Hisonotus acuen, a new and phenotypically variable cascudinho (Siluriformes, Loricariidae, Hypoptopomatinae) from the upper rio Xingu basin, Brazil

Abstract A new species of Hisonotus is described from the headwaters of the rio Xingu. The new species is distinguished from its congeners by having a functional V-shaped spinelet, odontodes not forming longitudinal aligned rows on the head and trunk, lower counts of the lateral and median series of abdominal figs, presence of a single rostral fig at the tip of the snout, absence of the unpaired figlets at typical adipose fin position, yellowish-tipped teeth, absence of conspicuous dark saddles and stripe on the body and higher number of teeth on the premaxillary and dentary. The new species, Hisonotus acuen, is restricted to headwaters of the rio Xingu basin, and is the first species of the genus Hisonotus described from the rio Xingu basin. Hisonotus acuen is highly variable in aspects of external body proportions, including body depth, snout length, and abdomen length. This variation is partly distributed within and among populations, and is not strongly correlated with body size. PCA of 83 adult specimens from six allopatric populations indicates the presence of continuous variation. Therefore, the available morphological data suggest that the individuals inhabiting the six localities of rio Xingu represent different populations of a single species. Low intraspecific variation in mitochondrial Cytochrome oxidase subunit I (COI) provides corroborative evidence.


Introduction
The subfamily Hypoptopomatinae is a monophyletic group of Loricariids (Schaefer, 2003) composed of 19 genera and 139 species (Eschmeyer and Fong 2014). Within this subfamily, Hisonotus Eigenmann & Eigenmann, 1889 comprises 33 valid species (Eschmeyer 2014). The genus Hisonotus was resurrected from the synonymy of Otocinclus by Schaefer (1998) based on the reduced or absent snout plates anterior to the nostril, rostrum with enlarged odontodes, and thickened plates forming the lateral rostral margin. However, the phylogenetic relationships in this genus are not well resolved (Britski and Garavello 2007) and, according to molecular (Chiachio et al. 2008;Cramer et al. 2011) and morphological (Martins et al. 2014) data, Hisonotus is a polyphyletic genus.
Although there is no definition of Hisonotus that supports its monophyly, many authors have considered this genus as valid. In the past decade, 18 species of Hisonotus have been described (Britski and Garavello 2007;Carvalho et al. 2008;Carvalho and Reis 2009;Carvalho and Reis 2011;Martins and Langeani 2012;Carvalho and Datovo 2012;Roxo et al. 2013;Roxo et al. 2014). Recently, during a collecting trip in tributaries of the rio Xingu basin, we found fish specimens that have the generally accepted characteristics of Hisonotus listed above but do not match any known species. Herein we describe the rio Xingu specimens as a new species.

Material and methods
All measurements and counts were taken on the left side of specimens. Measurements were taken point to point to the nearest 0.1 mm with a digital caliper. Body plate and osteology nomenclature followed Schaefer (1997) and measurements followed Carvalho and Reis (2009), as shown in Table 1. Abbreviations used in the text followed Carvalho and Reis (2009). Morphometrics are given as percentages of standard length (SL), except for subunits of the head region, which are expressed as percentages of head length (HL). Specimens were cleared and stained (c&s) according to the method of Taylor and Van Dyke (1985). Vertebral counts also include the five vertebrae that comprise the Weberian apparatus. Dorsal-fin ray counts include the spinelet as the first unbranched ray. All examined specimens were collected according to the Brazilian laws, and are deposited under permanent scientific collection licenses. After collection, specimens were euthanized using 1% benzocaine in water, fixed in 10% formaldehyde for morphological studies and preserved in 70% alcohol. For molecular studies specimens were fixed directly in 95% alcohol. Sequencing and molecular analysis followed Roxo et al. (2012). Institutional acronyms follow Fricke and Eschmeyer (2014

Principal component analysis (PCA)
Principal component analysis (PCA) was used to check overall variation among samples, including differences in morphometrics among species. PCA is a statistical procedure that uses orthogonal transformation to convert a set of observations of possibly correlated variables into a set of values of linearly uncorrelated variables called principal components (Jolliffe 2002). The analyses were made using all measurements listed above. Juvenile specimens below 18.0 mm SL were excluded from the analyses. PCA on covariances of base 10 logarithmically transformed measurements to reduce the influence of size were obtained using Past version 1.28 (Hammer et al. 2004). The PCA Loadings are presented in Table 2.     Fig. 3b), by having the caudal-fin color pattern mostly hyaline, except for dark blotch on origin of rays, and dark brown chromatophores largely concentrated on rays near lower caudal spine, Fig. 3c (vs. caudal-fin mostly dark brown with chromatophores largely concentrated on rays and membranes, and with two hyaline spots on middle of the fin, Fig. 3d); from H. paresi by the absence of conspicuous dark dorsal saddle and longitudinal stripe on the body (vs. inconspicuous dark saddles and stripe of the body) and from H. insperatus by the higher number of premaxillary (14−27 vs. 6−12) and dentary teeth (12−23 vs. 5−11).

