Environmental Biology of Fishes 42: K-23,1995. 0 1995 Kluwer Academic Publishers. Printed in the Netherlands On the habitat and isolation of Gymnocharacinus bergi (Osteichthyes: Characidae) Roberto C. Menni’ & Sergio E. G6mezz ’ Museo de La Plata, Paseo de1 Bosque s/n, 1900, La Plata, Argentina 2 Institute de Limnologia ‘Dr. R.A. Ringuelet’, Casilla de Correo 712,1900, La Plata, Argentina Received 13.4.1993 Accepted 3.2.1994 Key words: Thermal water, Water chemistry, Zoogeography, Fish ecology, Patagonia, Neotropical, Ostario- physi Synopsis Gymnocharacinus bergi is a rare Paranensean fish which is the only characiform almost lacking scales in the adult. It is endemic and the only species in a peculiar spot-a tributary of the Valcheta creek-in the Somuncu- ra plateau in northern Patagonia, Argentina, over 300 km from the nearest place with a paranensean fish fauna. Besides its geographical isolation, G. bergi occurs within an area with climatic features drastically different from those currently associated with fishes from Neotropical temperate zones. We tested the as- sumption that water temperature in the naked characin habitat do not agree with the northern Patagonia climate. We also considered the isolation of G. bergi within the framework of an increasing inpoverishment of the paranensean ichthyofauna along a NE-SW axis in the Buenos Aires province. For this we applied a decre- ment equation used in island biogeography. Our findings demonstrate that the existence of G. bergi in its isolated habitat is possible because of the thermal traits of the water at the sources of the creeks, its temper- ature being independent of the climate of the area. The chemical composition of water was found to be within the range of common environments in the Buenos Aires ‘pampas’ inhabited by several species of Paranensean fishes. Geographically, G. bergi lives in the last of a series of habitats which show a decreasing number of species correlated with the increasing distance from the La Plata River. Conservation status of the species is briefly discussed. Introduction Gymnocharacinus bergi Steindachner, 1903, a unique characoid fish with almost no scales, has an isolated distribution in a small tributary near the headwaters of the Valcheta Creek in the Somuncu- ra Plateau in Argentina. The Valcheta Creek is in northern Patagonia (40°50’ S, 66”40’ W), 329 km from the nearest locality with paranensean fishes, which is the Colorado River. Thus this species is the southernmost characid of the world. Its small pop- ulation (personal observation), isolated distribu- tion and still unsolved phylogenetic relationships (Gery 1978) makes G. bergi an appealing fish for study. Lopez et al. (1981,1982,1987,1989) quoted no less than 24 references in which the species is men- tioned. Important contributions by Gery (1978) and Miquelarena & Aramburu (1983) referred to oste- ology. Ltiling (1978a, b, c, d) dedicated several pa- pers to the natural history of G. bergi. Pozzi (1936) included a detailed description, measurements and a photograph. During a 1979 trip when the material cited by Mi- Atlantic Ocean Fig. I. Province of Buenos Aires and Northern Patagonia (Argentina) displaying Pampasic 'lagunas', Sierra de la Ventana and the Somuncurh Plateau. Locality numbers refer to Table 1. quelarena & Argmburu (1983) was collected, the senior author gathered samples and data about wa- ter characteristics from the naked characin habitat. We analyzed temperature and chemistry of water from the G. bergi habitat, to determine if this water is thermal and could explain the presence of the species in its isolated location. A comparison was also made with the water chemistry and ichthyofau- na of Sierra de la Ventana and other localities with- in the Buenos Aires province to consider the distri- bution of G. bergi in a wider faunal and ecological background. Sierra de la Ventana is the nearest area with a variety of lotic habitats occupied by several Paranensean species. Material and methods Topographic references are based on Military Ge- ographic Institute charts N 4166-14 and 4166-15. Linear regressions between pairs of variables were calculated by the least squares method, and the cor- relation coefficient according to Sokal & Rohlf (1979). An exponential decrement equation (Mac- Fig. 2. Annual variation of temperature (” C), air (M, MM, AM) and water (A, J) in Sierra de la Ventana (left) and the Somuncura Plateau (right). 