ANNOTATED LIST OF THE SPECIES 349 The alimentary canal practically without a stomach, the intestine long, much longer than the fish, sometimes twice as long.
The most obvious differences between Empetrichthys and Orestias are that the former has a double series of teeth in the jaws, the mouth is terminal (though oblique), and that the dorsal fin is in a more posterior position. Myers points out that more deep-seated differences exist (see above) which throw Empetrichthys nearer the funduline group, the outstanding one being that in Empetrichthys parietals are found, absent in Orestias.
Cyprinodontiform fishes isolated almost altogether from all other types by the great altitude of the altiplane on which they live, upland plateaux or pampas between the great cordilleras. These extend for more than a thousand miles from near Cerro de Pasco to southwestern Bolivia and Chile, mainly at 12000 to 13500 feet above sea level. At some points the fishes have attained elevations of approximately 16000 feet, while in a few limited areas they spill over the rim of the interandine basin down the eastern slope to about S000 feet. We found them on the western slope at only one point, a small lake near Casapalea, hanging on the precipitous rim of the divide. Co®xtensive with the altiplane, the northern barrier is the great Nudo de Paseo, while southward distribution has been cut off by the rugged voleanic terrain along the headwaters of the Pilecomayo, and the Rio Grande of Bolivia. At one point at least the genus has been hemmed off by a complete ring of active and extinct volcanoes and exists in a land-locked terrain (LL. Ascotan of Chile) cut off from another land-locked system (the Uyuni basin); the latter in turn is hemmed off by low mountains and deserts from the third land-locked system (the Titicaca basin), yet through all these metamorphoses of the land the Orestias have held their own in all three basins.
The student of the genus Orestias is confronted not only with imadequate descriptions, but with poor and often misleading figures.
All Valenciennes figures give a very erroneous conception of the mouth. The fins are inaccurate, the arrangement of scale-rows mechanical. Pentlandi, e.g., is shown as having definite rows of scales, equally with other species, while in fact they are rounded and irregular throughout the length, where other species have this irregularity upon the forward parts at most. In none does his artist show the scalation of the caudal base, vertical, broken series such as all species have. No Orestias is fully scaled below, yet three of his figures could be so understood. He describes O. albus as always with exposed areas alongside the back, but these areas are figured with scales, and as a matter of fact they often are scaled. His O. agassizvi fairly well shows the appearance of a young specimen, but in no Orestias, young or adult, is the armor of thickened scales completed so far back. This caused later writers to confuse it with jussiev.
Castelnau’s three figures are a much brighter green than any member of the genus, in which all greens are “‘olivaceous.”’ He says the eye of O. tschudii is smaller and more ventral, but shows it like jussze7 in his figure. His fins are untrue to life, and like Valenciennes, insists on having the scale-rows of pentlandi quite regular. Deep-sea fishes are often brought to the top with the viscera everted under
350 FISHES OF WESTERN SOUTH AMERICA the diminished pressure at sea-level. Our Orestias come from a region having the atmospheric pressure two-fifths less than at sea level. This may account for the fact that they usually have the scaleless belly inverted to an unusual degree, in our collections, when transported to the greater pressures at low elevation. Although the published history of the genus Orestias is generally dated 1846, with the brief description, Valenciennes had first published a brief notice in the Institute, VII, p. 118, 1889. The type was not clearly stated but inferred in the Histoire Naturelle des Poissons, 1846, XVIII, at the bottom of page 225 and the top of page 226, where he says: ‘‘La premiére espéce de ce genre sera pour nous celle dont les individus paraissent atteindre 4 la plus grande taille, et qui ont dans leur dentition, dans le developpement de leurs boucliers granuleux et de leurs écailles, un ensemble de caractéres dont les autres espéces n’offriront que des modifications.” Thus O. cuvieri is well understood to be meant as his type. Prior to this time, however, Valenciennes informs us that at least two of the species were known, O. pentlandii and agassizii, at least to a small world, through drawings that had been made by M. Joseph de Jussieu, and transmitted to him by the grand-nephew, M. Adrien de Jussieu.
