Vulnerability of pejerrey Odontesthes bonariensis populations to climate change in pampean lakes of Argentina


Tel.: +49 421 2380075; fax: +49 421 2380030; email:


The vulnerability of the pejerrey Odontesthes bonariensis population in Lake Chasicó was assessed under different climate change conditions. During the sampling period, the water temperature was adequate for fish reproduction and to sustain an adequate sex ratio. Climate-driven higher temperatures, however, may severely distort population structure and cause drastic reduction or local extinction of stocks. Lake Chasicó can be classified as eutrophic with clear waters and cyanobacteria that regularly cause fish mortality were identified as Nodularia spumigena and Oscillatoria sp. Global warming may strengthen the effects of eutrophication (e.g. toxic blooms or anoxia). Since many Cyanophyta species tolerate higher temperatures better than other algae, toxic blooms could increase. Furthermore, cyanobacteria have low nutritional value and could decouple the low-diversity food web. Lake Chasicó has currently the salinity optimum (c. 20) for the development of the early life-history stages of O. bonariensis. Climate change, however, is likely to amplify the intensity of droughts or inundations. Floods can endanger O. bonariensis development due to its sub-optimal growth at low salinity and droughts could increase lake salinity and also temperature and nutrient concentration. In order to reduce some of the effects of climate change on the O. bonariensis population in Lake Chasicó, integrated basin management based on an eco-hydrological approach is proposed.


In the last 30 years of the 20th century, warmer temperatures affected the phenology of organisms, the range and distribution of species, and the composition and dynamics of communities (Walther et al., 2002). Climate change has had detectable effects on fish distribution through changes in growth, survival, reproduction or responses to alterations at other trophic levels (Perry et al., 2005).

Changes in climate, and in particular temperature, affect and will continue to affect fishes at all levels of biological organisation, influencing physiological and ecological processes in many direct, indirect and complex ways (Graham & Harrod, 2009). Pronounced effects of climate warming are now being observed in many inland water bodies with potentially severe consequences for the aquatic biota (Williamson et al., 2009).

A typical water body in the La Pampa region of Argentina can be described as a relatively large (>100 ha), permanent, shallow lake (Torremorell et al., 2007). Pampean lakes vary from eutrophic to hypertrophic and present a highly variable hydrochemistry (Quirós et al., 2002). Furthermore, they frequently generate blooms of potentially toxic Cyanophyta (bluegreen algae or cyanobacteria) (Quirós, 2000).

The pejerrey Odontesthes bonariensis (Valenciennes) (Atheriniformes, Atherinopsidae) is the emblematic native fish characteristic of the pampean lakes and has been introduced in numerous water bodies around the world, particularly in Japan for aquaculture. This fish is generally zooplanktivorous (Da Silva Cassemiro et al., 2003), has thermolabile sex determination (Strüssmann et al., 1996, 1997) and thermal induction of germ cell deficiency (Strüssmann et al., 1998). Lake Chasicó represents a model of a pampean lake with a high O. bonariensis biomass and illustrates the instability of the hydrology of these water bodies.

In the early 1980s, extraordinary floods caused large increases in the areas of pampean lakes, particularly in Lake Chasicó. This phenomenon has been hypothesized to be related to strong El Niño events (Lara, 2006). The El Niño–Southern Oscillation (ENSO) phenomenon involves two extreme phases characterized by anomalously warm (El Niño) and cold surface (La Niña) waters in the eastern tropical Pacific Ocean. The ENSO influence on rainfall is faint but discernible south of 31° S (Pasquini et al., 2006). During La Niña events in the Argentinean pampas, maximum temperatures and solar irradiance tend to be higher and minimum temperature and rainfall tend to be lower than normal. El Niño events show an opposite trend (Podestáet al., 1999).

The models used for climate change projections leave an uncertain ENSO forecast. Timmermann et al. (1999) predicted more frequent El Niño-like conditions, whereas Collins (2005) found no trend towards either mean El Niño-like or La Niña-like conditions. Besides ENSO oscillations, non-ENSO long-term rainfall anomalies have had catastrophic consequences throughout the region (Scian et al., 2006). Climate change, however, is likely to increase the risk of droughts and floods (IPCC, 2007).

Hydrological changes have a dramatic effect on continental water bodies and associated biota. The current condition of a lacustrine ecosystem must be understood to assess the magnitude of shifts in future environmental conditions. The efficacy of lakes as sentinels of climate change depends on understanding internal lake processes (Adrian et al., 2009). The aim of the present work was to evaluate the current state of Lake Chasicó, to establish a baseline for monitoring future changes and to assess the general effect of climate change on pampean lakes in Argentina.