Hisonotus acuen
Description. Morphometric and meristic data presented in Table 1. Maximum body length 29.0 mm SL. Dorsal profile of head in lateral view convex to straight from upper part of rostrum to posterior margin of nares, slightly curved from eyes to posterior margin of parieto supraoccipital, almost straight to dorsal-fin origin. Dorsal profile of trunk almost straight, descending from base of dorsal-fin origin to caudal peduncle. Ventral profile slightly concave from snout tip to anal-fin origin, slightly convex to caudal peduncle. Greatest body depth at dorsal-fin origin (13.5−22.8% SL). Greatest body width at cleithral region, gradually decreasing towards snout and caudal  g- fin. Cross-section of caudal peduncle almost ellipsoid; rounded laterally and almost flat dorsally and ventrally.
Head rounded in dorsal view. Snout slightly pointed, its tip rounded, elongated (34.2−57.2% HL) and depressed in front of each nostril on dorsal surface. Dorsal and ventral series of odontodes completely covering anterior margin of snout; odontodes of snout similar in size to remaining ones found on head. Snout tip lacking band devoid of odontodes. Odontodes on head and trunk well defined and not forming longitudinal rows. Usually no tufts or crests of odontodes on head, in some juvenile specimens, a tiny tuft of odontodes at posterior tip of supraoccipital. Eyes small (11.2−16.2% HL), dorsolaterally positioned. Lips roundish and papillose; papillae uniformly distributed on base of dentary and premaxilla and slightly decreasing in size distally. Lower lip larger than upper lip; its border fringed. Maxillary barbel present. Teeth slender and bicuspid; mesial cusp larger than lateral cusp. Premaxillary teeth 14−27. Dentary teeth 12−23.
Dorsal-fin ii,7; dorsal-fin spinelet short and V-shaped; dorsal-fin lock functional; its origin slightly posterior to pelvic-fin origin. Tip of adpressed dorsal-fin rays slightly surpassing end of anal-fin base. Pectoral-fin i,6; tip of longest pectoral-fin ray almost reaching half of pelvic-fin length, when depressed. Pectoral axillary slit present be- tween pectoral-fin insertion and lateral process of cleithrum. Pectoral spine supporting odontodes anteroventrally. Pelvic-fin i,5; its tip not exceeding anal-fin origin when depressed in both sex. Pelvic-fin unbranched ray with dermal flap along its dorsal surface in males. Anal fin i,5; its tip reaching 7th and 8th plate from its origin. Caudal-fin i,14,i; distal margin forked. Adipose-fin absent. Total vertebrae 27.
Body covered with bony plates except on ventral part of head, around pectoral and pelvic-fin origin and on dorsal-fin base. Cleithrum and coracoid totally exposed. Arrector fossae partially enclosed by ventral lamina of coracoids. Abdomen entirely covered by plates in adults (about 23.0 mm SL); lateral plate series with elongate and large plates, formed by two lateral plate series, similar in size; median plates formed by four to five irregular plate series reaching anal shield (Fig. 4a). Lateral side of body entirely covered by plates; mid-dorsal plates poorly developed, reaching middle of dorsal-fin base; median plates not interrupted in median portion of body, but with 2 or 3 plates not perforated before end of series; mid-ventral plates exceed end of anal-fin base.
Color in alcohol. Large inconspicuous brown lateral stripe extending from tip of snout through inferior orbit to end of caudal peduncle Fig. 5a, e (very weak in some specimens, such as holotype Fig. 5c). Body ground color brown on dorsum, yellowish on ven-  tral region under lateral stripe. Some specimens with dark saddle on mid-ventral to ventral portion of body (Fig. 5b). Dorsal, pectoral, pelvic and anal fins with brown dots on rays, varying in concentration of chromatophores from one individual to another; inter-radial membranes hyaline. Caudal fin hyaline, except for dark blotch on origin of rays, and dark brown chromatophores largely concentrated on rays near lower caudal spine (Fig. 5d). In some specimens, chromatophores forming two dark bands on middle of rays (Fig. 5b, e).
Color in life. Similar to pattern described for alcohol individuals, but with ground color light brown (Fig. 6).
Sexual dimorphism. Males bear a papilla posterior to urogenital opening and present the pelvic-fin unbranched ray with dermal flap along its dorsal surface. Both characters are absent in females.
Distribution. Hisonotus acuen is known from small to median-sized streams of the upper rio Xingu basin, Mato Grosso State in Brazil (Fig. 7a).
Habitat. Hisonotus acuen was collected on flat areas in creeks of headwaters of the rio Xingu basin in places of shallow clear waters with low current. The fishes are found associated with vegetation that covers the bottom and the border of the headwaters (Fig. 7b).
Etymology. The specific name "acuen" is in reference to the Xavante indigenous peoples, who in anthropological literature are known as "acuen". These people are constituted by the natives inhabiting the east of the Mato Grosso State, living in the margins of the rivers Culuene, Xingu, Mortes and Araguaia.

