa, with a minimum distance with H. eques (13.9%) and a maximum with H. rosaceus (18.4%) ( Table 2).
Etymology. The name piorskii honors the ichthyologist Nivaldo Magalhães Piorski for his contributions to the ichthyologic knowledge of the Maranhão State.

Discussion
Despite Hyphessobrycon, as defined today, being a non-monophyletic group (Mirande 2010, Oliveira et al. 2011, Carvalho et al. 2017, Ohara et al. 2017, Moreira and Lima 2017), a few putative groups within the genus were proposed in the literature. One such case is the Hyphessobrycon sensu stricto as defined by Carvalho (2011) andCavalho andMalabarba (2015). According to those authors, this group is composed by approximately 25 species. Among the species considered by Weitzman and Palmer (1997) as possibly related to the "rosy tetra clade", only H. hasemani and H. pulchripinnis were considered to belong to Hyphessobrycon sensu stricto (Carvalho 2011, Carvalho andMalabarba 2015). Pristella maxillaris (Ulrey, 1894) is the sister-group of the Hyphessobrycon sensu stricto (Carvalho 2011), and corroborated in our analysis ( Figure 6).
Hyphessobrycon piorskii sp. n. exhibits all the diagnostic features that define Hyphessobrycon sensu stricto (see introduction and diagnosis section). The new species differs from the other possible species of this assemblage, which also occur near Maranhão (e.g., lower Amazon River basin, Guamá River basin, and São Francisco River basin), such as H. bentosi, H. copelandi, H. eques, H. dorsalis, H. hasemani, H. haraldschultzi, H. micropterus, and H. werneri, by a set of features listed below.
Hyphessobrycon piorskii sp. n. possesses an inconspicuous vertically elongated humeral spot, distinguishing it from all the species cited above, except for H. bentosi and H. hasemani (see morphological diagnosis section). The shape of the dorsal-fin spot is also useful to distinguish H. piorskii sp. n. from H. eques, H. hasemani and H. micropterus, which possess dorsal fin spot vertically extended, reaching the tip of the fin, while in H. piorskii sp. n. the black spot of dorsal fin never reaches the tip of the fin. The new species also differs from H. eques by the color pattern of the anal fin: H. eques possess a conspicuous black anal-fin margin on preserved species, while H. piorskii sp. n. does not exhibit this feature at the anal fin.
The number of teeth cusps was also revealed to be a useful feature for species discrimination. Hyphessobrycon piorskii sp. n. possess all of its teeth with one to three cusps (never pentacuspid), while H. eques possess pentacuspid teeth on the maxillary and inner row of premaxillary, and H. copelandi and H. hasemani on the dentary and inner row of the premaxilla (see Lima et al. 2013). The new species differs from H. bentosi by not having extended and pointed dorsal and anal-fin tips and by having bone hooks on anal-fin rays of mature males (Figure 3). The dorsal and anal fins of H. bentosi have pointed and extended tips, and it has not bony hooks on anal-fin rays (see Carvalho 2011, Zarske 2014. Hyphessobrycon copelandi possesses only ten teeth on the dentary, and dorsal-fin black spot reaching to the posterior margin of the fin (see Lima et al. 2013), while H. piorskii sp. n. possesses 11-15 teeth on dentary, and dorsal-fin black spot restricted to the anterior half of the fin's length. In addition, Hyphessobrycon piorskii sp. n. is easily distinguished from the sister-species of the clade Hyphessobrycon sensu stricto, P. maxillaris and M. hemigrammoides, by the absence of a black oblique stripe or band on the anterior portion of the anal fin (Figures 1, 2) [vs. presence (Carvalho et al. 2017, figure 7; pers. obs.)].  The description of H. piorskii sp. n. was based on morphological and molecular species delimitation methods, using the congruence of multiple operational criteria for determining species boundaries. As mentioned earlier, evidence corroborated from multiple operational criteria is considered to produce stronger hypotheses of lineage divergence (de Queiroz 2007, Goldstein andDesalle 2010), thus congruent to the proposal for an integrative taxonomy (Goldstein andDesalle 2010, Padial et al. 2010). The morphological criteria (PAA) distinguished the new species from all of the other congeners by unambiguous character states (see diagnosis). The DNA barcoding (DBC) criteria also revealed that H. piorskii sp. n. is a new species with an average sequence divergence of 17% from the other taxa (Table 2). In addition, H. piorskii sp. n. is placed in an exclusive and highly supported clade in the Bayesian tree ( Figure 6). Haplotypes clustered as an exclusive and high supported group, with geographical concordance area is evidence of lineage divergence, therefore a good and strong evidence for delimit species, and consequently describe them (Wiens andPenkrot 2002, Costa et al. 2014).
Our Bayesian tree also recovered H. piorskii sp. n. within the Hyphessobrycon sensu stricto group with high support (posterior probability = 0.94), which fits the morphological evidence, since H. piorskii sp. n. exhibits the three main diagnostic character states of the group (see introduction and diagnosis section). Hyphessobrycon piorskii sp. n. was recovered as the sister-group of the clade including H. bentosi, H. socolofi, H. megalopterus, H. erythrostigma, and H. pyrrhonotus, however this relationship was supported by a lower support value (posterior probability value = 0.55). Only posterior probability values about or higher than 0.95 are considered as statistically significant (Alfaro and Holder 2006). Therefore, any discussion about the relationship and supposed shared morphological features between H. piorskii sp. n. and this clade is speculative (Figure 7). To a better understanding of the internal relationships of the group, an analysis including more genes, especially from nuclear genome, is highly recommended. However, this was not the scope of the present paper. Pristella maxillaris and Moenkhausia hemigrammoides were recovered as the sister-clade of the Hyphessobrycon sensu stricto group, corroborating partially the results of Carvalho et al. (2011) and Carvalho and Malabarba (2015), who argue that P. maxillaris is the sister-clade of the Hyphessobrycon sensu stricto group.