Following this only a few small random collections were made by Castelnau in the 1850’s, Orton in the 60’s, and Agassiz and Garman in the 70’s. The second resulted in the publication of a few ill-considered species by Cope. The lastnamed were the basis of Garman’s very incomplete revision in the Cyprinodont report of 1895. An equally unsatisfactory report of the fishes resulted from the incidental collections made by MM. Créqui-Montfort and de la Grange in 1908, published by Pellegrin (1904, and 1906). Starks (1906) describes some 37 specimens from Titicaca, belonging to five species. In 1917 Evermann and Radcliffe brought together all the literature on the genus and assembled thirteen species which they could not remove from the list considered valid. While on the Irwin Expedition as traveling fellow of the University of Illinois my major efforts were given to the more elevated regions. It follows that Pygidium and Orestias were by far the greater part of my collections for some four months of my work in the highlands of central and southern Peru, Bolivia, and northern Chile. Some thousands of specimens of Orestias, large and small, were brought home. At intervals from that time until the death of Dr. Eigenmann I was working on them. In 1929, through error, the Orestias materials were sent to the California Academy of Sciences with the Jordan-Eigenmann collections. Other portions of the material had been sent to the University of Illinois. The following pages will show the results of the study of only a part of the entire material available. Many specimens were dissected for parasitological materials, and for stomach examinations. The field notes on food may be summarized as follows: Orestias agassizii. Intestine 1.5 times the length of the fish. Algal filaments, Cladocera, Amphipoda, insect fragments. No indication of shells or grit between scissor points. Specimens of small size from streams and shores with larger fragments of food, taken among water plants, than the larger species taken offshore, containing plankton. Those from shallows of Ascotan and Huancané had fed on Amphipods and insects. One specimen from R. de Eucaliptus
ANNOTATED LIST OF THE SPECIES 351 had only Corixid nymphs and adults; another with lower intestine packed with Corixid remains, the rest of the gut with plankton, mainly diatoms. Orestias miilleri. Cheese-like masses containing the tests of minute Cladocera; no plant, no larger Crustacea. Some from Lake Poopo, 3.5 inches long, insects and fragments. From Moho, much red mud and grit; fragments of higher plants than algae; no shells. Intermittent rain with much mud entering through streams from the red sandstone formations of the region.
Length of intestine 1.5 times length of fish.
Orestias pentlandi. Many Cladocera, few algal filaments; mass of much finer material. Fishes more pelagic, intestine shorter than body, food finer than either of preceding.
Orestias luteus. Intestine packed with Crustacea and great numbers of small Amphipods. The smallest specimens contained the largest prey. Orestias albus. Gastropod shells, many broken, but others entire. An older specimen, above 6 inches, with most of intestine packed with woody fragments, stems, leaves, ete., and a small portion of gut at anterior end filled with Gastropod shells.
Intestine nearly 1.5 times body length.
Zuniga, 1941, expresses the relative length of the intestine of several species as follows: O. pentlandii, vegetarian, 1.39; O. neveur?, piscivorous, 0.5; O. tschudii, crustaceans, 1.89; O. luteus, mollusks, 1.92. (From Biol. Abstr., article not seen by me.)
