Systematics of Spintherobolus (Teleostei: Characidae: Cheirodontinae) from eastern Brazil Author Weitzman, Stanley H. Division of Fishes, Department of Vertebrate Zoology, MRC- 159, Smithsonian Institution, Washington D. C. 20560, USA. Author Malabarba, Luiz R. Laboratório de Ictiologia, Museu de Ciências e Tecnologia, Pontifícia Universidade Católica do Rio Grande do Sul, Av. Ipiranga, 6681, CEP 90619 - 900, Porto Alegre, RS, Brazil, and Departamento de Zoologia, IB, Universidade Federal do Rio Grande do Sul, Porto Alegre, Brazil. text Ichthyol. Explor. Freshwaters 1999 1999-01-01 10 1 1 43 journal article 10.5281/zenodo.10881992 0936-9902 10881992 DB264AEA-C4EC-48B7-BE35-56761CA13225 Spintherobolus broccae Myers ( Figs. 2 , 32-35 ) Spintherobolus broccae Myers, 1925: 143 , 2 figs. (new species description, type locality: hills behind Rio de Janeiro , holotype : CM 7979a (now FMNH 58863 ), one paratype : CM 7980a , five additional spms in author’s collection, # 56, three of these are USNM 92959 , but 1 spm missing; see also CAS-SU 24073 below) Specimens examined. USNM 92959 , 2 males , paratypes , 20.0- 20.3 mm SL; near Rio de Janeiro (city), Rio de Janeiro , Brazil ; 1924 , R. Brocca . - CAS-SU 24073 , 1 male , 20.3 m SL; paratype ; Brazil , Rio de Janeiro , near Rio de Janeiro ; 1924 ; R. Brocca & Salathe , 1924. - USNM 342062 , 1 male , 19.2 mm SL, 3 females , 21.5-22.1 mm SL; 5 km East of Sampaio Correa , Saquarema ; 27 Nov 1979 , C. Cruz , S. Weitzman et al. - USNM 287324 , 1 male , c&s, 18.2 mm SL, 2 females , c&s 18.3- 19.0 mm SL; 13 males , 17.1-18.9 mm SL, 27 females , 17.5-20.9 mm SL; MCP 19196 , 4 males , 17.5-17.9 mm SL, 3 females , 17.4-18.9 mm SL; rio Macacu , near town of Cachoeiras de Macacu , small tributary about 1 -2 km from town, highway bridge over stream; 27 Nov 1979 , C. Cruz , S. Weitzman et al. - MNRJ 8796 , 3 males , 18.1 - 18.4 mm SL, 3 females , 18.0- 18.5 mm SL, 6 immature , 13.6-17.4 mm SL; swamp or marshy area near Imbarié , Baixada Fluminense ; 4 Aug 1954 , L. Travassos et al. - MNRJ 6264 , 62 , adults + immature, 14.6-18.8 mm SL; old road to Petrópolis , about 1-2 km from new road to Petrópolis, Raiz da Serra , Rio de Janeiro ; 28 Sep 1942 , L. Travassos et al. - USNM 342064 , 13 males , 18.6- 22.4 mm SL, 10 females , 20.1 -22.9 mm SL, 1 (c&s), 19.4 mm SL; Estacion Ecologia , Universidade Federal Rural do Rio de Janeiro , at km 43 on the road from Rio de Janeiro to Sao Paulo ; 15 Dec 1972 , E. Izecksohn et al. - USNM 297997 , 3 females , 20.8-25.6 mm SL, 21 young males , 16.4- 18.8 mm SL, 214 immature , 13.0- 20.1 mm SL; unnamed clear water stream crossing road SP-98 at km 94 , northwest of Bertioga ; 23 Feb 1988 , R. M. C. Castro . Fig. 34. Spintherobolus broccae, USNM 92959, paratype, adult male 20.3 mm SL, Brazil, vicinity of Rio de Janeiro. Fig. 35. Spintherobolus broccae, USNM 297997, adult female, SL uncertain, Brazil, S ã o Paulo, Bertioga. Diagnosis. This miniature species lacks infraor ­ bital bone 3. The other miniature species, S . ankoseion and S . leptoura, as well as the non-miniature S . papilliferus have the third infraorbital bone well ossified and, in addition, have significantly lower anal-fin ray counts ( Fig. 20 ) and ventral procurrent caudal-fin ray counts ( Fig. 23 ). Although statistically significant, there is overlap in these counts (see counts given below), especially in the ventral procurrent caudal-fin ray count, and species determination using these counts must be made by examining population samples, not individual specimens. However, specimens bearing overlapping anal-fin ray counts are uncommon. Table 4.Morphometriesof Spintherobolus broccae . Table includes measurementsofspecimens from USNM USNM 342062, USNM342064and MCP19196.

