ANNOTATED LIST OF THE SPECIES 365 wider at the opercle (1.4 compared with 1.6); head 3.0 in the length, as wide as long and equal to greatest depth of the body; eye smaller (4.6 as against 4.2), slightly oval, equal to snout (while in luteus of this size the eye exceeds the snout); eye in the interorbital width 1.7—2.0 as compared to 1.3. The mouth in these specimens is, as the authors state, vertical, wider than in luteus, the gape equal to the interorbital width; teeth much fewer, and in a single series, compared with a double rank in luteus. Opercle of about equal extent, but ending in a much less acute angle. Caudal peduncle shorter, squarely truncate rather than expanded at the base of the fin.
Preorbital scaled; cheek scales reaching to the anterior vertical from the eye, or nearly so; three vertebral series rather regular, 12 scales in middle series; granulations of anterior scales beginning to appear.
Numerous clumps of chromatophores falling into 2-3 longitudinal rows, forming a reticulate pattern. Dorsal and caudal fins from nearly hyaline to several broken rows of fine punctulations, very few on anal and pectoral. Head and shoulder scales darker about margins, with lighter centers.
Turning to the bay of Huancané, we find a series of smaller specimens, some of which are more slender, and may be luteus. In this age range the scales are more deciduous, the vertebral series more regular, 12-14, only the largest beginning to show any granulation of scales, the opercle often showing silvery through the still-thin scales. Colors similar. O. luteus of this size more arched on the nape, less robust, with more tapering caudal peduncle, lateral angle more acute, mouth more vertical and more deeply cleft, eye more elevated, larger, 4.3 in the head; crown flatter, overhanging the larger eye; snout shorter, 0.5 times the eye; more granulations begun on the enlarged scales; finrays all separated by a dark membrane; reticulated patches wanting.
16080, 4, 97-120 mm., Puno, Allen, November, 1918.
Separable from nearby O. luteus by the extreme breadth of the opercle, the pointed head, small, narrow much upturned mouth with fleshy lips, pointed snout, often greater convexity of the predorsal arch; shorter body, abruptly compressed. The chunkiest of all the species, with the most pointed head, the contours of a plummet.
Postorbital portion of the head 1.4-1.7 greater than the anterior moiety (end of snout to posterior border of orbit); opercle acutely angular; body becoming compressed rapidly from opercle to anus; caudal peduncle deep (but less so than OQ. luteus), its depth nearly equal to the postorbital portion of the head, shorter than the peduncle of luteus, widening less at the base of the caudal. Mouth usually described as larger than that of lutews. This should be interpreted as meaning wider, but with the gape crescentic and not curved sharply downward into a horseshoe. Except possibly for O. olivaceus, the most superior mouth of all Orestias species; thick-lipped, almost without teeth, as described by Valenciennes.
These specimens may at least be reconciled with the much smaller ones above,
366 FISHES OF WESTERN SOUTH AMERICA since the difference might be accounted for by age; the mouth is similar, a rather unique character, the light-centered scales of the shoulder, size of eye, paucity of teeth.
Eye small, 5.5 in the head, rounded, equal to luteus in its distance from the end of the snout; eye without the protuberant supraorbital; width of head surpassing its depth; head 2.7 in body length; interorbital 2.7 times the eye; crown considerably elevated and convex in both dimensions, with a short, flatter area at the occipital; snout sometimes expanded to give the effect of a proboscis. Scales similar to luteus in the extent and form of the granulations; resembling it also in the naked infraorbital, outlined by a row of pores, and snout naked; surpassing it in the investiture of the prepectoral, represented by Valenciennes with pectoral fin surrounded by a scaleless area, which Starks makes a diagnostic character; cheek scales barely reaching, not passing, the anterior level of the orbit. Vertebral series three, regular, as described by Garman, without lateral desquamated areas.
D. 14; A. 18-14; P. 19; C. 28.
While O. luteus belongs to the parchment-finned group of species, those of albus are smaller, fleshier, more rigid in alcohol, thick-skinned, and heavily coated in mucus. Garman says they are medium, I should say small, with the dorsal basis less than its distance from the first caudal finray; pectoral broad; caudal deeper than long, its depth half that of trunk, while /uteus exceeds half, and is conspicuously larger; I am unable to find 35 rays as did Valenciennes. Sharing with O. luteus the extensible, muscular, unscaled vent region, as though used for oviposition.
Colors in aleohol varying from red-brown to gray-green and nearly black; fins colorless or nearly uniformly brown.