THE SPECIES OF ORESTIAS As the synonymies will show there has been some confusion in the literature and numerous doubtful identifications, arising from the fewness and the imperfection of specimens, the brevity and ambiguity of descriptions, and the inadequacy of the illustrations. My study began in the hope of reducing the number of species materially. Such a number of species in waters so intimately connected seemed doubtful. But, as a matter of fact, one by one they have turned up either in our collections or, in two cases, specimens borrowed. They appear not to have had the advantage of isolation during their long history, but perhaps we can explain the evolution of so many species, mostly in one drainage area, from the long series of more or less violent changes in the habitat, their rise from sea level to as much as 12000-13000 feet, diastrophic upheavals, volcanic activity, and the change from saline to fresh water. Perhaps the species of Orestias have been the most bombarded of all fishes, with the greatest variety of external forces, some of which may have had genetic effects, even with our present understanding of genes. Confusion of species has also arisen from the choice of characters by which to separate them. The various authors have resorted to such characters as these: Finray counts which frequently overlap, never varying much Position and form of fins uniform for most species Teeth descriptions nearly identical Naked areas not consistently present or absent Depth of cleft of mouth hard to determine for lack of points of reference Striation of scales often not as given in deseriptions
SDYSILOJOL T [eoneqjodAy payeos 104 -U9A ‘O}BLIYS “yjoouls sa[Bos peytpou 23 :
Sa[BOS autos eee epun __toquaa ye Spayeu tau ~pad Suny sd ysadaep Apoq {]BOUIOA UNTO S4.20ys opounpad Aysay sdry | a 2 Ypjoquin fanbyqo yynour er RLOquany ‘poyutod gDIUe ‘ ' suo] peoy Spepuno. Apoq !
a SSa[Bos | ey I; ae) P Jo worjRoy snapsfidwua -1pour yysiys uryy pasiejue Japlnoys “peoy TenUBD []eus sayeos opt __uo paytu iL asAvy Sy, Sdepuas SyRoa | soyBos =109 “pasive[uo Japynoys puw | Oye + onIpD0 CaIPaG | | Yoouls sayeos pBay Jo sayros | nor Spoye ayeoundy -1apoOur sopBos uty fapotedo poyoue SULLO} aye SU0TA supbaja ye 4sousry, adeu Sysnqot SS1n0zU09 SULLOJ oys a punoa “ysnqor | | posreypue qonur aBIV] | F syed 101} AIOA Yynoul ayRNsuB YOU vizisspbp Cae ajafosqo suoTy =, x OFBl 3vee uv JO SaTwos ovis SSUO] AIA PBL ‘ 3 £ J it -epnmesd uty) suo ; peoy S1opues Apoq sdqy “ot | re }B.1Op' ul pBol] -radns yynout | Wauind | ; — Spatmgdn | ipunjjuad HEYA 4h ‘Opla peal apa ‘sutted =2y ‘oyeBUOya LUDSNprs [ ; yey UMOTO AGL ywon -ARPNGUG TOVUT XOAUOD UAMOID “\10ys UWOLsaL fomiey ao “4,10Ys you Ay § Sq1oys ynous s‘s miaqyut jepneo :
weet WLOYS } ynous £s1n0} :
daap “y10ys | -W09 popunor snayn) | | core [jews oso Spunos aotwodns yestop unyo “duo] ynous ynour Spaumy -aid payoue apa Jappnoys {A}IXOATIOD -dn ‘api ynous yout ‘poyutod qusreays ut yenba simo}4 SMO] [BSLopaad pvoy ‘possns 12 [BqquaA ‘xaAu0d -u0d [Bsop { ‘doap sv opin ~~ MOJO [Bs1op pue Baja | | | SNIIDAY]O snq]vD Luwano | | doap AraA apie] aX !tapuals “Q1oys opoun apunped pep -pad !