1 and J localities without G. bergi. M = monthly mean maximum, MM = monthly mean, AM = annual mean. A = April 1983 (Menni et al. 1988), B = May 1987 (G6mez pers. comm.), C = November 1979 (Menni et al. 1988), D = ‘22.5 to 22.6” C all the year’, Valcheta Creek (Liiling 1978d), E = October 1975, Valcheta Creek (Liiling 197&), F = ‘20 to 22” C, also in winter months’, El Rinc6n creek (Cei 1980), G = November 1979, Estancia El Rinc6n y Chipauquil (Miquelarena & Aramburu 1983), H = April 1968. Quebrada de El Rinc6n (Cei 1%9), I = March 1968, Raimunda pond (Cei 1969), J = December 1967, Raimunda pond (Cei 1969). Arthur & Wilson 1967, Pianka 1982), was used to describe the impoverishment of the Parenensean fish fauna along a NE-SW axis in the southern Pam- pasia. Air temperature data were obtained at the Maquinchao station, at 800 m above sea level (a.s.l.), an elevation very close to that of Valcheta Creek sources. Climate data for Sierra de la Venta- na are values for the city of Sierra de la Ventana station. Chemical analysis were done at the Water Chemistry Laboratory of the Instituto de Limnolo- gia de La Plata, according to APHA (1985)3 meth- ods. The Somuncura Plateau (Fig. 1) (Cei 1969) has a surface of 1.5 000 km*. It is roughly ellipsoid and is located between 40’50’ and 41”45’ S and 66” and 68” W. It was formed as a result of volcanic activity during the Mio-Pliocene. Neither the Entrerriana ingression nor the glaciations seem to have affected its characteristics. Three altitude levels or floors are recognized within the plateau, related to the mor- pho-hydrological features: the volcanic heights, be- tween 1500 and 2 000 m; the lowland floor of the 3 American Public Health Association. 1985. Standard methods for the examination of water and wastewater. American Public Health Assoc., Washington. clay ponds between 900 and 1500 m; and the ‘que- bradas’ or filter creeks floor, between 500 and 900 m. The Valcheta Creek headwaters, where G. bergi lives, are placed at the latter level, which is al- so inhabited by the endemic (Cei 1969,198O) anuran Telmatobios somuncurensis. The naked characin specimens studied here were collected at few me- ters of tributary origin. The source of the water is a limnocrene in the sense of Ringuelet (1962), i.e. sur- facing waters forming first a small lentic water body which then produces a small lotic course. The fauna of the Somuncura Plateau has a zooge- ographic relict character, because of the degree of adaptation and high endemism, particularly evi- dent in the Telmatobiinae (Amphibia). The plateau vegetation is ecotonal between the Patagonic and Monte formations (Cei 1969). Results The habitat of Gymnocharacinus bergi Water temperature We consider that the water at the Valcheta Creek tributary inhabited by G. bergi is thermal in charac- Fig. 3. Regression lines based on data for localities 1 to 8 in Table 1 (Loc. 9 = Valcheta Creek headwaters). Dotted line: regression line between number of permanent species (n) and mean annual temperature (t) for eight localities (open circles) in the Buenos Aires province. Solid line: regression line between number of permanent species (n) and the logarithm of the distance (km) to La Plata harbour (d) for the same eight localities (solid circles) in the Buenos Aires province. ter. Cei (1969) mentions the presence of G. bergi in the creek origin at El Rincon neighborhood (700 m height), stating that at this level temperature during April between 15:OO and 17:OO h was 15-16" C. Our temperature data at the Valcheta Creek were ob- tained during spring, but appeared high for the ar- ea; so Miquelarena & Aramburu (1983) describe it as 'slightly thermal'. Figure 2 summarizes all available information on air and water temperature for Sierra de la Ventana (left) and Valcheta area (right). Solid lines show monthly maximum means, monthly means and an- nual means air temperature values. Dotted boxes (and dotted line) represent water temperature val- ues. The range