Astyanax henseli , a new name for Tetragonopterus aeneus Hensel , 1870 from southern Brazil ( Teleostei : Characiformes )

Tetragonopoterus aeneus Hensel (1870) is removed from synonymy with Astyanax fasciatus (Cuvier, 1819), revalidated and redescribed as A. henseli to avoid homonymy. Astyanax henseli differs from A. fasciatus stricto sensu and other Astyanax species by the presence of two to four maxillary teeth, two vertically-elongated humeral spots, dentary tooth cusps positioned close to one another, and the slender form of the dentary teeth.

Through most of the 20 th Century, Tetragonopterus aeneus Hensel continued to be regarded as a junior synonym of A. fasciatus (e.g.Fowler, 1948), or has been treated as a species of uncertain status (e.g.Malabarba, 1989, who listed it as "Astyanax aff.fasciatus", along with T. obscurus).Eigenmann (1921) proposed the following diagnostic attributes of Astyanax fasciatus: aligned predorsal scales, third infra orbital not covering the cheek, elongated umeral spot, 25 or more anal-fin rays, a silvery lateral band becoming black on the caudal peduncle and continued as a black streak to the end of middle fin rays, origin of dorsal fin equally distant from snout to caudal fin and one or more maxillary teeth, rarely three.Recently, however, several authors (Garruti & Britski 2000:84;Lozano-Vilano & Contreras-Balderas, 1990;Schmitter-Soto, 1998;Melo, 2001) questioned the validity of A. fasciatus sensu Eigenmann (1921) and suggested that many species may be included in Eigenmann´s (1921) broad definition of A. fasciatus.Here we refer to these species as the "Astyanax fasciatus species complex".Buckup in Lima et al. (2003:111) recognized the southern Brazilian taxon as a valid species, separate from A. fasciatus.Those authors, however, maintained an association between T. obscurus and T. aeneus Hensel, and used A. obscurus as the valid name.
As part of an ongoing revision of the Astyanax fasciatus species complex, we examined the holotypes of Tetragonopterus aeneus Hensel and Tetragonopterus rutilus and the syntypes of Tetragonopterus aeneus Günther and Tetragonopterus obscurus, and concluded that they represent four different (valid) species.Here we describe Astyanax henseli as a new name to replace A. aeneus (Hensel, 1870), which remains unavailable more than a century after its original publication.

Material and Methods
Counts and measurements follow Fink & Weitzman (1974).Additional measurements include (1) the distance between the dorsal-fin origin and the adipose fin, (2) the distance between the anal-fin origin and the tip of the longest ray (3) the distance between the dorsal-fin origin and the tip of longest ray.For counts recorded in the description of the species the observed range is followed in parentheses by the mean and total number (n) of specimens.Morphometric data for the holotype are presented separately in the tables.All measurements other than standard length (SL) are expressed as a percentage of SL except subunits of the head, which are recorded as a percentage of head length (HL).In all tables, SD stands for 'standard deviation', and n stands for the number of specimens measured.Vertebral counts were taken from radiographs, and cleared and counterstained (c&s) specimens according to Taylor & Van Dyke (1985).This number includes the four vertebrae incorporated in the Weberian apparatus and considers the fused PU 1 +U 1 as a single element.Astyanax henseli is distinguished from A. saguazu, A. elachylepis and members of the Astyanax fasciatus species complex by the autapomorphic presence of a space between the dentary teeth, which are relatively slender, with cusps positioned close to one another (Figs.1-2), and by the presence of a maximum of five cusps in each dentary tooth (vs.maximum of seven cusps).Astyanax henseli is also distinguished from A. fasciatus, A. rutilus, and A. elachylepis by the presence of two to five, usually five, tricuspid teeth, on the anterior margin of maxilla (vs.one to two teeth, usually one, on anterior margin of maxilla).Description.Morphometric data are presented in Table 1.
