A revision of Astyanax (Characiformes: Characidae) in Central and North America, with the description of nine new species Author Schmitter-Soto, Juan J. text Journal of Natural History 2017 J. Nat. Hist. 2017-07-03 51 21 - 24 1331 1424 http://dx.doi.org/10.1080/00222933.2017.1324050 journal article 10.1080/00222933.2017.1324050 1464-5262 5181888 ABC57223-DF66-49B6-8FE0-87CFF5D3EA03 Astyanax aeneus ( Günther 1860 ) ( Figures 3 (a), 5(b), 13–15) Tetragonopterus aeneus Günther 1860: 319 . Astyanax rutilus aeneus , comb. nov. , Eigenmann 1908: 24 . Astyanax fasciatus aeneus Günther 1860 , comb. nov. , Eigenmann 1921: 306 . Tetragonopterus humilis Günther 1864: 327 . Tetragonopterus microphthalmus Günther 1864: 324 . Astyanax fasciatus non Cuvier, Martínez-Ramírez 2000. Astyanax sp. 1 , Ornelas-García et al. 2008 . Diagnosis Diagnosed from other Astyanax species in northern Middle America and southern Mexico as follows: total gill rakers on first arch, mean 20–21 (mean 22–25 in A. angustifrons and A. caballeroi ; also southern populations of A. aeneus may have a higher number of gill rakers); longest dorsal-fin ray, mean 25.4% SL (mean 23.6%, in A. fi nitimus ); distance between origins of pelvic and anal fins, 18–22, mean 20.2% SL (16–19, mean 17.4%, in A. fi nitimus ); supraoccipital in lateral view, usually caudally concave (vs usually angled); ceratohyal foramen, comet-shaped or absent (vs oval). Redescription A species of Astyanax , subgenus Astyanax (i.e. with a complete predorsal series of scales). Head profile straight to concave; snout round or squarish. Lips even, mouth terminal. Pectoral fins may reach pelvic fin origin; anal and dorsal fins may overlap. Lobes of caudal fin, subequal. D. 9–11; A. 22–31, modally 25; pect. 10–11. Procurrent unsegmented dorsal rays on caudal fin, variable. Gill rakers on first arch, 18–23, modally 21; on lower limb, 10–14, modally 12. Scales on lateral line, 32–39, mean 34.5; predorsal scales, 9–14, modally 11; scale rows from lateral line to base of first dorsal-fin ray, 6.5–8.5, modally 7; to base of pelvic fin, 5.5–7; to base of pectoral fin, 3.5–5, modally 4; circumpeduncular scales, 14–18, modally 16. A single scale row on anal fin base, about 10 scales long. Nuptial tubercles may be bifid. Total vertebrae 32–33, 18–19 caudal. Detailed frequencies are given in Table 3 . Largest examined specimen, 90.5 mm SL. Body usually rather deep, 34–42% SL, mean 38%. Head length 27–33% SL; orbital diameter, 27–33% HL; interorbital distance, 6.6– 10.1% SL, mean 8.7 (further morphometric data appear in Table 4 ). Anterior fontanel long, straight-sided, blunt-tipped. Supraoccipital process in dorsal view, long, narrow-based, usually slightly concave in lateral view. Vomer rostrally slightly concave. Arms of premaxilla, subequal; 0–4 teeth. Highest tooth on dentary, first or third; the posterior teeth, abruptly smaller. Dorsal edge of longer articular arm, variable. Maxillary, with a convex anteroinferior edge; 1–3 teeth. Quadrate, dorsal process not expanded. Metapterygoid, rostral arm longer than ventral; dorsorostral projections, usually 2 (only 1, in the Río Papagayo population). Infraorbital II, triangular, usually with a short convex base (base angled in Honduran populations). Infraorbital III, usually inferoposteriorly semicircular; infraorbital IV, squarish with a projection; a narrow contact between infraorbitals II and III. Urohyal rostral end turned up, its ventrorostral edge convex, its ventral apex about equidistant between rostral and caudal ends (closer to caudal end in southernmost populations); ceratohyal foramen comet-shaped or absent, rostral vertices and ventral side variable. Epibranchial III, insertion of uncinate process widely open; upper pharyngeal bones, oval; lower pharyngeal plate, single; its caudal side, concave. Hyomandibular, variable. Opercular dorsal edge, convex; sides of dorsal half, parallel; caudoventrally straight-convex, caudodorsally concave. Interopercular posterior edge, straight-convex. Preopercle, ventral rim, curved; posteroventral edge, curved; 2 canals at angle. Four or 5 predorsal bony elements; rostral edge of first pterygiophore, angled. Coracoid with 2 interdigitations in suture to cleithrum, a concave caudal edge. Caudad process of postcleithrum, with parallel sides and convex tip. Proximal edge of pelvic bone, convex. Dorsal and caudal tips of scaphium, truncate; dorsoposterior edge, usually slightly concave. Neural spines under dorsal fin, straight. Postanal