THE NATURE AND ORIGIN OF THE FISHES OF THE PACIFIC SLOPE OF ECUADOR, PERU AND CHILI.1

(PLATES VIII-X)

BY CARL H. EIGENMANN.

(Read April 22, 1921.)

There are three distinct faunas on the Pacific slope of South America and west of the Cordillera of Bogota.

One of these, the richest, occurs in Panama, southeast of the Canal Zone, and in Colombia, west of the Cordillera of Bogota.

The second of these faunas occurs in the Guayas basin of Ecuador, and trails southward at least to the Rio Rimac at Lima.

The third occupies the Pacific slope of Chili trailing northward to the Rio Rimac or possibly the Rio Santa.

The first two are part of the tropical American fauna. The third belongs to the south temperate fauna.

In a series of articles I have dealt with the nature and origin of the freshwater fishes of Panama,2 of the Pacific slope of the Cordillera Occidental of Colombia, of the Magdalena river basin,1 of the Cordillera of Bogota, and of the general horizontal distribution of all of the fishes west of the basins of Lake Maracaibo and Titicaca, and west of the Orinoco and Amazon river basins.

1 Contribution from the Zoological Laboratory of Indiana University, No. 181.

2" The Freshwater Fishes of Panama East of 80° West," Indiana University Studies, No. 47, pp. 3–19, 1921.

3" The Fishes of the Rivers Draining the Western Slope of the Cordillera Occidental of Colombia," 1. c., No. 46, pp. 1–20, 1921.

4" The Magdalena Basin and the Horizontal and Vertical Distribution of its Fishes," 1. c., No. 47, pp. 21-34, 1921.

5"The Fish Fauna of the Cordillera of Bogota," Journ. Wash. Acad. Sci., X., pp. 460-468, 1920.

6" South America West of the Maracaibo, Orinoco, Amazon, and Titicaca Basins, and the Horizontal Distribution of its Freshwater Fishes." Indiana University Studies, No. 45, pp. 1-24, 1920.

These articles deal with the first of the faunas mentioned above. I propose in the present paper to deal with the nature and origin of the freshwater fishes of the Guayaquil basin of Ecuador, of the rivers and lakes of Chili north of Puerto Montt and of the rivers between these areas, i.e., of all the Pacific slope rivers between northern Ecuador, near the equator, and Puerto Montt, 41° 28′ S. The material for this study was collected by Mr. Arthur Henn of the Landon Ecuadorian Expedition of Indiana University in 1913, and by Dr. William Ray Allen, Dr. Adele Eigenmann and myself of the Irwin Expedition of Indiana University and the University of Illinois between June, 1918, and June, 1919.7

THE GUAYAS BASIN AND ITS FISHES (PLATE VIII.).

North of the desert of Tumbez the coast range of Ecuador consists of cretaceous formations trending from Guayaquil northwestward and reaching a height of 2,300 feet. North of about 1° 50′ south latitude the cretaceous joins tertiary hills reaching a height of from 600 to 1,000 feet and extending north to the Rio Santiago. North of the Rio Chone the hills approach the coast and are relatively younger (late tertiary and quaternary). South of the Rio Chone a wider or narrower quaternary territory extends

Several additional rivers should receive consideration at the earliest moment, the Esmeraldas in Ecuador; the Santa, the largest river of Peru; the lower Loa, an isolated river in northern Chili; and the Bio Bio, the largest river in Chili. Between Puerto Montt and Cape Horn there is a series of large rivers and lakes concerning which we know nothing. It has been suggested that they be searched for Cerotodus.

The Esmeraldas drains the area immediately north of the Guayas basin. Very little is known of its fauna. The Santa is the largest river of the Pacific slope of Peru and may be expected to contain a more complete complement of the ancient fauna of the Pacific slope of Peru than any of the rivers examined. The Loa in northern Chili is widely separated by deserts both from the nearest rivers to the north and the nearest rivers to the south. Its fishes, if there are any, should determine whether this portion of Chili belonged in the past to the tropical American faunal area or to the Patagonian.

The Rio Bio Bio, the largest basin in Chili, contains all of the fishes found in the rivers of Chili north of it and is "farthest north" of the peculiar fauna with Australian affinities which finds its culmination in the south of Chili.

8

along the coast to the Gulf of Guayaquil. East of the coast range, between the Guayas and the Esmeraldas rivers, lies a quaternary plain 60 to 250 feet high and 30 to 50 miles wide. East of this rises the Cordillera Occidental. About Guayaquil and along the coast to the Rio Tumbez there are recent alluvial flat lands. The coastal quaternary lands are drained by the Rios Chone and Portoviejo, two short rivers emptying into the Pacific between 30′ and 40′ south. The interior quaternary lowlands are drained in the north by the Quininde and Toachi which empty into the Esmeraldas. The greater, southern part is drained by the affluents of the Rio Guayas, the Chan Chan, Chimbo, Caracoles, Vinces and Daule, with their tributaries. I have elsewhere compiled an account of the Pacific slope of Ecuador and need repeat only that the height of land between the Rio Esmeraldas and some of the head-waters of the Guayas is negligible as a barrier to fresh-water fish migration. The usual ability to drag a boat from one system to the other is reported. (Sievers, "Süd und Mittelamerica," p. 459.) The fishes should, therefore, be the same in the two streams.