Garman was in error as to Lake Umayo, type locality, being without outlet. It drains by a small stream, the Rio Ilpa, to the bay of Puno. 481. ORESTIAS SILUSTANI Allen, sp. nov.
Plate XIX, fig. 1 16097, 5, 55-86 mm., Lago Umayo, Allen, January, 1919.
Catalogued as O. olivaceus. Comparison with the cotype of the latter brings out numerous differences. They resemble one another in the extremely superior mouth and convexity of the ventral surface of the head. My type has an even straighter dorsal profile with less indication of a predorsal hump; like olivaceus the mouth and cheeks widely and roundly expanded, without the angularity of luteus, but the width of the head not equal to that of olivaceus; differing from it in the smooth, moderately thickened scales of opercle, without the granulations. Differing in the lack of the almost uniform green color of olivaceus. The type is of the same size as the olivaceus cotype, is more compressed, more tapering, with a slenderish caudal peduncle, body form much less expanded and robust. My paratypes are sufficiently unlike the type to be mistaken easily for luteus
ANNOTATED LIST OF THE SPECIES 367 or albus at first, and a considerable metamorphosis must be assumed between the stages here represented. They cast some doubt upon the validity, otherwise satisfactory, of the assortments of genes by which my two main branches of the genus are separated.
Dorsal profile very slightly and uniformly elevated from snout to dorsal fin, thence very slightly concave; ventral profile very convex from the elevated mouth to the throat, thence slightly convex, anal basis a little oblique; general form tapering from head to tail; caudal peduncle thinly compressed, rather short, not deep nor expanded at the caudal basis; width at opercle moderate, 1.6 in the depth, which is equal to the length of the head and 3.3 in the length to the base of the caudal. Head much shorter than in olivaceus, 3.6 compared to 3.0. Kye rather small, circular, elevated, 5.0 in the head, 2.0 in the interorbital space; mouth very superior, rather small, less than interorbital space; lips thin, teeth in one weak series in each jaw.
Scales striate, infraorbital sparsely scaled, scales of shoulder, cheek, and opercle only moderately thickened, lustrous; 15 scales in vertebral series; those of prepectoral area sparse, but postpectoral well scaled downward to a narrow naked band on the belly.
ID),WakeANSlegEXPAO):
Origin of dorsal midway from occiput to caudal basis; fins small, thin, membranous, not roughened or fleshy.
Light gray to light brown above; whitish ventrally; conspicuously yellow on opercle; numerous small patches of minute chromatophores on the body; fins with staggered rows of fine, elongate punctulations, more regular on caudal, united into larger spots on the dorsal.
The smaller specimens, paratypes, were at first mistaken for luteus or albus, being more convex on the back, but are of more elongate form, and begin to show the smooth-scaled epaulettes of the miilleri group at the age when albus and luteus are taking on the granulate form of scale. Differing from albus of equal size in the color and much more compressed rather than rounded; unlike luteus in the less pointed lateral angle, more pointed and not truncate head, oblique gape, smaller eye, more elongate caudal peduncle.
The name silustani is that of the nearby Inca ruin on the shore of Lake Umayo, overlooking the weedy habitat from which the species was collected. 482. ORESTIAS EMPYRAEUS Allen, sp. nov.
Plate XX, fig. 2 Central Peru; upland pampas, upper Mantaro 15232,many,20-115mm.,Huancayo,Eigenmann,September,1918.
15233, many, Jauja, Peru, Eigenmann, September, 1918.
15234, many, Pachachaca, Eigenmann, September, 1918.
15235,1,35mm.,R.Mantaro,EigenmannandAllen,September,1918. 15236,many,25-63mm.,Zigzag,Allen,September,1918.
15237, 22, 27-94 mm., Tilarnioc, Kigenmann, September, 1918. 15238,many,16-150mm.,LagoChinchaycocha,Allen,October,1918.
368 FISHES OF WESTERN SOUTH AMERICA 15239, 38, 18-114 mm., Rio de Junin, Allen, October, 1918. 15241,many,24-78mm.,LagoYanamate,Kigenmann,September,1918. 15244,6,21-34mm.,Tuctu,nearMorococha,MissHeywoodandMr.A.8.Kalenborn,September, 1918(elevationabout14400).
15245, Challhuacocha, near Quishuarecancha, Mr. Murdock, September, 1918 (elevation 13200 feet).
15246, Pocobamba, Adele Eigenmann and Mr. Emerson, September, 1918 (elevation13700feet).