element, short. Sixth or seventh caudal vertebra from tail, with a haemal spine displaced caudad. Rostral edge of largest hypuric plate, straight. Epuric plate on last neural spine, edge straight. Humeral spot, oval to rhombic or P-shaped; often, a second humeral spot (sometimes only an intensification of the lateral stripe). Pigment on anal fin, sparse. Caudal spot, both on peduncle and on fin rays. Live colour was described for Salvadoran populations as ‘silvery with bluish green above; sides with an ill-defined bright silvery streak, most evident posteriorly; a dark shoulder spot, and frequently an indication of a second black spot behind it; dorsal olivaceous; caudal greenish, with an elongate black spot at its base extending to the end of the median rays; anal pink to deep red; ventrals usually more or less reddish; pectorals yellowish to red; upper part of iris red to reddish yellow’ ( Hildebrand 1925 , p. 245). A live specimen from Oaxaca is shown in Figure 13 . Figure 13. Astyanax aeneus , live specimen from San Lorenzo Cacaotepec, Oaxaca. © E. Martínez- Ramírez and V. Ortiz. Reproduced by permission of E. Martínez-Ramírez and V. Ortiz. Permission to reuse must be obtained from the rightsholder. Type material and depositor Lectotype BMNH 1907.4 .10.3, 67.5 mm SL, ‘Oaxaca’, Mexico , coll. A. Sallé , ca. 1850 ( Figure 14 ) . Paralectotypes : BMNH 1860.6 . 17.41–42 ( 2 specimens ), same collection data as lectotype . Figure 14. Astyanax aeneus , lectotype, BMNH 1907.4.10.3. Photo and permission to publish it: © The Natural History Museum, London. Reproduced by permission of The Natural History Museum, London. Permission to reuse must be obtained from the rightsholder. Figure 15. Map of northern Central America and southern Mexico, showing distribution localities of Astyanax aeneus , A. altior , A. angustifrons , A. bacalarensis sp. nov. , A. caballeroi , A. finitimus , A. macal sp. nov. and A. ocotal sp. nov. A larger symbol denotes the type locality (uncertain for A. angustifrons ). Distribution From Río Papagayo , Guerrero, Pacific coast of Mexico , to Río Choluteca , Pacific Honduras ( Figure 15 ) . The type locality is ‘ Oaxaca’ , Mexico ( Arroyo Las Flores , tributary to Río de Los Perros , Pacific slope; see Remarks and Discussion ) . Remarks Astyanax aeneus s. str. corresponds to clade Id of Ornelas-García et al. (2008) , their ‘species 1’, and clade F of Strecker et al. (2004) . Valdez-Moreno (1997) considered populations of Astyanax from Pacific Chiapas to be osteologically different from those in the Atlantic versant of Mexico . Schuppa (1984) and Paulo-Maya (1994) also observed morphometrical differences between the two forms. Martínez Ramírez (2000) found two quite distinct molecular clades in Oaxaca , one for each versant, arbitrarily calling the Pacific form ‘ A. fasciatus ’ (= A. aeneus s. str. ) and the Atlantic form ‘ A. aeneus ’ (= A. fi nitimus ). There are several clines in meristic traits along the distributional range of A. aeneus . Honduran populations have 34–39 scales on the lateral line, modally 35, whereas Mexican populations have 32–37, modally 33; the number of scale rows between the lateral line and the dorsal-fin origin is 6.5–7.5 in Guerrero and Oaxaca, vs 7–8 in Chiapas, Guatemala , El Salvador , and Honduras . Carr and Giovannoli (1950) claimed that ‘a detailed survey of the group [ A.aeneus ] might disclose at least 1 Honduran stock that required naming’. I checked their specimens at UMMZ (see Material examined), and I cannot discern two forms in the area, unless they refer to unmentioned specimens from the upper reaches of the Patuca River, Caribbean drainage, which indeed belong to a different species ( A. belizianus ). It is not possible to identify with certainty the nominal species A. oaxacanensis , because the types are lost (J. Gregorio, pers. comm.) and the type locality is not defined, given only as ‘Oaxaca’, a Mexican state with both Atlantic and Pacific slopes. However, see Discussion. Locality vagueness used to affect A. aeneus too. Regan (1908) thought that the type locality was the City of Oaxaca, not the state, but I cannot find the rationale for this. However, the research of Eschmeyer ( in litt . 2011) indicated that the precise type locality is Arroyo Las Flores , tributary to Río de Los Perros, Pacific slope. The French collector Auguste Sallé had moved his working area ‘from Vera Cruz into the State of Oaxaca on the Pacific side of the dividing range’ ( Sclater 1858 , p. 227).