Discussion
The new species has a functional V-shaped spinelet (Fig. 2a-f ). Carvalho and Datovo (2012) first reported this structure in H. bockmanni, H. chromodontus, H. insperatus and H. luteofrenatus. Subsequently, Roxo et al. (2014) reported this character in H. oliveirai and H. paresi. The functional V-shaped spinelet (Fig. 2a-f ) is a putative apomorphic character within Hisonotus, and may distinguish a monophyletic group within the genus (Carvalho and Datovo 2012). However, Martins et al. (2014) have a different interpretation, in which the spinelet in H. chromodontus, H. luteofrenatus and H. piracanjuba is reduced, and the locking mechanism is not functional (Martins et al. 2014, Fig . 11A), the same character state found in H. armatus, H. depressicauda, H. francirochai and H. notatus. Martins et al. (2014) also suggested that in H. insperatus the spinelet is absent (Martins et al. 2014, Fig. 11b). However, in our interpretation, H. acuen, H. chromodontus, H. bockmanni, H. insperatus, H. luteofrenatus, H. oliveirai, H. paresi, and H. piracanjuba exhibit a functional V-shaped spinelet, which is not present in H. depressicauda and H. notatus (Fig. 2). Therefore, despite the fact that the genus Hisonotus may not represent a monophyletic unit, we include H. acuen within Hisonotus pending a formal phylogenetic analysis of Hypoptopomatinae, and the species-level composition is established.
Hisonotus acuen exhibits a large amount of variation in external body proportions across its range (Fig. 5), especially in body depth at dorsal-fin origin (13.5-22.8% of SL), snout length (34.2-57.2% of HL), and abdomen length 10.2-24.4% of SL). This variation is partly distributed within populations, and partly between populations, and is not strongly correlated with body size. We performed a PCA to evaluate the morphometric variation within this new species. We compared the morphometric data of six populations found in different tributaries of the rio Xingu, and our results suggest that the range in morphology has a continuous distribution. The lack of phenotypic discontinuities among populations suggests they are not different species (Fig. 8). Additionally, we found that the genetic variation of the Cytochrome Oxidase I (COI) gene within the populations of H. acuen is 1%, and that variation among closely related congeners (i.e. H. chromodontus, H. insperatus, H. oliveirai, H. paresi and H. piracanjuba) is more than 17% (see Table 3 and Fig. 9; sequences can  be downloaded from GenBank using the accession numbers provided in Methods). Therefore, the available morphological and molecular data support the recognition of the individuals inhabiting the six localities of rio Xingu and representing different populations as a single species.
The new species H. acuen is the first described species of Hisonotus from the rio Xingu basin, and is externally very similar to H. chromodontus, a species from the rio Tapajos basin. The coloration of the caudal fin and the tip of the teeth distinguish these species that also are very different genetically (i.e. 19.3% of genetic divergence; Table 3 and Fig. 9). Britski and Garavello (2003) discussed the coloration of the teeth of H. chromodontus, reporting that in more than one hundred specimens examined, varying from 12.0 to 32.2 mm SL, all tooth-tips have a reddish-brown color. We analyzed more than one hundred specimens of H. chromodontus from the museum collections of LBP and NUP, and found the same reddish-brown tooth tips. This tooth features appears to be unique within the genus Hisonotus. A very similar external morphology, as well as the presence of the functional V-shaped spinelet among H. acuen and H. chromodontus, could suggest a close relationship between these species.