of temperatures at Sierra de la Ven- tana are very close to the monthly mean air temper- ature throughout the year (observations for April, May and November: A, B and C, left). This situa- tion is expected when only climatic local factors de- termine water temperature, which is not the case in the Valchete Creek. Data on water temperature at Valcheta Creek, in localities where G. bergi occurs (right, D, E, F, G and H), are similar to, or higher than, the maximum mean monthly air temperature. Data are known for April, October and November. Liiling (1978c, d) states that water temperature is 22.5-22.6" C during the entire year. Cei (1969) ex- plicitly says that water temperature is 20-22" C 'al- so in winter months'. Thus we conclude that water in Valcheta Creek is of thermal origin. This conclu- sion is strengthened by the fact that a lagoon at 1400 m a.s.1. also in the Somuncurh plateau, has wa- ter temperature values below the monthly mean (right, I, J). Number of species, distance, and temperature Past ecological conditions in Patagonia had a marked temperate and moist character, providing conditions for a wider distribution of the Brazilian or Paranensean fauna (Cei 1969, Ringuelet 1975, 1978,1981). With increasing aridity and lower tem- peratures in north Patagonia and the southern Pampasia, the Brazilian fauna suffered a marked retraction. This is confirmed for many animal groups (testacean amoebae, leeches, aquatic oligo- chaetes, bivalves, gastropods, harvestman (Opi- liones), many aquatic and terrestrial orders of in- sects, crustaceans, fishes, amphibians, birds and mammals) by historical, paleontological and eco- logical data (Ringuelet 1978,1981). The presence of G. bergi can be related to the present ecological conditions along an environmen- tal gradient across the Buenos Aires province. These conditions were analyzed from two points of view: the relationship between species number and temperature, and the relationship between the number of species and the relative distance to a point with a richer fauna. We selected as this point the Rio de La Plata, which is the source of renewal of fish populations for a variety of environments in the Buenos Aires province under different condi- tions (communication channels, flooding). Exam- ples of this are man-made ponds studied by Menni & Almiron (personal communication), and the use of the Salado River by many species as a population corridor to Bonaerensean 'lagunas' (Menni 1984). For this analysis eight localities are considered, for which the number of permanent fish species is well known (Fig. 1, Table 1). Application of the ex- ponential decrement equation to the number of fishes in the eight localities, taking the Rio de La Plata as the distance origin, gives the following rela- tionship between the number of species (n) and the log distance (log d): n = 71.84 - 24.01 (log d), with r = - 0.991 (p < 0.01). The equation predicts that the 0 number of species is reached at 981 km from the origin point. The headwaters of the Val- cheta Creek are at 992 km (Fig. 3). Figure 3 also shows the relationship between the number of species and the mean annual temper- ature for the eight localities in Table 1. Los Talas (35" S) is the nothernmost point, and the drainage channels of the Colorado River (39" S at the mouth) are the southernmost point (Fig. 1). Using again n as the number of species and t as the mean annual temperature, the relationship is: with r = 0.9499 (p < 0.01). The dotted line in Fig. 3 shows the decreasing number of species with decreasing values of tem- perature. Lower temperatures correspond to south- ern localities. Display of Valcheta Creek values (point 9) shows that they do not agree with the cal- culated equation. This demonstrates that the spe- cies number at Valcheta (one) is not the expected for the normal air temperature there (annual mean 9.6" C), a value at which no Paranensean fish can live permanently. The equations predicts zero spe- cies at about 14.4" C mean annual temperature, which is lower than that of Sierra de La Ventana. All mean values for all localities are slightly above this one, while those for Valcheta are well below (Table 1). Similar conclusions are reached using