passoad neo Sayoyduioo -W09 S parpoq ATO LOU -daap AIoA S18] ‘X9AUO0 AUDA -nsue ites INO}UOd [BSLOp | Lpnyas} vaissne SVIUSAU() SAND) AHL AO ATU, IWOWOTOANTY “TTA WTAVL
354 FISHES OF WESTERN SOUTH AMERICA KEY TO THE SPECIES OF THE GENUS ORESTIAS a.Scalesnearlyuniformthroughout;venterscaled;smallfishes b. Deep, compressed, pointed equally toward head and tail; greatest depth midway; lips hts) «hz Pe Rees crs Det shes 2 A Te a ne RE Teac ll IMENT BEd LN eM FP Bs mncae bb.Elongate;verticalchin,thin-lipped;caudalpedunclelong,slender;deepestabovepectoral AUTO eclectic et a EN eee ae? RADE EP? UA, ARUP ar oneal oh humboldtr aa. Seales enlarged and thickened cn anterior parts c.Scalesofthehead,opercleandshoulderenlarged,irregular,granular,especiallyinadults d.Fishesofelongateform;caudalpeduncleslender e. Mouth, teeth, and eye large (6.6 in long head); mouth wide, oblique; 20 scales in vertebral series...............0.0.0000ne sues ... cuvtert ee. Mouth smallish, nearly vertical; snout roundly pointed; head moderate, 3.7-4.0; nape with low arch; granulations not extreme; scale rows Irrepular= throuchouteylemo hie: ees ee ea eye eee, pentlandi dd.Fishesofshort,robustform;napearched;caudalpeduncleshortanddeep f. Eye large, 4.8-4.5, elevated, supraorbital jutting over eye; snout short, lessthaneye;mouthhorseshoe-shaped,vertical,thin;headangular; lateral angele acute; ‘crowm® flate.....e5..... snes ee eer luteus ff. Eye smaller, round; snout equal or exceeding eye; mouth oblique, fleshy; crownmoreorlessconvex;lateralanglebroad e¢.Mouthupturned,backnearlystraight;roundlycompressed;caudal bases sexpandediy: seme) ee eee ee en: olivaceus ge. Head a wide cone, moderately rugged, pointed at chin; wide as deep;abruptlycompressedcaudally;highpredorsalarch..albus ec.Enlargedscalesonanteriorpartssmoothandmoreorlesspolished,withoutgranulations h.Sealesofheadandshouldermoderatelythickenedonsmallareas;littlemodificationofgeneralscalepattern i. Head wide and upturned, mouth superior, thin-lipped; dorsum straight; body wtapermne. from eeadh tos taille 0s eae eee stlustand ii. Body outlines rounded; caudal peduncle abruptly constricted far back on body; head pointed; mouth oblique, lips fleshy. ........... empyraeus hh.Scalesofshoulderandheadmostlythickenedandcornified,formingamoreorless densearmorofclose-setbucklers j. Armor of anterior parts reaching to or beyond tip of pectoral fin; scales moderatelyspecialized;predorsalslightlyormoderatelyarched;scalerows little disrupted or erowded; body not broadened at shoulder k.Small,elongate,scalesmostlythin,littleconvex,deciduous..elegans kk. Larger, deeper, wider fishes; scales smooth, more permanent in older specimens 1.Archedpredorsal,highabovetheendofopercle,taperingthence to tail; robust, not expanded; chin truncate; mouth and snout wide, lips thin; fins small, fleshy........ agassiziv ll.Predorsalarchedequallywithventralprofile;spindle-shaped forwardfrommiddleofbody;chinroundlypointed,lips thick, preorbital portion of head long; tapering weakly Aovad,(VSLGliJORMABUAoeeaseaweesanosSueowentt jj.Bodydeepandwideatlevelofpectoralbasis;thicklyanddenselyarmored; thickenedandconvexscalestobeyondmiddlelevelofpectoral;chin vertical m.Elongate,mcderatelydeep;headinlength3.8;taperinggradually frcm the base of the pectoral fin to the tail............ miillert mm. Shorter form; deeper; head in length 3.3; depth in length 2.7-2.8 n. Dorsal and anal fins oblique and caudal peduncle thin and shallow; its depth in its own length 1.3.......... Jussier nn. Tapering gradually shoulder to tail; caudal peduncle very deep, short, its depth equal toitslength.......... tschudit
ANNOTATED LIST OF THE SPECIES 355 474. ORESTIAS INCAE Garman Plate XVII, fig. 1 Orestias incae Garman, 1895, Mem. Mus. Comp. Zool., XTX, 155; Eigenmann, 1910, Rept. Princeton Univ. Exped. Patagonia, III, 461; Evermann and Radcliffe, 1917, Bull. U. S. Nat. Mus., no. 95, 42. Lago Titicaca 3948 MCZ, 1, 55 mm., cotype, Moho, Lake Titicaca, Garman, January, 1875. As in the case of O. olivaceus the kind loan of this specimen enabled me to retain O. incae as a valid species, since my collections unfortunately do not include it. The single specimen before me is extremely short, high, and compressed. The dorsal profile is rather uniformly convex from the oeciput to the end of the dorsal basis, thence becoming sharply concave on the caudal peduncle; slightly concave on the head; ventral contour equally rounded from the vertical through the opercle to the caudal peduncle. Depth greater than the length of the head, 2.7 in the length of the fish without caudal fin, the point of greatest depth farther back than in any other species, being about midway of the body and toward the extremity of the pectoral (not at the first dorsal rays as stated by Garman), tapering uniformly toward head and tail. Head unusually rounded in cross-section; mouth oblique, lips fleshy, teeth scarcely perceptible; eye moderately large, greater than the snout; 3.75 in the head, which is 3.3 in the length. Caudal peduncle slender, 4 in the length of the fish.