15248,many,25-65mm.,railwaytoGoyllarisquisga,Eigenmann,September,1918. Although impressed by the intense coloration and finely modeled form of these fishes while collecting them in the crystal waters of Chinchaycocha, against the Fia. 44. The Colombian Charapero, or turtle-dealer, on the upper Peruvian Amazon. His raft isofBalsa-wood;histurtle-penformstheground-floor,hisdwellingtheupperstory. equally vivid green aquatic plants and sponge of the false bottom of that lake, I had been content to catalogue them as a variety of O. agassizii. Evermann and Radcliffe had so regarded the younger specimens of the nearby streams at La Fundicion. Resuming the study of Orestias more recently I find these fishes very distinct from agassizii, or any other species, showing more resemblance to the specimens of intermediate size of var. inornata Pellegrin than to any other. Comparing adult specimens with those of O. agassizii of equal range in size, I find them:
More elongate, narrower, with more rounded contours transversely; the width
ANNOTATED LIST OF THE SPECIES 369 only slightly exceeding the length of the head, the dorsal profile straight, or nearly straight, on the head, while convex in agassizii; head pointed, its depth through center of eye less than that of caudal peduncle; mouth wider, equal to interorbital, lips thickened, fleshy; little enlargement or convexity of the shoulder scales; median fins more rectilinear, but caudal more rounded. Colors in alcohol, redbrown upper parts, merging into maroon below; fins unmarked or with dark band at base of dorsal and anal.
A comparison of smaller specimens reveals such differences as the following: More elongate than O. agassizii, head more pointed; more rounded form: dorsal profile less convex; crown and snout narrower, interorbital space narrower; eye less oval; chin less vertical; fins larger in both dimensions, pectoral especially broad, median fins considerably broader, but caudal not so elongate in extreme young; colors begin early to differentiate as above. Thick lips are a good character for separating adults, but I do not find this consistent in the smaller specimens. Considering the small size and juvenile character of the Valenciennes types, less than 4 inches, our young specimens of this species could readily be confused with them, especially with their great variation and the metamorphoses through which they pass.
I find the present species, by reason of its more minnow-like form and less specialized scales, to be nearer than O. agassizii to the primitive Orestias stock. Since nowhere in the northern range of the genus have we collected large specimens similar to the O. agassizii of Lake Titicaca, and since all the larger specimens taken in the north are identical with the present type, I am convinced of their complete isolation geographically. The great gorges of the Apurimae and other streams appear to be an effective barrier between them.
Description of the type, 120 mm. long, Lake Chinchaycocha, (15238). Dorsal contour straight on the head, faintly arched above the opercle, slightly convex to origin of the dorsal fin, descending slightly along its base to rectilinear caudal peduncle. Contours rounded, never angular. Caudal peduncle slightly longer than dorsal basis.
Width in depth 1.5, shghtly greater than length of head, and 3.5 in body length; eye 4.8 in the head, equals snout, 1.4 in the interorbital space; head 3.7 in the body length; snout narrow, equals width of mouth and length of opercle; chin roundly pomted, mouth oblique, crescentic in front view, lips thickened and fleshy; teeth sparse.
Scales showing striae, becoming moderately thickened on forward parts, but well-embedded, not deciduous anywhere; scale-rows somewhat irregular on nape, head, and shoulder; nowhere polished or granular; few scattered scales on suborbital area; smooth, tightly embedded, but not dense, on cheek; similar, but closer, on opercle; very few on pre-pectoral area; preventral widely exposed, more narrowly on post-ventral to anus.
Ree Ava @. 28:
The fins somewhat thickened, but scarcely fleshy, with coarse rays, somewhat roughened; dorsal and anal more rounded than in young; caudal somewhat broadened fanwise; pectoral broad, fan-shaped.
370 FISHES OF WESTERN SOUTH AMERICA Colors as mentioned above, most readily confused with O. agassizii var. inornata, such as those of Maravillas, but readily distinguished when brought together.
Garman’s description of O. agassizvi collected at Puno applies to larger specimens than the types of Valenciennes, and was intended to supplement the original description. This it does very satisfactorily as applied to my larger specimens. His description and that of Evermann and Radcliffe, the latter taken from small specimens in central Peru, and large specimens from southern Peru, fail to take into account larger specimens from central Peru, such as the present species. The name empyraeus is in allusion to the extremely elevated habitat, mostly 10000-14000 feet, the species being at its best in Lake Chinchaycocha, elevation 13500 feet.
