Interciencia Asociación Interciencia [email protected] ISSN (Versión impresa): 0378-1844 VENEZUELA 2000 Barry Chernoff / Antonio Machado Allison / Phil Willink / Jaime Sarmiento / Soraya Barrera / Naercio Menezes / Hernan Ortega FISHES OF THREE BOLIVIAN RIVERS: DIVERSITY, DISTRIBUTION AND CONSERVATION Interciencia, septiembre, año/vol. 25, número 006 Asociación Interciencia Caracas, Venezuela pp. 273-283 Red de Revistas Científicas de América Latina y el Caribe, España y Portugal Universidad Autónoma del Estado de México FISHES OF THREE BOLIVIAN RIVERS: DIVERSITY, DISTRIBUTION AND CONSERVATION BARRY CHERNOFF, ANTONIO MACHADO-ALLISON, PHIL WILLINK, JAIME SARMIENTO, SORAYA BARRERA, NAERCIO MENEZES and HERNAN ORTEGA he forested lowlands of eral ichthyological results, including a Fishes were collected us- most Amazonian tribu- breakdown by sub-basin; (ii) compare the ing a variety of nets and netting tech- taries remain poorly ex- diversity with those of other South Amer- niques. Each group was equipped with plored. Lists of freshwater fishes do not ican regions and discuss any distribu- seines (5m x 2m x 1.25cm, 5m x 2m x exist for more than one or two of the tional aspects of the data; (iii) discuss the .63cm, 1.3m x .7m x .37cm), dip nets and major arms of the Amazon, a feature that economic and conservation importance of experimental gill nets (40m x 2m, mon- was prominently absent from an other- the results (iv) consider threats to the ofilament, with five 8m panels, mesh size wise remarkable work on the fishes and species richness; and (v) to propose rec- from 1.25cm to 6.25cm). Team 2 pulled an fisheries of the Rio Madeira (Goulding, ommendations for conservation. otter trawl (mouth 3m wide) with two 15kg 1981). In Bolivia there have been at- doors where the depth of the water was tempts to bring together information for Methods >2m over sandy or muddy stretches in a Amazonian streams (e.g., Lauzanne et al., manner modified from that described by 1991). More recently, a well documented Two teams, three-four Lopez-Rojas et al. (1984). Additionally, fauna from the Rio Gaupore/Itenez was persons each, made collections between 4 one of the river pilots threw a 2m cast net provided by Sarmiento (1998). and 21 September, 1996. The teams were in some deeper lakes or cochas. Lauzanne et al. (1991) usually isolated and worked in different Fishes were preserved recorded 389 species of freshwater fishes tracts of the Upper Rio Tahuamanu for in buffered 10% formalin solution. All from the Bolivian Amazon to which Sar- the first five days. Eighty-five collection specimens captured at the same place and miento (1998) added an additional 21 stations were established, each receiving a time were maintained separately from all species bringing the total to 410 species. unique, sequential field number. The field other collected specimens. Larger speci- This work reports upon the ichthyologi- stations were enumerated separately for mens were tagged individually using fine cal results of the first AquaRAP carried each group, identified as P1 and P2. At wire and punched cardboard tags and ei- out in a relatively small area of the upper each station, longitude and latitudes were ther placed in large liquidpacks contain- Rio Orthon. Two teams of ichthyologists registered from hand held GPS units. ing formalin with other specimens or surveyed aquatic habitats in the rivers At each field station a were skeletonized, soaked in 40% isopro- Tahuamanu and Manuripi and their tribu- number of ecological variables corre- panol and dried. All material was wrap- taries over 17 days in 1996. Amazingly, sponding to a description of the habitat ped and shipped to Chicago for sorting, 313 species of fishes were recorded. This were recorded. These included the shore, identification and enumeration in the Di- value includes a spectacular number of substrate, type of habitat (e.g., river, lake, vision of Fishes, Department of Zoology, new records for Bolivia and several spe- flooded area, etc.) as well as the water Field Museum of Natural History cies new to science, including a new pi- type. The classification of water type (FMNH). Fifty percent of the specimens ranha, genus Serrasalmus. (black, white, turbid) was checked with are housed in the Museum of Natural The purpose of this con- the results obtained by the limnology History, La Paz, Bolivia; the remaining tribution is to: (i) make known the gen- group. specimens were shared among the partici- KEY WORDS / Fhises / Diversity / Distribution / Conservation / Bolivia / Barry Chernoff and Phil Willink: Field Museum of Natural History, 1400 Lake Shore Dr. Chicago, Illinois 60605-2496, USA. e- mail: [email protected] Antonio Machado-Allison: Laboratorio de Ictiología, Instituto de Zoología Tropical, Universidad Central de Venezuela, Caracas, Venezuela and Field Museum of Natural History. e-mail: [email