males				females
n	low	high	mean	n	low	high	mean
Standard length [mm]	31	17.1	22.4	18.8	40	17.5	22.9	19.8
Percentage of standard length
Snout to anal-fin origin	31	55.9	61.5	59.7	40	57.8	63.9	60.5
Snout to dorsal-fin origin	31	51.3	59.5	56.2	40	52.6	60.6	55.7
Snout to pelvic-fin origin	31	36.2	44.7	42.2	40	39.8	44.7	42.1
Dorsal-fin base length	31	9.0	14.0	10.8	40	9.1	12.0	11.0
Anal-fin base length	31	18.8	26.7	23.4	40	19.9	26.5	23.7
Caudal peduncle length	31	12.4	19.1	15.7	40	13.1	17.7	15.4
Caudal peduncle depth	31	12.6	15.1	13.8	40	11.2	14.9	13.1
Depth at dorsal-fin origin	31	29.2	32.8	30.5	40	28.1	35.1	30.8
Dorsal-fin length	31	22.4	29.9	26.9	39	22.1	30.3	26.1
Pelvic-fin length	31	16.4	20.7	18.4	38	13.5	19.5	16.9
Pectoral-fin length	29	15.2	18.7	16.9	38	12.9	18.3	15.6
Bony head length	31	25.4	30.4	28.6	38	25.6	30.3	28.3
Percentage of head length
Snout length	31	17.0	23.6	20.3	39	17.6	21.8	19.9
Upper jaw length	31	17.3	23.6	21.0	39	17.2	25.0	21.6
Horizontal eye diameter	31	26.3	32.7	29.1	33	24.5	32.1	28.3
Least interorbital width	31	26.8	34.6	29.7	33	25.0	30.8	28.1

(18, state 2) Infraorbital 3 absent. This is an autapomorphy for S. broccae . See apomorphy 18 for a discussion of infraorbital loss in species of Spintherobolus . The anal-fin ray count for S. broccae is 13-16, x = 14.1 versus 11 -14, 12.6 for both S. ankoseion and S. leptoura . The ventral procurrent caudal-fin ray count for S. broccae is 14-17, x = 15.4, versus 12-16, x = 14.0 for S. ankoseion and 13-15, x = 14.1 for S. leptoura. These relatively high counts for S. broccae relative to all other species of Spintherobolus appears autapomorphic for this species. Description. Morphometric data are summarized in Table 4 . Body compressed, moderately elongate. Predorsal body profile slightly convex. Body profile along dorsal-fin base posteroventrally inclined, becoming straight or slightly convex between dorsal and caudal fins in females and nearly straight to slightly convex in males. Ventral profile convex from anterior tip of mandible to pelvic-fin origin. Belly profile slightly convex between pelvic-fin and anal-fin origin in females, slightly concave in males. Body profile along anal-fin base concave in females; slightly convex at anterior anal-fin lobe of males and concave along posterior short finned section Ventral profile of caudal peduncle concave i females. Keel formedby prominent ventral pro current caudal-fin rays of males makes ventra profile of caudal peduncle nearly straight bu inclined ventrally in a posterior direction. Snout short, shorter than eye diameter. Moutl terminal, gape angled posteroventrally; maxil short, not reaching line drawn vertically throug anterior margin of eye. Dorsal-fin rays, ii,8-9 (x = 8.9, n = 28). Dorsal fin origin behind midbody. Adipose fin absent Anal-fin rays iii-iv,13-16 (unbranched rays x 3.1, n = 88; branched rays x = 14.1, n = 90) Anterior lobe of anal fin of males rounded, long er than posterior lobe, concave. Anal fin in fe males deeply concave, decreasing in size from posterior unbranched ray to sixth branched ray following rays all of equal size. Anal-fin origin a vertical line drawn through origin of most poste rior dorsal-fin ray in females and at vertical line drawn through middle of dorsal-fin base in males Hooks in anal-fin rays rare among examine males. When present, hooks much reduced in size, found on posterior face of two to six bon segments of posterior branch of first and some times second branched rays. When present, each ray segment bears two to three hooks on each side. Anal-fin rays one through five in males large, second to fourth branched rays more than five times thicker in sagittal dimension than following rays. These large rays with a process on anteroproximal face of each lepidotrich, these being projections of rays two through four, each inserted between lepidotrichia of next anterior ray. Each bony ray segment of posterior branch of branched rays two through four and sometimes also the first and fifth rays, not relatively square like other branches of same ray, but with their posteroventral