483. ORESTIAS ELEGANS Garman Plate XX, fig. 1 OrestiaselegansGarman,1895,Mem.Mus.Comp.Zool.,XTX,149; Eigenmann, 1910, Rept. Princeton Univ. Exped. Patagonia, III, 461; Evermann and Radeliffe, 1917, Bull. U. S. Nat. Mus., no. 95, 40; Kigenmann, 1922, Mem. Carnegie Mus., [X, 184.
Central Peruvian Andes 15242,46,37-102mm.,smallglaciallake,15000-15500feetelevation,nearCasapalea,Peru,Eigenmann, September, 1918.
The only known Orestias of the Pacific slope; small lakes at the headwaters of the Rio Santa Eulalia, a small precipitous river arising just west of the continental divide near Casapalea; flowing into the Rimac above 3000 feet elevation and above the flume of the Chosica hydroelectric plant. Although populated at the sources, I found no fish life in the lower courses of the stream, near Chosica. D. 15; A. 16; P. 16-18; Ll. 34-36.
Elongate, minnow form, ventral contour nearly straight, dorsal uniformly and moderately arched; head rounded above; eye usually circular equal to snout, more than 4 in the length of the head; H. 4.3 in length without caudal; greatest depth equals or exceeds length of head, and 3.75-4.3 in the length to the caudal. Mouth small, nearly vertical, the cleft reaching below the level of the eye, its width 1.4 in the interorbital space; teeth small, hooked, in a complete single series. P. short, not attaining to the midpoint from its origin to that of the anal; ‘audal only about the length of the head.
In form like an elongate O. agassizii, with the shoulder scalation more like that of O. miilleri, but without the widening of the body at that point; all other scales discernible small, weak, thin, mostly smooth; scales of cheek extending forward of the eye; few or none found in infraorbital region, which is encircled by a row of pores.
Specimens badly preserved, fins badly broken, all nearly devoid of scales, emaciated. Naked skin pale yellow, with red chromatophores on dorsal areas.
ANNOTATED LIST OF THE SPECIES 371 Garman reports a 23-inch specimen with well-developed eggs; some of mine had developed gonads at 70 mm., (22 inches).
15238 (part), 2, 85 and 92 mm., Chinchaycocha, Allen, September, 1918. Two poor specimens of quite different nature from any other, at least near their own size, having somewhat the general form of mature O. agassizii, but much less robust, and differing in various respects; close to elegans in a number of ways, and since they were taken in contiguous territory to that of elegans, and are difficult to separate without better material, I place them here not too confidently. Similar to elegans in the proportions of the head regions, mouth, teeth, eye, size of head, shoulder scalation, finray counts. Rather less slender and less circular in cross section, compressed; caudal peduncle measurably deeper. Scales thin, flat, not much striated, deciduous, regular except upon the back; shoulder scales somewhat thickened, polished; in one specimen a dense group of scales in the infraorbital area; cheek well filled with close scales extending before eye; opercle similar except anterior horn.
Fins moderately long.
Yellowish-white below, sides yellow-brown, nearly uniform; fins hyaline without markings.
484. ORESTIAS AGASSIzIT Valenciennes Plate XX, figs. 3-8 Orestias agassizii Cuvier and Valenciennes, 1846, Hist. Nat. Poiss., X VIII, 238, Corocoro, Bolivia; Eigenmann and Eigenmann, 1891, Proc. U.S. Nat. Mus., XIV, 65; Garman, 1895, Mem. Mus. Comp. Zool., XIX, 150; Steindachner, 1902, Denksch. Kix. Akad. Wiss. Wien, LX.XII, 58, pl. iii, fig. 3; Pellegrin, 1904, Bull. Soc. Zool. France, XXTX, 93; Starks, 1906, Proc. U. 8S. Nat. Mus., XXX, 780; Pellegrin, 1906 (1907), Poiss. Lacs Hauts Plat. Amér. Sud, 19, pl. xiv, figs. A-D; Evermann and Radcliffe, 1909, Proc. Biol. Soc. Wash., XXII, 165; Evermann and Radcliffe, 1917, Bull. U. 8S. Nat. Mus., no. 95, 40; Rendahl, 1937, Ark. Zool. Stockholm, X XIX, A. 11; Eigenmann, 1927, Mem. Nat. Acad. Sci., XXII, 52, pl. xiii, fig. 8. OrestiasagassiziEigenmann,1910,Rept.PrincetonUniv.Exped.Patagonia,III,461. Orestias ortonia Cope, 1875, Jour. Acad. Nat. Sci. Phila., 186. Orestias frontosus Cope, 1875, Jour. Acad. Nat. Sei. Phila., 187. Orestias owenit Garman (part), 1895, Mem. Mus. Comp. Zool., XIX, 152. OrestiasalbusGarman(part),1895,Mem.Mus.Comp.Zool.,XTX,153. Orestias tirapatae Boulenger, 1902, Ann. Mag. Nat. Hist., (7), X, 153; Eigenmann, 1910, Rept. Princeton Univ. Exped. Patagonia, III, 461. Higher Andes of southern Peru to northern Chile 15239 (dupl. number), 22, 43-84 mm., Rio Mauri, Calacoto, Bolivia, Allen, February, 1919, (type localityPellegrin’svar.typica).