minimum annual means (r = 0.9046) and the mean temperature of the coldest month of the year (July, r = 0.8935). Water chemistry Water in the Valcheta Creek section inhabited by G. bergi is hypohaline. Due to the C a w and Mgi-t con- tent it is moderately soft (total hardness 111 to 118.5 mg C0,Ca 1.'). It is of the C0,H -, Na +, Ca tt, Mg tt type, slightly alkaline and with a rela- tively high buffering power in relation to the inorga- nic carbon system (alkalinity 3.39 meq 1.'). Data available on Valcheta Creek water chemis- try include a set from Liiling (1978d) and two sets from samples obtained by RCM during November 1979. These data are in Table 2, together with in- formation from Sierra de La Ventana in the high- lands area (Menni et al. 1988), from the middle Par- an8 River (Bonetto & Lancelle 1981), from the Vitel pond (Olivier 1961), and from the Chascomus pond (Conzonno & Claverie 1990). These data are in- cluded because Sierra de La Ventana creeks are considered as the nearest environments to Valcheta inhabited exclusively by paranensean fishes. The Vitel and Chascomlis 'lagunas' (permanent ponds, lakes of third order), are typical pampasic environments (Ringuelet 1972, Ringuelet et al. 1967b). Table 3 shows percentage similarity values among the considered environments, based in the number of factors with overlapped ranges. The pre- Table 1. Number of permanent species (n), linear distances to La Plata harbour (d, in km) and air temperature in " C (tl = annual mean, t2 = minimum annual mean, t3 =July mean) for eight localities in the Buenos Aires province (1 to 8) and the headwaters of the Valcheta Creek (9). Locality n d t l 12 t3 1. Los Talas 2. Chascomus 'laguna' 3. Salada Grande 'laguna' 4. Alsina and Cochico 'lagunas' 5. Sierra de la Ventana 6. Naposta Grande Creek 7. Chasico creek 8. Drainage channels Colorado River 9. Valcheta Creek headwaters sent analysis shows that there is no chemical factor with values exclusive for Valcheta. Though water temperature in Valcheta creeks obviously depends upon geological factors, its chemical composition does not differ markedly from water of Vitel and Chascomus. It is unlikely that the chemical composition of water there has a significant influence on fishes; any more or less eu- rytopic paranensean fish species could tolerate these values (Menni et al. personal communica- tion). This chemical 'normality' of Valcheta water, indicates that although its temperature is high enough for G. bergi to live there, it does not alter the geological substrate and modify its own composi- tion. Ionic content of Valcheta waters are very far from C1-, SO,- and Na +values of reached by high- ly thermal waters (idiotrophic in the sense of Ring- uelet 1962) from a locality in NW Argentina - with several fish species - where values for those ions are 266.9,556.9 and 469.7 mg l", respectively. Thermal environments inhabited by cyprinodontiform fishes in U.S.A. (Deacon & Minckley 1974) show the fol- lowing ranges for C1-, SO, - and Na +, respective- ly: 8.86 - 21 000,40.8 - 9 658 and 12.9 - 24 440 mg 1.'. Discussion Some proposals, based on the assumption that the present distribution of G. bergi was not reached by dispersion appear of interest. Ringuelet et al. (1967a), and Ringuelet (1975,1978,1981), have ex- plicitly stated that the presence of G. bergi in north- ern Patagonia indicates a wider distribution of the Paranensean ichthyofauna in the past. Ringuelet supported the hypothesis, now widely accepted, that the increasing aridity of the climate was the ma- jor cause of the restriction of that fauna during the Tertiary. Typical Paranensean fish species are known from a Lower Pleistocene level, near Bahia Blanca (about 39" S), as Pimelodella sp. (nearly identical to f? laticeps) (Pimelodidae) and Callichthys sp. and Corydoras sp. (Callichthyidae) (Cione et al. person- al communication). A good indicator of the Para- Table2. Chemical data for Valcheta creek tributaries, creeks and 'lagunas' in the Buenos Aires province and the middle Paranh River. A = Paraje Chipauquil (Valcheta Creek), 16-11-1979. B = Establecimiento El Rinc6n (Valcheta Creek), 16-11-1979. C = Valcheta Creek, 25-10-1975 (Liiling 1978d). D = Creeks in Sierra de La Ventana, November 1979 (Menni et al. 1988). E = 'laguna' Vitel (Salado River basin), 1956-1957 (Olivier 1961). F = 'laguna' Chascomus (Salado River basin), 1983-1985 (Conzonno & Claverie 1990). G = Middle Paranh River (main course), 1968-1969 (Bonetto & Lancelle 1981). SC = specific conductivity US cm-'. TDS (total dissolved solids), anions and cations in mg 1.'