angle acute and elongate, angled towards distal tip of that fin ray. Segments ofbranched fin rays two and four progressively fused to each other. Pectoral-fin rays i,8-10 (x = 8.6, n = 36). Posterior tip of longest pectoral-fin rays extending beyond pelvic-fin origin in both males and females. Pelvic-fin rays i, 5 in all counted specimens (n = 36). Tip of largest ray reaching to or beyond anal-fin origin of males, but not reaching that fin in females. Pelvic-fin rays of males with hooks on unbranched ray and 2 or 3 branched rays, 2 each bone segment. Hooks on ventromedial face and along all length of unbranched to second branched ray and along half length of third branched ray. Principal caudal-fin rays count usually 10/ 9 ( one specimen 10 / 10 and one specimen 9 / 9, n = 85). Dorsal procurrent caudal-fin rays 9-13 (x = 11.0, n = 86). Ventral procurrent caudal-fin rays 14-17 (x = 15.4, n = 89). Ventral procurrent caudal-fin rays of males large, well-developed, with their proximal tips fused, slab shaped and expanded in sagittal plane. Four to five most posterior caudal vertebrae (including posterior ‘half centrum’) supporting principal caudal-fin rays and procurrent rays, with enlarged and modified hemal spines. Those elements anterior to the hemal spine of antepenultimate vertebra fused to each other and with lateral bony projections between muscles and skin. Anteriormost ventral caudal procurrent rays with proximal tips absent or abbreviated in length. Scales cycloid. Lateral line with 2-4 perforated scales (x = 3.4, n = 34). Scales in a lateral series, 31 -33 (x = 31.8, n = 23). Scale rows between dor ­ sal-fin origin and lateral-line series, 4-6 (x = 5.0, n = 31), and between lateral line and pelvic-fin origin, 4-5 (x = 4.7, n = 31). Scale rows around caudal peduncle, 14 in all specimens (n = 31). No modified scales on caudal fin. Scale sheath along anal-fin base absent. Teeth. All jaw teeth with a basal pedicle and laterally expanded conical distal tip. Anterior ­ most dentary teeth tricuspid followed lateroposteriorly by conical teeth. Premaxillary and max ­ illary teeth conical. Premaxillary teeth 5-7, maxillary teeth 5-8 and dentary teeth 5-8 in four c&s specimens. Vertebrae 32-34 (x = 33.0, n = 88). Branchiostegal rays 4 in four cleared and stained specimens, with same attachments as in S . papilliferus. Color in alcohol. See Figure 2 , 32-33 for the approximately equivalent preserved color patterns of adult males and females. A discrete lateral body stripe extends from just dorsal to the dorsalmost part of the gill opening to the caudal peduncle where it continues onto the caudal fin as a very solid dark wedge with its apex covering the two middle caudal-fin rays. In the photographs of a male and female, there appears to be a dark humeral spot just posterior to the more dorsal region of the opercular flap. This is not a humeral spot. Instead it is the area of the anterior and posterior pseudotympanums and the black pigment coating the swimbladder is visible through the skin. The scale borders are covered by a single row of dark chromatophores above the lateral stripe, giving the dorsal part of the body a reticulate pattern. The abdomen and sides ventral to the lateral stripe is pale cream color or yellowish (the color of preserved muscle tissue) and the scale borders are devoid of dark chromatophores. The posterior termination ofthe caudal peduncle, both dorsal and ventral to the lateral stripe, is white, without dark pigment of any kind. The pectoral and pelvic fins of both sexes are hyaline with a few scattered dark chromatophores, especially on the anterior undivided ray of both fins. The dorsal fin is mostly hyaline except for its anterior dorsal border where the anterior undivided rays and the first branched ray have numerous dark chromatophores. The rays posterior to these dark rays progressively lose the dark chromatophores except distally where the posterior tip of the second and often the third branched rays have many dark chromatophores. The caudal fin is hyaline with a thin scattering of small dark chromatophores