15249, many, Rio Mauri, Desaguadero, Allen, February, 1919. 15250, many, 50-85 mm., Rio Mulato, Bolivia, Allen, February, 1919. 15347, many, Lago Ascotan, Chile, Allen, February, 1919. 16060, many, same data.
16067,manyverysmall,LagoHuaypo,Eigenmann,November,1918.
372 FISHES OF WESTERN SOUTH AMERICA 16089,5,foursmall,58-73mm.,onelarge,180mm.,bayofHuanecané,Allen,December,1918. 16093, many, Calacoto, Bolivia, Allen, March, 1919.
16101, many, small to minute, R. Colorado; Viacha, Allen, March, 1919. 16106, many, Rio de Juli, saline, Allen, December, 1918.
16108, many, Rio Colorado, Viacha, Bolivia, Allen, March, 1919. 16112, many, wet meadows, Pampa de Acora, Allen, December, 1918. 16114, many, 28-8 8 mm.,R. de Eucaliptus, Bolivia, Allen, January, 1919. 16116, 18, 56-74 mm., roadside pool near Juli, Allen, December, 1918. 16117, many, pools and creek, Ilave, Allen, December, 1919. 16118, 15, Rio Caminaque, Have, Allen, December, 1918.
16119, many, Rio de Have, Ilave, Allen, December, 1918.
16121, many, millrace, Maravillas, R. de Lampa, Allen, November, 1918. 16122, many, 29-74 mm., Lagunillas, Allen, March, 1919.
16124, 20, 28-82 mm., R. de Lampa and lagoons, Allen, January, 1919. 16125, 18, 31-94 mm., Rio Grande de Lipez, Bolivia, Allen, February, 1919. 16126, 11, 40-103 mm., R. de Tiahuandco, pond, Allen, March, 1919. 17867,many,40-100mm.,R.deSantaRosa,Eigenmann,December,1918. ———, many, Puno mole, Allen, November, 1918.
——, 2, 86 and 117 mm., Chucuito, Allen, November, 1919.
Of all the members of the genus this is the most widely dispersed and the most given to variation, a circumstance which has been a constant incitement to the creation of species and varieties. It has the greatest range, from the mountains about Cuzco and the passes of the western cordillera, throughout the Titicaca basin southward farther than other species to the southern extremity of the Bolivian and Chilean altiplano.* With two exceptions it is the nearest the primitive ancestral form, being little specialized from the ancestral minnow-like form, especially in the younger stages, and resembling in its juvenile stages those of other species, most if not of all. I have found it the most adaptable of all the Orestias, occurring in all types of marshy lands, streams and lakes. It has the greatest range in latitude, is found at higher and lower altitudes than any other, and has adjusted itself to all degrees of salinity and freshness of the waters of the interandine valleys. Hence the variations of form and color, especially of color. In some of the intermediate areas agassizii is replaced locally by other species, such as miilleri, their distribution being reminiscent of that of the two dominant tribes of the land, the Quichuas and the Aymaras.
Although not esteemed for food to the degree I have indicated for O. cuviert and pentlandi, yet among the aborigines it is a food fish of considerable importance, despite its small size. The larger ones are marketed fresh, while the smaller ones are dried entire and are a reserve food for use when the fresh ones are unavailable. In the shops at Puno, Cuzco, La Paz, and elsewhere I encountered large burlap bags of dried Orestias agassizii (and the other small species) displayed with the tops of the bags rolled back, and standing about on the earthen floors among the still larger bags of the dried leaf of the coca plant.
Known today as in Pentland’s time as the carache. This name Pentland varied as carachito, but I did not encounter the diminutive form except when young *Higenmann,1927,page5,correctlystatesthatthisspecieswasfoundinLakeAscotan.His captionunderfig.31sinerrorinstatingthatO.agassiz7iextendsdownthewesternslopetotheRio LoaatCalama,whichIfounddevoidoffishesnow,althoughfromtheextensivetamperingwiththat streamsomeformerpopulationmayhavebeenexterminated.