Body moderately elongate and compressed (Figs.3-4); deepest at dorsal-fin origin.Dorsal body profile convex from snout to dorsal-fin origin, straight along dorsal-fin base, slightly convex between insertion of last dorsal-fin ray and adipose fin.Ventral profile of head and abdomen convex; straight along anal-fin base.Dorsal and ventral profiles of caudal peduncle slightly concave.Snout convex.First infraorbital expanded over small portion of maxilla.Two series of premaxillary teeth; teeth of outer series usually tricuspid, smaller than those of inner series, numbering four to five, usually five; teeth in inner series five with three to seven cusps.Two to five tricuspid teeth on anterior margin of maxilla, usually five teeth (mean=4.7,n=25).Dentary teeth slender, more so in young specimens; anterior four to five teeth larger than remaining teeth; larger teeth with three to five cusps; fifth tooth usually with three cusps; remaining teeth with one or two cusps.Space between dentary teeth, wider in young specimens.
Dorsal-fin rays ii,8-9 (mean=9.0,n=25); first unbranched ray about half the length of second unbranched ray.Dorsalfin origin located just anterior to middle of body.Adipose-fin origin located dorsal to posterior anal-fin ray insertion.Pectoral-fin rays i,11-13 (mean =12.6, n=25); tip of longest ray extending to or posterior to pelvic-fin origin in both males and females.Pelvic-fin rays i,7; pelvic-fin origin anterior to vertical line passing through dorsal-fin origin; longest pelvic ray reaching anal-fin origin in both sexes.Anal-fin rays iiivi, 22-27 (mean=24.3, n=25).Anal-fin origin posterior to vertical line through dorsal-fin origin.
Color in alcohol.Body pale brownish yellow.Lateral body stripe broad and dark posteriorly, becoming pale and narrow anterior to dorsal-fin origin.Lateral body stripe silvery in some specimens.Two dark humeral spots, slanted posteriorly and  slightly wider dorsally; first spot vertically elongate, centered on third and fourth scales of scale row just dorsal to lateral line; second spot, very diffuse, located on first series of scales above lateral line, between seventh to tenth series of scales and extending over two or three horizontal series of scales.Middle caudal-fin rays darkly pigmented with pigmentation extending to tips of rays.Exposed borders of scales delineated by dark chromatophores.Dorsal and anal fins covered with scattered dark chromatophores, except for distinct unpigmented area at tip of dorsal fin and tip of anterior anal-fin lobe.Adipose fin mostly white.Head black to gray dorsally, especially dark near nape.Sides of head and opercles silvery.
Sexual dimorphism.When present, sexual bony hooks distributed over anal-fin and pelvic-fin rays.Pelvic fins with retrorse bony hooks on branched rays and posterior unbranched rays.Anal-fin rays of males with small retrorse bony hooks present from longest unbranched ray to last branched ray; hooks mostly present on posterior branches of rays; one pair of bony hooks per ray segment.Females without fin hooks (gonad examined in specimen MNRJ 14186, 60.8 mm SL).
Distribution.Known from the rio dos Sinos and rio Taquari drainages in Southern Brazil.Rio dos Sinos and rio Taquari are tributaries of the lago Guaíba which is located next to the city of Porto Alegre, and is the northernmost component of the laguna dos Patos system.
Etymology.The specific name henseli is in homage to R. Hensel for his contributions to Ichthyology in Southern Brazil.

Discussion
Due to similarity of overall body shape, A. henseli may be mistakenly identified as A. fasciatus Cuvier (1819) according to the definition of this species proposed by Eigenmann (1921).The original description of A. fasciatus was based on a specimen (MNHN A.8653) collected in the rio São Francisco drainage, which was sent to MNHN by Delalande (Eigenman, 1921).Astyanax fasciatus has been reported from many river drainages from northern Mexico to northern Argentina.Recently, Melo (2005) concluded that the name A. fasciatus should only be applied to specimens from the rio São Francisco drainage; specimens usually identified as A. fasciatus from rio Paraná, eastern Brazil and Central America represent a complex of similar species.Astyanax fasciatus is diagnosed by the presence of an elongated dorsal fin in mature males (Fig. 5).Astyanax henseli, however, may be easily distinguished from populations of A. fasciatus from the rio São Francisco drainage, and other species from the A. fasciatus species complex, based on dentary tooth shape (teeth are wider in A. fasciatus), number of maxillary teeth (only one in A. fasciatus), shape of maxilla (straight maxilla vs. anteriorly curved in A. fasciatus), position of cusps on tooth crown (cusps widely spaced in A. fasciatus), and presence of a second humeral spot.