over the rays and membranes, especially those of the dorsal parts of the dorsal lobe of the male. The anterior base of the anal-fin is the origin of a second dark, relatively broad stripe which covers the muscular anterior base of the fin and extends posteriorly to almost completely cover the short anal-fin rays that follow the large, elongate, anterior anal-fin lobe. The anterior lobe of the fin, distal to its muscular base, is hyaline except for the distal tip of the first branched ray. The head is pale brown around the mouth with a scattering of dark chromatophores over the ventral surface of the snout, jaws, and oper ­ cular region. In the photographs ( Figs. 2 , 4 , 32- 33 ), the snout region, especially the ventral part of the jaws, is overexposed and the dark chromatophores do not show. The dorsal surface of the snout and the area between the eyes are pale brown, while the dorsum of the cranium and the nape are dark brown. The head area posterior to the eye and extending ventrally from the parietal region to the preopercle and branchiostegal rays is white. The circumorbital region ventral to the eye is dark, with a series of dark chromatophores, giving the impression of a spot ventral to the eye. The opercular region is dark dorsally and pale brown to white below in both sexes. Color in life. See Figure 35 of a female, taken immediately after capture, from near Bertioga, São Paulo. Life colors are very much like those in preservative except that the dark stripes are black, the abdominal region bluish white and the two white spots at the posterior termination of the caudal peduncle are yellowish to white. The body is translucent greenish yellow. The black chromatophores anterior, dorsal and posterior to the pseudotympanums are expanded and the area they occupy is black; those posterior to this are somewhat contracted and the lateral stripe remains less dense until approximately the anterior end of the caudal peduncle where the black chromatophores are again expanded and the stripe appears black. The black spot ventral to the eye is quite dark with expanded black chromatophores. The premaxilla and maxilla are black. The dentary is white except for scattered black chromatophores in its anterior region. The relatively hyaline regions of the caudal fins are pale reddish and a bit of this color also occurs on the anterior two branched rays of the anal fin. Sexual dimorphism. Males and females can be recognized externally by the anal fin-shape; compare Figures 32 and 33 . The anal fin of the females is deeply concave and divided into a prominent anterior lobe that has a rather straight posterodistal border and a narrow posterior lobe made up of relatively short fin rays. The anal fin of males is very similar except the anterior lobe has a rounded or circular posterior border. Not all males have hooks on the anal fin but all of them have hooks on the pelvic-fin rays and anal-fin rays one through five enlarged. Males have a keel in the ventral profile of the caudal peduncle. This is formed by the ventral procurrent fin rays. The ventral procurrent rays of female are unmodified. Discussion. Spintherobolus broccae was described as a new species by Myers (1925: 143) from aquar ­ ium specimens and the species was previously misidentified as Hasemania bilineata Ellis (1911: 1950) by Rachow (1924: 601) . Meinken (1929 a-b) provided the history of the species discovery and importation. Myers (1925: 144) considered S. broccae and Phoxinopsis typicus Regan (1907: 262) possibly generically the same and Myers (1926: 273) was apparently convinced of this for, he listed the species as P. broccae in his list of tropical fish names published by Innes. However, he had not seen Regan’s type of P. typicus and the illustrations of the two species, Regan (1907: 262) and Myers (1925: 144 , pl. 10), although not exactly the same, look similar enough so that the distinctness of the species was justifiably questionable, especially since Regan ’ s drawing was a hasty sketch. It wasn’t until Innes (1935: 130) , on the advice of G. S. Myers, listed the species