The Guayas basin differs from all others south of the Rio San Juan of Colombia. The Guayas and the Vinces have a course parallel with the general trend of the Andes and for many miles flow through lowland. Farther north the Dagua and the Patia, and

8 Mr. Arthur Henn, who visited this region, reports:

"As I recall the rivers at Chone and Portoviejo, they are approximately of the same size. These rivers arise in seasonally humid hills of about 1,200 feet in height. The coastal area is quite arid. At Bahia when I was there they were bringing in drinking water by tank car on the railroad from the interior, and Manta on the coast west of Porto viejo is almost an absolute desert. Water for cooking and washing was secured by means of wells sunk in the sand. Water filtered in from the sea and was slightly saline. Drinking water was brought in by donkey from the interior. South of Manta, near Santa Elena, the same desert continues with vast arid dunes. At Chone and Porto viejo which are more in the interior, the people can always secure water from pools in the river. During the rainy months these rivers are wide, deep, muddy streams and I believe it is possible to get up in a small launch from Bahia to Chone. When I was there, however, the rivers were small creeks with fords at numerous intervals, and numerous bamboo bridges which would be washed out in time of flood. The only water over waist-deep was the occasional deep pool. I understand from the people that in times of severe drought the river disappears entirely except for these occasional pools."

farther south the Jequetepeque, Rimac, etc., flow from the Cordilleras directly to the sea. The Guayas basin lies between the great coastal desert of Peru and the extremely wet region of southern Colombia. It is the largest river basin of the Pacific slope north of Chili.

The fish-fauna of the Guayas basin is old and highly specialized. It gradually tapers off southward as one and another genus and species have been excluded or exterminated by the elevation of the Andes, resulting in torrential courses, high seasonal variation in the amount of water and great fluctuations in the amount of silt carried, all conditions unfavorable to fishes.

The fishes on the Pacific slope of Colombia are quite distinct from those of Ecuador and Peru.

The fishes of the Patia, in southern Colombia, are essentially like those of the San Juan, in central Colombia. On the other hand, the fishes of the Guayas and the San Juan basins, the former in the wet zone, the latter in the dry zone, are essentially different. The line separating the two faunas lies somewhere near the Mira. and Esmeraldas basins.

Little is known of the fishes of the Esmeraldas. Several notable contributions to the fauna of the Guayas have been published. One of these is Kner and Steindachner's "Neue Gattungen und Arten von Fischen aus Central-America gesammelt von Prof. Moritz Wagner" (Abhandl. k. bayer. Akad. Wiss., München, X., 1864). The general bearing of the facts was discussed by Wagner, in the same volume, pp. 93-109. Another is Steindachner's "Zur FischFauna des Cauca und der Flüsse bei Guayaquil" (Denksch. K. Akad. Wiss., Wien, XLII., 1880, pp. 55-104, 9 plates).

A third contribution of note to this region is: Boulenger's Poissons de l'Equateur."

In this paper Boulenger describes or lists the specimens collected by Festa in a trip across southern Ecuador from Santa Elena through the Guayas basin to the Rios Zamora, Santiago and Bamboiza rivers. The last three are parts of one system tributary to the

"Viaggio del Dr. Enrico Festa nell'Ecuador e regione vicine,” Bolletino, Musei Zool. Anat. comp. della Univ. di Torino (Première Partie), XIII., No. 329, Dec. 2, 1898; (Deuxième Partie), XIV., No. 335, Feb. 15, 1899.

Marañon, in eastern Ecuador. He refers in the same paper to Brycon atricaudatus collected by Rosenberg at Paramba in the Mira basin, and describes as new Tetragonopterus simus from the Chota valley in northern Ecuador (Mira basin).

The localities given for a number of species make it doubtful whether the identification is correct or whether the locality label with the specimen has not been misplaced. It is questionable, for instance, whether Astyanar fasciatus was taken in the Peripa, whether Brycon atricaudatus came from east of the Andes, whether the specimens identified as Salminus affinis from east of the Andes are the same as the Salminus affinis from the Cauca River; whether the Leporinus from the Vinces is the frederici of the eastern rivers, whether the specimens listed under Brycon striatulus from the Rio Santiago actually belong to that west coast species; whether the Pygidium tænium is the same as the tænium from the elevated portions of the western slope of the Andes; whether, finally, Chatostomus dermorhynchus is actually present both on the Atlantic and the Pacific sides of the Andes of Ecuador.

Sixty-odd species of fresh-water fishes have been taken in the Guayas basin. Of these the species of Astroblepus and Pygidium belong to the highest altitudes and follow their own laws of dispersal. Twenty belong to families and genera that are found only in the lowland, some of them indifferently in salt or fresh water (Hexanematichthys, Stolephorida, Hamulidæ, Tylosurus, Centropomide and Gobiida).10 This leaves of strictly fresh-water species only forty. Of the forty only Sternopygus macrurus is certainly found east of the Andes. Only four of the species extend north of the Rio Esmeraldas.12 These are Chatostomus fischeri north to the Chagres; Chatostomus marginatus north to the San Juan;

11

10 They are marked with * in the table.

11 Hemibrycon polyodon is recorded from the basin of the Santiago east of the Andes. It is quite within possibilities that the specimens recorded from Guayaquil also came from the east. Pellegrin has recorded Ancistrus bufonius from the Rio Pove at Santo Domingo de los Colorados, 560 m. I am not sure whether the Pove drains into the Esmeraldas or the Guayas, but the identification may be doubted for the present.

12 Brycon atricaudatus has been recorded from the R. Mira just north of the Esmeraldas.