protected] Jaime Sarmiento and Soraya Barrera: Museo Nacional de Historia Natural, La Paz, Bolivia. Naercio Menezes: Museu de Zoologia, Universidade de São Paulo, São Paulo, Brasil. Hernan Ortega: Museo de Historia Natural (UNMSM), Lima, Perú. SEP 2000, VOL. 25 Nº 6 0378-1844/00/06/273-11 $ 3.00/0 273 pating institutions: FMNH; Museu Zoolo- gia do Universidade do São Paulo, Bra- zil; Museo de Historia Natural San Mar- cos, Perú and Museo de Biología de la Universidad Central de Venezuela. The identifications were made in a careful but relatively rapid fashion. General works such as Eigenmann and Myers (1929) or Gery (1977) were used but preference was always given to systematic revisions (e.g., Vari, 1992; Mago-Leccia, 1994) and recent species descriptions (Stewart, 1985) if available. In many cases speci- mens were compared to types or historic material referenced in the literature and housed at FMNH. However, some identi- fication to the level of species or even genus was not possible. To do so would represent a less than scholarly approach to the taxonomy. Instead we rely upon Figure 1. General Map of the Rio Orthon Basin, showing the Collecting Stations on the Manuripi morphospecies – the number of distin- ( ) and Lower Tahuamanu (+) rivers (11º8-10'S to 67º33'W) guishable entities present in our samples. This bears the assumption that such The Rio Nareuda is a major tributary of well as in the Lower Tahuamanu, just discernable entities or morphospecies are the Rio Tahuamanu. above its mouth in the Rio Manuripi. River putative taxonomic entities (i.e., species). The gear effort used to conditions permitted trawling. We were careful to check for sexual and make the 85 field stations, were as fol- ontogenetic differences. All specimens lows: seine = 73; trawl = 6; gill net = 5; Results and Discussion were examined critically and identified to cast net = 2. This adds up to 86 because their lowest taxonomic level (Table I). one station P2-04 included both gill net Diversity and distribution: General Another issue inheres and seine collections. The gill nets were in the appropriate selection of taxa across usually set for several days. Because no The species richness of lists in order to judge new additional or striking differences were noted for day fishes was spectacular. A total of 313 new records. We chose a conservative and night samples within the gill nets, species were captured and identified. Be- approach and did not include all of the they were recorded as single stations. cause an additional 5% that could not be taxa that we had collected. We elimi- Due to low water con- identified unambiguously were discarded nated from comparison about 5% of ditions during the time of the field sam- from further evaluation, the actual num- those taxa whose identifications were am- pling, motors had only limited use in the ber of species in the entire region is even biguous or unknown in our list and in Upper Tahuamanu and could not be used greater. The fishes (Table I) included published lists. So for example, if Hemi- in the Rio Nareuda except for in the vi- members of all trophic or activity groups, grammus sp. or H. sp. 1 occurred in both cinity of its confluence with the Rio ornamentals (e.g., Abramites hypselono- lists, it was not counted as a similarity or Tahuamanu. During the first portion of the tus), food fishes (e.g., Pseudoplatystoma a difference because there is no way to expedition team 1 worked in the Upper fasciatum), as well as miniatures (Scol- ascertain that the taxonomic designations Tahuamanu and also in the lower end of oplax cf. dicra, <20 mm SL) and large represent the same biological entity. the Rio Muymanu whereas team 2 col- fishes (e.g., Prochilodus cf. nigricans, However, Gephyrocharax sp. occurs in lected in the Upper Nareuda as well as in Doras cf. carinatus, >200 mm SL). our list but only G. chapare is reported a number of small streams (garapés) that Together Lauzanne et by either Lauzanne et al. (1991) or Sar- drained independently into the Tahua- al. (1991) and Sarmiento (1998) record miento (1998). In this case we count the manu. The Upper Tahuamanu and the Up- 410 species from all rivers within the G. sp. as a new record because we com- per Nareuda systems as well as their Bolivian Amazon. The 313 species of pared our material to G. chapare and it tributaries were surrounded by terra firme. fishes, therefore, represents a fauna is different. The possible error in this lat- For the middle portion greater than 76% of the number of spe- ter case is identical to the possible errors of the expedition, both teams were cies previously reported from all other in a list containing misidentified species camped together but team 1 focused Amazonian tributaries within Bolivia. bearing specific epithets or not. upon the Lower Nareuda down to its The number of fishes discovered in a mouth in the Rio Tahuamanu.