. COD =chemical oxygen demand mg 0,l ' . Total hardness and total alkalinity in mg 1" of C0,Ca. For D, E, F and G ranges for the period are shown, but those with * are mean values. Lack of data as -. SC TDS pH co3- C0,H- C1- SO,-- Ca++ Mg++ Na+ K+ COD MgICa Mg + CalNa + K Total hardness Total alkalinity nensean character of this assemblage is also the presence of the aquatic turtle Hydromedusa tectif- era. The Patagonian cold temperate water genus Percichthys was also found there (A. Cione person- al communication). There are fossil characiforms in the Lower Eocene near the San Jorge Gulf in Pata- gonia (46" S) (Cione 1986). If the Callichthys sp. belongs in the same or a sim- ilar species to C. callichthys, mean annual temper- ature in the environment to which the fossil be- longed must have been no less than 16" C (Gdmez 1993). PI laticeps and Corydoraspaleatus still live in the area (Menni et al. 1988). The present boundary of the Paranensean fish fauna, as proposed by Mac Donagh (1934), is ap- proximately at the latitude of the noted fossil loca- tions, at the south of the Buenos Aires province, somewhat to the south of Sierra de la Ventana. This has been discussed in detail by Ringuelet (1975, 1981). The composition of the Sierra de la Ventana fish fauna is provided by Menni et al. (1988), togeth- er with distributional data and environmental par- ameters, supporting previous hypotheses of Mac Donagh (1934) and Ringuelet (1975). The Paranensean fauna (fishes included) in the Buenos Aires province exhibits a clear impoverish- ment roughly from NE to SW. Ringuelet (1975) sug- gests that this feature is due to decreasing temper- ature and the progressive salinization of the waters toward the SW of the area. This explanation was contended regarding salinity, but confirmed regard- ing temperature by Gdmez (1993), who obtained evidence of a strong correlation between the num- ber of species and the yearly mean temperature. In spite of the accuracy of these explanations in the present, historical factors as the climate changes, are also likely important to the distribution of G. bergi. The isolated distribution of G. bergi can be placed in the context of a progressive desertifica- tion of the area between the Buenos Aires province and the Somuncurfi Plateau. Consideration of this process as a vicariant event or as an insularity case does not produce any new information for the mo- ment. Weitzman & Weitzman (1982) stated that as- sumptions of past distributional limitations based on knowledge of those limitations in recent orga- nisms, although never fully reliable, can greatly help our understanding of the patterns to be dis- cerned by other methods. Our demonstration of the thermal character of the water explains, from an ecological point of view, why this species remains in its present habitat. This habitat is a peculiar combination of a Sierra de la Ventana type physiography, a pampasic type water chemistry, and a warm constant water temperature. Available information (Menni et al. personal com- munication) on the ecological ranges of several var- iables tolerated by Cheirodon interruptus and Je- nynsia lineata beg the question of why these species do not accompany G. bergi at its isolated spot. If the area covered by the creeks inhabited by G. bergi is near the minimum required for a freshwater species (Horn 1972, Welcomme 1979), perhaps this is the re- sult of previous competition. Restricted distribution and limited habitat of G. Table 3. Similarity between considered localities based in chemical characteristics. Similarity values computed as number of characters with overlapping ranges over total number of compared characters (in parentheses). Letters in parentheses refer to values in Table 2. Valcheta headwaters Sierra de la Ventana Vitel and Chascomds Maine course of the ( A + B + C ) creeks (D) 'lagunas' (E + F) Parana River (G) Valcheta headwaters (A + B + C) 1 0.25 0.62 0.13 (4116) (10116) (2115) Sierra de la Ventana creeks (D) 1 0.31 0.47 (5116) (7115) Vitel and Chascomds 'lagunas' (E + F) 1 0.27 (4115) Main course of the Paranh River (G) 1 https://www.researchgate.net/publication/50337599_Fisheries_Ecology_of_Flood_Plain_Rivers?el=1_x_8&enrichId=rgreq-e396fff2-87e9-4f62-9456-3a7a3a6bf0ab&enrichSource=Y292ZXJQYWdlOzIyNjU5OTcwMztBUzoxNjQ2ODI2MTA1MjAwNjRAMTQxNjI3NDgwMjY1Ng== https://www.researchgate.net/publication/259037243_Los_peces_continentales_del_Cenozoico_de_Argentina_Su_significacion_paleoambiental_y_paleogeografica?el=1_x_8&enrichId=rgreq-e396fff2-87e9-4f62-9456-3a7a3a6bf0ab&enrichSource=Y292ZXJQYWdlOzIyNjU5OTcwMztBUzoxNjQ2ODI2MTA1MjAwNjRAMTQxNjI3NDgwMjY1Ng== bergi implies the possibility of this species being af- fected by habitat alteration and/or competition. The first possibility is less probable, as human pop- ulation in the area is scarce and agricultural activ- ities meager. The second one deserves some atten- tion, because it appears that introduced salmonids live in the Valcheta River, a few kilometers down- stream of the G. bergi habitat. Wegrzyn & Ortubay (1991) reported informal data implying that G. bergi extended in the past downstream to the river, from where they have been displaced by trouts (species no cited), introduced for sport fishing. The naked characin is the only fish species includ- ed in a list of argentinean vertebrates in danger of extinction (Chebez 1993), qualified as an endan- gered species proposed to the Red Data Book. A proposal for the consideration of the Valcheta area as a Provincial Reservation, in order to protect G. bergi, has been submitted to the Rio Negro provin- cial government (Cussac personal communication). We think that the conservation status of this species should be treated with care. For the moment, an un- due emphasis on its protection could lead to an in- creased interest with undesired results. Acknowledgements We thank the staff of the Chemical Laboratory (IL- PLA-MLP) for analysis of our samples and V. Con- zonno for help and criticism. To R.H. Aramburu and H.L. Ldpez for assistance with field work and discussions on Somuncura features. A. Almirdn and M. Garcia for critical readings. We specially thank A.L. Cione for many suggestions and data on the evolution of climate in Patagonia. This work was funded by a Comisidn de Investigaciones Cien- tificas de Buenos Aires grant (field work) and by the Research Project PID 160 from the Consejo Na- cional de Investigaciones Cientificas. References cited Bonetto, A.A. & H.G. Lancelle. 1981. Calidad de las aguas del rio Parana medio. Principales caracteristicas fisicas y quimicas. Comunicaciones cientificas CECOAL 11: 1-22. Cei, J.M. 1969. La meseta basaltica de Somuncura, Rio Negro. Herpetofauna endemica y sus peculiares equilibrios biocenot- icos. Physis 28: 257-271. Cei, J.M. 1980. Amphibians of Argentina. Monitore Zool. Ital. (NS) Monogr. 2: 1-609. Chebez, J.C. 1993. Listado de vertebrados argentinos en peligro de extincion. Nuestras aves (Asoc. Omit. del Plata), Supl. 1: 1-14. Cione, A.L. 1986. Los peces continentales del Cenozoico de Ar- gentina. Su significacion paleoambiental y paleobiogeograf- ica. Actas IV Congr. Arg. Paleontologia y Bioestratigrafia, Mendoza 2: 101-106. Conzonno, V.H. & E.F. Claverie. 1990. Chemical characteristics of the waters of Chascomus pond (Provincia de Buenos Aires, Argentina). Limnological implications. Rev. Bras. Zool. 50: 15-21. Deacon, J.E. & W.L. Minckley. 1974. Desert fishes. pp. 385488. In: G.W. Brown (ed.) Desert Biology, Vol. 2, Academic Press, New York. Gery, J. 1978. Results of Dr. K.H. Liilingresearch in Argentina in 1975. The relationships of the naked characin Gymnocharaci- nus bergi (Pisces, Cypriniformes, Characoidei) from northern Patagonia. Zool. Anz. 205: 403409. Gomez, S.E. 1993. Susceptibilidad a diversos factores ecologicos extremos, en peces de la Pampasia bonaerense, en condiciones de laboratorio. pp. 165-174. In: Tesis Doctorales, Resumenes 1986-1990, Fac. Cienc. nat. y Museo, Univ. Nac. La Plata. Horn, M.H. 1972. The amount of space available for marine and freshwater fishes. U.S. Fish. Bull. 70: 1295-1297. Lopez, H.L., R.C. Menni & R.A. Ringuelet. 