(Garman) D. 16-18; A. 16-18; P. 18-20.
Scales small, striate, without granulations, belly completely scaled. Many scattered red-brown chromatophores about over the body, and broken rows of fine markings on all fins.
Not at all closely allied to albus and luteus as believed by Garman, except as to the short form and the colors, but nearer to the primitive stock and to humboldtii, lacking the specialized forms and scales of both the major divisions of the genus. Placed here with little hesitancy, since it differs clearly from the specimens, which I have identified as humboldti, in the much less slender form, greater depth rounded head, more oblique mouth with fleshy lips, and teeth less evident. 475. ORESTIAS HUMBOLDTI Valenciennes Plate XVII, fig. 2 Orestias humboldti Cuvier and Valenciennes, 1846, Hist. Nat. Poiss., XVIII, 233. Orestias cuviert Eigenmann and Higenmann, 1891, Proc. U.S. Nat. Mus., XIV, 65; Eigenmann, 1910, Rept. Princeton Univ. Exped. Patagonia, IIT, 461. Lake Titicaca 16078, 31, 53-100 mm., Moho bay, Allen, December, 1918.
This species was collected for Valenciennes by Pentland, and has not since been collected. My specimens from Moho bay were long mistaken for Garman’s incae until one of Garman’s cotypes was sent me from the Harvard Museum,
TABLE VIII.—Gernetic ANALYSIS OF THE SPECIES OF ORESTIAS 356 a Der hast ee x x | XS NX 4 Buoy SJoywropour wo} x x x |x|] x] X i MOLIBU YPM x x x x x x x xX e Jopus[s opaunpod x OS x Da SX aloes juosqe suoryenUBIs x FG NK ee ex x a Ayo} Ba9pour seed x x x x x x x x Suue AyLepnsue x x | X x | X x x ay EEE LS) x Ka xX x x MS OS x adany your | x x x x x Ss anbryqo qnou DSW Pe OOS | Galle Oe Sli Cl oe x S uy sd) xX | xX | xX PAW x YSN Se || OS x cd YOOUS So] vos Ae | ee aK XX i ayRaopow opounped x | x | OS lex SS ULLOJIUN So]BoOS x x = _ ayRTUOO WAOJ YSN PS |e eX x | X = JoyR1opour suo BpnuBAs x x co | oyesopour Aytiepnsue x x x x nN doop apunped | x x x< x = yen erie: (Pociees al el al ee «al Me AEX ax = pouoyoryy sdy = -dn eas Ganou x x x x x 2 qrews yynour x NS x PO ae I OS TX ex n yeus eso xX | xX x | Xx DS eS eS Wes x 2° outaryxo AjIBpnsuR KK SK i Ayyeoad igeporl x x ell x + | oWadjxe suUOTyE[NUBID | xX x | O° apn Ayo} R.ALopout x x x x x a doop sv opin) xX x | | 4 gOS WILO} | d< |) <7) OX x PS | 2% I 2X ale vAae oe ane z S/S |S) els \s)2| 2 S| s| 2 2 sieie 2 Sune es | Soe pS A IS Ste | © fo) S R Q, = = ie 3 7 Sa cans |