as P. typicus , that S. broccae came to be considered a junior synonym of P. typicus . Nearly all subsequent aquarium fish literature considered the species to be P. typicus. Travassos (1953: 512) recognized the species as S. broccae . Böhlke (1954: 30-31) tentatively rec ­ ognized P. typicus and S. broccae as distinct species and he included these two nominal species and S. papilliferus in Phoxinopsis . His reasons for doing this were based on information received from G. S. Myers sometime before 1953. According to Böhlke, “Prof. G. S. Myers a number of years ago sent a paratype of his Spintherobolus broccae to the late J.R. Norman of the British Museum (Natural History) for comparison with Regan’s type of Phoxinopsis typicus. Dr. Myers has informed me that after a point by point comparison ofthe two specimens (including the teeth, which are very important characters) Norman concluded that they were specifically identical”. Myers must have sent a paratype to Norman soon after publication of the original description because the catalog number of the specimen is BMNH 1925.10.19.1 (Eschmeyer & Ferraris, pers. comm.). This is corroborated by Myers (1926: 273) wherein he used the generic name Phoxinopsis rather than Spintherobolus . In 1973 Böhlke wrote to SHW that he examined the type of P. typicus and found it to be an Aphyocharax Günther, 1868 . Later the holotype of P. typicus was borrowed by SHW and identified as Aphyocharax anisitsi Eigenmann & Kennedy (1903: 517) . Thus S. broccae is not a junior synonym of P. typicus, but P. typicus is a junior synonym of A. anisitsi and Phoxinopsis is a junior synonym of Aphyocharax . Spintherobolus broccae and A. anisitsi are very different fishes and it is difficult to understand how Norman confused the two. If young specimens of A. anisitsi are compared with comparable sized adult female specimens of S. broccae , about 18-20 mm in SL, they appear somewhat similar in body shape and fin position, but not in color and certainly not in their teeth and mouth shapes. The three paratypes we have seen of S. broccae ( USNM 92959, CAS 24073) are now faded and have lost their dark markings. If they had lost these when Norman examined his specimen, the type of P. typicus and the paratype of S. broccae would have had no striking color pattern differences. Although both species have relatively elongate and somewhat similar jaw teeth, those of S. broccae are more slender. Furthermore the gape of S. broccae is wide and the toothed portions of the dentary and the premaxilla ex ­ tend laterally almost at right angles to the long axis of the head and body. In A. anisitsi the mouth opening is narrow, the snout conical, and the toothed portions of the premaxillae and dentaries are curved posteriorly. Sarraf (1997: 28) recorded the principal caudal-fin ray count as i, 18,i. This produces a principal ray count of 20 rather than 19. Yet when we counted the principal caudal-fin rays in figure 8 of Saraaf (1997) we counted 10 / 9 which is equivalent to nearly all of the counts given above in our description of S. broccae and would be i,17,i for a total 19 in Sarraf ’ s method of counting. The dorsal-fin ray count was recorded ii,8 by Sarraf (1997: 28) . We found the number to be ii, 9 in most specimens, and found ii,8 as an unusual count (x = 8.9, n = 28). Ecological notes. Myers (1944: 186) reported specimens from “behind the beach of Sernambetiba” to live in dark-brown swamp waters. SHW has taken this species ( USNM 287324, MCP 19196) in somewhat brown acid waters of a stream in the forests surrounding the town of Cachoeiras de Macacu, Rio de Janeiro and from clear waters of a stream draining into lagoa Saquarema ( USNM 34206). The area around lagoa Saquarema was at one time at least partly forested, but this is no longer true. Nevertheless, S. broccae has survived in the region. The species appears to be present in somewhat acid waters in relatively lowland areas, in both swamps and slow moving streams, to the north and south of the city of Rio de Janeiro, but only in areas where the streams remain unpolluted.