Materials and methods

Lake Chasicó (Fig. 1) is located in the south of the pampean region (38° 37′ S; 63° 05′ W) of Argentina. This endorheic lake has a maximum depth of c. 16 m and is topographically the lowest water body in South America (20 m below sea level).

Figure 1.

(a) Lake Chasicó (inline image) and other pampean lakes in Buenos Aires Province of Argentina. (b) Location of the main sampling stations (CV, EE, EV and RC) in Lake Chasicó.

In 1963, Lake Chasicó was a hypersaline water body (c. 100) with a surface area of c. 3100 ha and devoid of fishes due to its high salinity. This saline lake experienced several flood events and underwent a rapid increase in surface area in the 1980s, and currently has an area of c. 12 000 ha. Its salinity decreased to c. 20 due to inundation, although it increased during recent drought periods (reaching c. 23). Odontesthes bonariensis invaded the lake via the River Chasicó following the increase in surface area, encountering moderate salinities that favour growth during early development stages (Tsuzuki et al., 2000). Within a few years, fish biomass increased greatly with abundant specimens reaching up to 500 mm total length (LT) and c. 2 kg mass. A total of fishable O. bonariensis (>250 mm LT) of 420 t year−1 was estimated as maximum sustainable catch (Berasain & Argemi, 2006). In order to protect the O. bonariensis stock from overfishing, this lake was designated a nature reserve. The fish population in Lake Chasicó, however, is endangered by recurrent summer mortality, possibly caused by toxic cyanobacteria, as also seen in other pampean lakes.

Three sampling stations (Fig. 1) were chosen in relevant locations of Lake Chasicó: (1) station CV, close to Chapalcó village to evaluate possible anthropogenic disturbances; (2) station EE, situated in the site called El Embudo, where fish mortality has been frequently recorded; (3) station EV, in El Vivero, located near the mouth of the River Chasicó to estimate the influence of freshwater input in Lake Chasicó. Furthermore, a sampling station was located upstream in River Chasicó (station RC) to assess possible agricultural pollution in this nature reserve. Stations in Lake Chasicó were sampled once a month from July 2007 to September 2008 (n = 42) and in River Chasicó once every 4 months (n = 4).

Conductivity, temperature, dissolved oxygen, pH, salinity and turbidity were measured with a multivariable water quality meter (U-10, Horiba Ltd.; Additionally, water transparency was determined with a Secchi disk.

Lake water (1 or 0·5 l) was filtered through precombusted (at 450° C, 5 h) glass-fibre filters (Whatman GF/F; Filters were stored frozen at −20° C and filtrates were poisoned with HgCl2 and stored at 4° C in 50 ml PE bottles for nutrient analyses (Kattner, 1999). Nitrate, nitrite, ammonium, silicate and phosphate were determined according to standard methods (Kattner & Becker, 1991) with a nutrient analyser (Evolution III, Alliance Instruments; Water aliquots for the analysis of dissolved organic carbon (DOC) and nitrogen (DON) were adjusted after filtration to pH 2 with H3PO4 and kept frozen (−20° C) in precombusted sealed ampoules. DOC and DON were analysed by high-temperature catalytic oxidation with a Shimadzu TOC-VCPN and TNM-1 ( The GF/F filters were dried at 50° C for 12 h and analysed for particulate organic carbon (POC) and nitrogen (PON) by high-temperature flash combustion with a Carlo Erba NA 2100 elemental analyser (

For the quantification of pigments, filters were frozen and stored in the dark until analysis. Pigment extraction was performed with 90% acetone in water for 24 h at 4° C. The chlorophyll a content was quantified after Lorenzen (1967).

Fish larvae were caught by horizontal towing with a bongo net of 500 µm mesh-size at 1 m depth and a towing speed of c. 1 m s−1. The volume of filtered water was calculated with a mechanical flowmeter. Zooplankton were collected with 20 vertical hauls of a plankton net with 200 µm mesh-size. Samples were preserved with 4% formaldehyde and counted with a counting chamber for zooplankton (Hydro-bios; Phytoplankton was sampled using a Van Dorn bottle, preserved with Lugol solution and counted in duplicate using the method of Utermöhl (1958). The microalgae biovolume was calculated according to Hillebrand et al. (1999).