Astyanax henseli is synpatric with another species of the A. fasciatus complex which we refer to as "Astyanax sp.aff.fasciatus" (Fig. 6, see Comparative Material below).Astyanax sp.aff.fasciatus is widespread in the laguna dos Patos system and the Uruguay drainage.Specimens of Astyanax sp.aff.fasciatus are very similar to specimens of the São Fran- 45.9 42.9-48.946.0 1.4 24 cisco population of A. fasciatus.They differ from the São Francisco populations by usually having 5 [3(1), 4(9), 5(39), 6(1)] teeth in the outer row of the premaxilla, while the latter usually has 4 [3(1), 4(38), 5(4), 6(1)] teeth in the outer row of the premaxilla.Establishing the status of Astyanax sp.aff.fasciatus is beyond the scope of the present publication.However, A. henseli differs from Astyanax sp.aff.fasciatus in the same way it differs from populations of A. fasciatus from the São Francisco drainage, including dentition characters.Astyanax sp.aff.fasciatus and most Astyanax species have the cusps of the dentary teeth more evenly distributed over the relatively wider crown (Fig. 7).Steindachner (1876:17) included A. henseli (listed as Tetragonopterus aeneus Hensel) as a synonym of T. rutilus Jenyns, originally described from the rio Paraná.The latter is currently considered a synonym of A. fasciatus (e.g., Eigenman, 1921;Lima et al., 2003).However, our examination of the holotype of T. rutilus (BMNH 1917.7.14.14, Table 2) suggests that it may be a valid species (it has 37 total vertebrae vs. 35-36 total vertebrae in A. fasciatus, mean=35.7,n=11).Astyanax rutilus differs from A. henseli by the presence of only one maxillary tooth.
Astyanax henseli was listed by Buckup in Lima et al. (2003:111) as a junior synonym of A. obscurus.That listing was based on Steindachner's (1876:20) opinion that T. aeneus Hensel and T. obscurus were conspecific.However, Astyanax henseli is easily distinguished from A. obscurus based on the number of branched anal-fin rays (18-19 vs. 22-27 in A. henseli, Table 2), and the arrangement of dentary teeth.In A. obscurus the arrangement of the teeth is very similar to A. taeniatus Jenyns, A. giton Eigenmann, and A. intermedius Eigenmann.In those species the dentary teeth decrease gradually in size from anterior to posterior portion, unlike most species of Astyanax, which exhibit an abrupt reduction in size between the anterior and the posterior dentary teeth.
In addition to A. henseli and A. obscurus, five additional species of Astyanax from southern Brazil have been previously described: Astyanax jacuhiensis (Cope, 1894), A. eigenmanniorum (Cope, 1894), A. laticeps (Cope, 1894), A. brachypterygium Bertaco &Malabarba, 2001, andA. cremnobates Bertaco &Malabarba, 2001.Astyanax jacuhiensis belongs to Astyanax bimaculatus species-group and is distinguished from A. henseli by the absence of maxillar teeth, and by the shape of the humeral spot, which is horizontally oval.Astyanax laticeps, A. cremnobates and A. brachypterygium differ from A. henseli by their low number of branched anal-fin rays: 18-21, 14-18, and 12-16, respectively (Bertaco & Malabarba, 2001).Astyanax henseli is distinguished from A. eigenmanniorum by the presence of two or more maxillary teeth, in addition to the autapomorphic spacing and shape of the dentary teeth.
It is unclear to us why Hensel (1870) described A. henseli as a new species, but used the name Tetragonopterus aeneus, which was pre-occupied by a Mexican species described a decade earlier by Günther (1860).Hensel (1970) apparently knew about Günther's publication, as evidenced by his own citation of Günther (1860).Regardless of his motives for publishing a homonymy, it is necessary to provide an available name for the species from Porto Alegre.Tetragonopterus aeneus (Günther) is a valid species from Central America currently included in the genus Astyanax (Schmitter-Soto, 1998).In spite of their sharing the same name and generic allocation, the two taxa are not conspecific.Astyanax aeneus (Günther) does not have the diagnostic characters of A.  Table 2. Morphometric and meristic data of the syntypes of A. aeneus (Günther, 1860), BMNH 1860.6.17.41-2, BMNH 1907.4.10.3, and A. obscurus (Hensel, 1870), ZMB 7478, and the holotype of Astyanax rutilus (Jenyns, 1842), BMNH 1917.7.14.14.