1981. Bibliografia de 10s peces de agua dulce de la Argentina y Uruguay 1967-1981. Biologia AcuAtica 1: 1-100. Lopez, H.L., R.C. Menni & R.A. Ringuelet. 1982. Bibliografia de 10s peces de agua dulce de la Argentina y Uruguay. Suplemen- to 1982. Biologia Acuatica 3: 1-26. Lopez, H.L., R.C. Menni & R.A. Ringuelet. 1987. Bibliografia de 10s peces de agua dulce de la Argentina y Uruguay. Suplemen- to 1986. Biologia Acuatica 9: 1-61. LopCz, H.L., R.C. Menni & R.A. Ringuelet. 1989. Bibliografia de 10s peces de agua dulce de la Argentina y Uruguay. Suplemen- to 1988. Publ. Ministerio de Asuntos Agrarios y Pesca de la provincia de Buenos Aires, La Plata. 42 pp. Liiling, K.H. 1978a. Der Schuppensole Messingsalmer (Gym- nocharacinus bergi) aus Patagonien. Tetra Informat. 12: 28. Liiling, K.H. 1978b. Die Wiederentdeckung und die erste Auss- tellung des Patagonischen Messingsalmers, Gymnocharaci- nus bergi, in einem Schanaquarium. Zoologischer Garten 48: 167-176. Liiling, K.H. 1978c. Nuevo hallazgo y nota sobre Gymnocharaci- nus bergi Steindachner, 1903 (Characidae, Tetragonoperidae). Rev. Biol. del Uruguay 6: 19-29. Liiling, K.H. 1978d. Wissenschaftliche Ergebnisse des For- schungsaufenthaltes Dr. K.H. Liiling in Argentinien 1975: Ub- er den Biotop des patagonischen Messingsalmers Gymnoch- aracinus bergi (Pisces, Characoidei). Zool. Anz. 201: 410. Mac Arthur, R.H. & E.O. Wilson. 1967. The theory of island bio- geography. Princeton Univ. Press, Princeton. 203 pp. Mac Donagh, E.J. 1934. Nuevos conceptos sobre la distribution geogrhfica de 10s peces argentinos basados en expediciones del Museo de La Plata. Rev. Mus. La Plata 24: 21-170. Menni. R.C. 1984. Los peces en el medio marino. Ed. Sigma, Buenos Aires. 169 pp. Menni. R.C., H.L. Lopez & R.H. Aramburu. 1988. Ictiofauna de Sierra de La Ventana y Chasic6 (Provincia de Buenos Aires, Argentina). Zoogeografia y parametros ambientales. An. Mus. Hist. nat. Valparaiso 19: 75-84. Miquelarena, A.M. & R.H. Aramburu. 1983. Osteologia y lep- idologia de Gymnocharacinus bergi (Pisces, Characidae). Limnobios 2: 491-512. Olivier, S.R. 1961. Estudios limnologicos en la laguna Vitel (Par- tido de Chascomus, Bs. Aires-Argentina). Agro. 3: 1-128. Pianka, E.R. 1982. Ecologia evolutiva. Ed. Omega, Barcelona. 365 pp. Pozzi, A.J. 1936. Nota sobre Gymnocharacinus bergi Stein- dachner. Physis 12: 161-165. Ringuelet, R.A. 1962. Ecologia acuatica continental. Eudeba, Buenos Aires. 138 pp. Ringuelet, R.A. 1972. Ecologia y biocenologia del habitat lagu- nar o lago de tercer orden de la regi6n Neotropica templada (Pampasia sudoriental de la Argentina). Physis 31: 55-76. Ringuelet, R.A. 1975. Zoogeografia y ecologia de 10s peces de aguas continentales de la Argentina y consideraciones sobre las areas ictiologicas de America del Sur. Ecosur. 2: 1-122. Ringuelet, R.A. 1978. Dinamismo historic0 de la fauna Brasilica en la Argentina. Ameghiniana 15: 255-262. Ringuelet, R.A. 1981. El ecotono faunistico subtropical pampas- ico y suscambios hist6ricos. pp. 75-80. In: Symposia, VI Jorna- das Argentinas de Zoologfa, I1 Biogeografia, La Plata. Ringuelet, R.A., R.H. Aramburu & A.A. de Aramburu. 1967a. Los peces argentinos de agua dulce. Comision de Investiga- ci6n Cientifica, Province of Buenos Aires, La Plata. 602 pp. Ringuelet, R.A., A. Salibiin, E. Claverie & S. Ilhero. 1967b. Lim- nologia quimica de las lagunas pampasicas (Provincia de Bue- nos Aires). Physis 27: 201-221. Sokal, R.R. & F.J. Rohlf. 1979. Biometria. Principios y metodos estadisticos en la investigation biol6gica. H. Blume, Madrid. 832 pp. Steindachner, F. 1903. Uber einige neue Reptilien und Fischar- ten des Hofmuseums in Wien. Sitz. Akad. Wiss. Wien 112; 15- 22. Wegrzyn, D. & S. Ortubay. 1991. Nuestros salm6nidos. Dir. Pesca Prov. Rio Negro, Viedma. 120 pp. Weitzman, S.H. & M. Weitzman. 1982. Biogeography and evolu- tionary diversification in Neotropical freshwater fishes, with comments on the refuge theory. pp. 403-422. In: G.T. Prance (ed.) Biological Diversification in the Tropics, Columbia Uni- versity Press, New York. Welcomme, R.L. 1979. Fisheries ecology of floodplain rivers. Longman, London. 317 pp.