In Lake Chasicó, the surface water temperature increased from 5° C in winter to 25° C during late summer but did not present a thermal stratification throughout the year. Salinity increased from 21 during late winter 2007 to 23 in early spring 2008, and the water level dropped c. 1 m at the end of this period, due to low rainfall. The pH was slightly alkaline at mean ±s.d. 8·9 ± 0·6, and the waters were well oxygenated (mean ±s.d. 8·6 ± 2·1 mg l−1) throughout the year. The annual mean ±s.d. value of chlorophyll a was 5·8 ± 4·5 µg l−1. Turbidity (Secchi disk depth) was mean ±s.d. 1·4 ± 0·6 m.

This lacustrine system was not limited by nitrogen and ammonium (mean ±s.d. 5·2 ± 1·7 µM N) represented mean ±s.d. 83 ± 13% of the sum of inorganic nitrogenous compounds. Phosphate values were predominantly high (mean ±s.d. 3·6 ± 0·5 µM P) with only low variation throughout the year. In contrast, silicate concentrations had high variations mainly at station EV and were particularly related to the river discharge. Silicate reached very low values in summer (1·6 µM Si). During the entire sampling period, the mean ±s.d. N:P ratio was 1·8 ± 0·7. The situation was completely different at station RC: nitrate (mean ±s.d. 24·2 ± 12·4 µM N) was the main inorganic nitrogenous nutrient (82%) and the mean ±s.d. N:P ratio was very high (72·9 ± 35·1) due to low phosphate (mean ±s.d. 0·7 ± 0·7 µM P). The silicate concentration in this river was particularly high (mean ±s.d. 782·5 ± 134·3 µM Si).

The POC and PON mean concentrations were 195·2 and 23·2 µM, respectively, and the maximum values of POC (2254·2 µM) and PON (195·7 µM) were found at station EV associated with the plume of the River Chasicó in summer. The annual mean ±s.d. concentration of dissolved organic carbon and nitrogen were 3751 ± 960 and 230 ± 20 µM, respectively.

Throughout the year, phytoplankton was generally dominated in abundance by small flagellates (Chlamydomonas spp.) and in biovolume by diatoms (Cyclotella spp.). Chlorophyta (Oocystis spp.) and dinoflagellates were abundant during early autumn. Cyanophyta, such as Microcystis spp., Planktothrix sp., Anabaena spp., Nodularia spumigena and Oscillatoria spp. were found during summer, without associated fish mortality. The highest biovolume of Cyanophyta was found at station EE (Fig. 2). The maximum number of cyanophytes was c. 27·6 million cells l−1 and their relative abundance 53%.

Figure 2.

Monthly biovolume (mm3 l−1) of phytoplankton including flagellates (inline image), diatoms (inline image), chlorophytes (inline image), cyanobacteria (inline image) and dinoflagellates (inline image) at station EE (see Fig. 1).

The copepod Boeckella poopoensis dominated the zooplankton in terms of both abundance and biomass. This species was displaced by the cladoceran Moina eugeniae in December and by the rotifer Brachionus plicatilis in March.

Larvae of O. bonariensis were the only fish species captured. Temperature and larvae abundance at stations CV, EE and EV are shown in Fig. 3. One peak of abundance (2·5 larvae m−3) occurred during mid-spring (October) at station EE. Larvae were mainly found in the small bays at this station, where the water temperature was a few degrees higher than in other parts of the lake. During late summer (February), the small bays at station EE had dried up due to the drought. The other abundance peak occurred in early summer (December) at stations CV and EE with c. 2 larvae m−3. Two small peaks were found at station EE in January (0·5 larvae m−3) and at station EE and in February (0·6 larvae m−3). From March to May, larvae were mostly recorded at station EV (1·3–0·2 larvae m−3, respectively). The range of temperatures for larvae appearance was from 16 to 25° C.

Figure 3.

Monthly abundance of Odontesthes bonariensis larvae in stations CV (inline image), EV (inline image) and EE (inline image) and mean water temperature (inline image).


Lake chasicó and O. bonariensis populations

According to the hydrology and nutrient concentrations, Lake Chasicó can be classified as eutrophic with clear waters (Table I). One of the effects of the nutrient abundance is the high fish biomass. The catch per unit of effort (CPUE) of O. bonariensis was similar to highly turbid lakes. The current environmental conditions in conjunction with management guidelines today provide successful and sustainable fisheries in Lake Chasicó.

Table I.  Average and range values of Secchi disk depth, total phosphorus (TP), total nitrogen (TN), chlorophyll a (Chl a), cyanobacteria relative abundance (%) and catch per unit effort (CPUE) of Odontesthes bonariensis of Lake Chasicó and another 39 pampean lakes arranged according Quirós et al. (2002)
ConditionChasicó LakeClearPampean lakesHighly turbid
  1. *To ease comparison with Quirós et al. (2002), the values of this work were converted from µM to µg P and mg N l−1 as NH3 according to determination of Kjeldahl-N.

  2. **Phosphate (major phosphorus component).

Secchi (m)1·41 (0·03–3·40)1·50 (0·60–4·60)0·41 (0·15–1·10)0·23 (0·11–0·35)
TP (µg P l−1)*113 (70–170)**230 (23–1288)288 (25–1250)895 (466–1275)
TN (mg N l−1)*4·3 (2·9–8·0)6·4 (2·7–11·6)10·6 (2·9–28·8)7·4 (3·5–18·4)
Chl a (µg l−1)5·8 (0·4–30·5)19·2 (1·6–82·4)103 (12·8–405·0)113·0 (52·0–192·0)
Cyanophyta %10·0 (0·0–53·4)49·3 (23·9–90·4)70·9 (2·6–99·1)78·5 (76·4–82·5)
CPUE (kg day−1)337 (217–481)56 (0–205)96 (0–435)396 (0–1512)

During the whole sampling period, the water temperature (maximum 25° C) in Lake Chasicó was adequate for good fish reproduction. The lake is located in the south region of Buenos Aires Province (Fig. 1), in a colder climate than other pampean lakes. At temperatures >25° C, gonadal regression has been observed in O. bonariensis (Soria et al., 2008). Larvae and juveniles exposed for prolonged periods at incipient lethal temperatures (29° C) suffer a germ-cell deficiency up to complete sterility (Strüssmann et al., 1998). Water temperature in northern pampean lakes often rises above 30° C during summer (Boltovskoy et al., 1990). The situation in these lakes is completely different: in Lake San Miguel del Monte (35° 27′ S; 58° 47′ W) and Lake Lacombe (35° 50′ S; 57° 53′ W) where wild O. bonariensis adults with gonadal abnormalities were reported including partial and total sterilization probably caused by high water temperatures (Ito et al., 2008).

The temperature of Lake Chasicó also allows an adequate sex ratio to be maintained. Strüssmann et al. (1997) found that fish larvae exposed to c. 17° C from the hatching to the juvenile stage become all-female, while groups exposed to c. 25° C become male-biased. According to peaks in larvae abundance and the water temperature of Lake Chasicó (Fig. 3), the larvae in early spring (October) and late autumn (April to June) should become mainly female. Conversely, the larvae in February should become mostly male. During the rest of the year, the sex determination was highly variable.

Climate change scenarios

In response to higher climatic temperatures, the offspring of O. bonariensis may be considerably altered. For the next two decades, a warming of c. 0·2° C per decade is predicted (IPCC, 2007). The abundance of males could sharply increase due to higher temperatures and severely distort the population structure, causing drastic reduction or local extinction of O. bonariensis stocks. Additionally, this situation could be aggravated because larvae and juveniles generally develop in micro-environments formed by shallow, vegetated and food-rich areas (Loubens & Osorio, 1988), which could become an entirely unsuitable habitat space. Moreover, O. bonariensis adults with gonadal abnormalities caused by higher temperatures may be more frequently found in pampean lakes.

A nutrient scenario of changing concentrations may affect the O. bonariensis population. Climate change may strengthen the effects of eutrophication (Mooij et al., 2007). In Lake Chasicó, as in other pampean lakes (Table I), the concentrations of phosphate and total nitrogen, with an N:P ratio <3 clearly favour the occurrence of cyanophytes as high phosphorus and low nitrogen concentrations are the most important drivers for cyanobacteria growth in shallow lakes (Kim et al., 2007).

In a scenario of increasing cyanophyte abundance, toxic blooms may be responsible for causing considerable fishery problems through decrease in food quality or toxic effects. In Lake Chasicó, the maximum relative abundance of cyanophytes was 53%, whereas in turbid pampean lakes 99% can be reached (Table I). As toxic cyanophytes: Nodularia spumigena and Oscillatoria sp. were found previously described to be responsible for O. bonariensis mortalities in two lakes near the study region (Grosman & Sanzano, 2002). Cyanobacteria respond more strongly to rising temperature than green algae and diatoms, resulting in higher growth rates and more frequent bloom events (Paerl & Huisman, 2008). Cyanobacteria are known to be a relatively poor food item for zooplankton (Gliwicz & Lampert, 1990). The lower food quality could signify lower abundance and growth rate for the zooplankton and thus, may be detrimental for the nutritional requirements of O. bonariensis.

In addition to toxic effects, increasing plankton biomass can potentially also generate anoxia in the still well-oxygenated Lake Chasicó, particularly during warmer summers. Anoxia enhances regeneration of nutrients such as phosphorus that can in turn generate more severe cyanobacteria blooms (Wagner & Adrian, 2009). Climate-driven reductions in oxygen are predicted to increase summer kill of cold-water fishes (Williamson et al., 2009).

Extreme rainfall events may increase the phosphorus loading of lakes (Mooij et al., 2005) creating eutrophication problems for shallow lakes. Floods, however, could also prevent toxic blooms by shifting nutrient ratios. The present data show that the freshwater inflow to Lake Chasicó has had a low phosphate content and high N:P ratio. The poor adaptation to freshwater makes O. bonariensis particularly vulnerable to flood events. The best survival and growth rates of embryos, larvae and juveniles were usually obtained at intermediate salinities (Tsuzuki et al., 2000). If climate change leads to warmer temperatures and higher precipitations (i.e. lower lake salinity), O. bonariensis populations and its economic benefits may be critically damaged, but otherwise the biodiversity of warm-adapted (truly freshwater) fish species could increase.

According to Graham & Harrod (2009), fish communities in temperate freshwater ecosystems will be increasingly dominated by warm-adapted cyprinid and percid species. The common carp Cyprinus carpio L. was detected in River Chasicó (F. C. Piantanida, pers. comm.). Currently, C. carpio does not grow in Lake Chasicó because salinities >10·5 have a negative effect on its growth rate (Wang et al., 1999). Cyprinus carpio may be favoured by greater inundation and could affect O. bonariensis populations.

With increased drought, the freshwater discharge will diminish, evaporation and nutrient concentrations will increase and temperature changes would become more extreme affecting biota severely. Concerning the effect of the expected salinity increment, it is worth mentioning that O. bonariensis currently develops at the growth salinity optimum of c. 20 in Lake Chasicó. The history of this water body, however, has shown that salinity can be easily shifted from the optimum for fish growth. Resource management plans should take this hydrological instability into account.

Future perspectives and integrated basin management

The pampean lakes are extremely sensitive to climate change and may be one of the most affected environments, including local O. bonariensis populations. The climate sensitivity of pampean lakes is related to the fact that extreme climate-driven changes cause severe hydrological changes affecting the biota. Moreover, some effects of eutrophication (e.g. toxic blooms or anoxia) may be aggravated by climate change. The complexity of factors and processes involved in a simple environment like Lake Chasicó, such as fish physiology, biotic interactions, trophic state, hydrological variations and teleconnections with global acting phenomena (e.g. ENSO), make the forecast of these changes uncertain. Higher temperature and increased risk of floods or droughts will, however, affect O. bonariensis populations in Lake Chasicó negatively.

In order to reduce the vulnerability of the O. bonariensis population in Lake Chasicó, an integrated basin management plan is proposed (Fig. 4) based on an eco-hydrological approach (Zalewski, 2002), which takes into account that changes in land use could increase fertiliser utilization. This would involve the sequestration of nutrients derived from agricultural runoff by the creation of wetlands along the River Chasicó. The wetland creation may thus help avoid some effects of eutrophication on this lake. Stabilization of salinity values at the O. bonariensis physiological optimum requires the damping of hydrological oscillations due to droughts or floods, which could be achieved by constructing reservoirs linked to Lake Chasicó and alleviation channels connected to the neighbouring estuary of Bahía Blanca.

Figure 4.

Effects of climate change on Odontesthes bonariensis population in Lake Chasicó (inline image) and tentative basin integrated management (inline image).

We are grateful to the German Academic Exchange Service (DAAD), to the Ministry of Education of Argentina (ME), to the Leibniz-Zentrum für Marine Tropenökologie (ZMT), to the Ecological Chemistry Division (AWI), to the Environmental Chemistry Chair (UNS), to the Ministry of Education, Culture, Sports, Science and Technology (Monbukagakusho) of Japan and to the General Secretary of Science and Technology (UNS) for financing this work. We thank A. Rule from Chapalcó Ray for logistical support, S. B. José de Paggi and J. C Paggi (Instituto Nacional de Limnología; INALI-CONICET) for the taxonomic identification of Cladocera and Rotifera, Chasicó Park Rangers for their assistance and Chemical Oceanography (Instituto Argentino de Oceanografía; IADO-CONICET) for the instrument facilities. We also thank anonymous reviewers for their helpful comments on an earlier version of this manuscript.