Chemosphere 73 (2008) 365–370

Contents lists available at ScienceDirect

Chemosphere

journal homepage: www.elsevier.com/ locate /chemosphere

Discriminating the pH tox­icity to Poecilia reticulata Peters, 1859 in the Dunas

Lake (Camaçari, BA, Brazil)

0045-6535/$ - see front matter © 2008 Elsevier Ltd

doi:10.1016/j.chemosphere.2008.05.061

Cristiano V.M. Araújo *, Salomão J. Cohin-de-Pinho, Carla B.A. Chastinet, Jéssica S. Santos, Eduardo M. da Silva

In­stituto de Biologia, Federal Un­iversity of Bahia, Cam­pus de On­din­a, 40170-115 Salvador, BA, Brazil

a r t i c l e i n f o a b s t r a c t

Article history:

Received 26 September 2007

Received in revised form 28 April 2008

Accepted 27 May 2008

Available online 15 July 2008

Keywords:

Acidifi­cation

Acute Tox­icity

Bioassays

Biomonitoring

Survival time

* Corresponding author. Tel.: +55 71 3263 6525; f

E-m­ail address: cvmaraujo@yahoo.com.br (C.V.M

Tox­ic potential of the pH reduction to fi­ngerlings of Poecilia reticulata, through acute tox­icity bioassays,

as well as the influence of increased pH on the tox­icity were assessed. Acid lake samples (Dunas Lake)

were collected during 19 months, and assessed with following treatments: water at local pH (±3.0) and

samples with modifi­ed pH to 3.5, 3.8, 4.0, 4.3, 4.6, 5.0, 5.5, 6.0, and 6.5. Culture water samples with pH

reduced to 3.0 were also assessed. Newborn P. reticulata were ex­posed during 96 h, and dead/immobile

organisms were counted at various time intervals during ex­posure (short intervals in the beginning and

long towards the end). Mean results of LT50 and confi­dence intervals from the Dunas Lake and control

water with reduced pH were 1.36 (±0.48) h, and 1.03 (±0.50) h, respectively, with no statistical difference.

Samples with increased pH showed a signifi­cant reduction in tox­icity, with no tox­icity detected at pH 6.0

and higher. Relationship between pH and lethal tox­icity for fi­ngerlings of P. reticulata demonstrated that

pH ex­erted a strong effect on the survival of this species at the Dunas Lake, ex­plaining about 80% of the

tox­icity observed.

© 2008 Elsevier Ltd. All rights reserved.

1. Intro­duc­tio­n

The knowledge about ecosystem acidifi­cation goes back to the

mid-18th century (Gorham, 1998), but, over the past 30 years, sci-

entists have been working to understand the causes and results

of the acidifi­cation process, and how acidic deposition might influ-

ence different ecosystems (Driscoll et al., 2003). Acidifi­cation is

one of the main causes of environmental degradation in aquatic

ecosystems in the temperate regions, not only due to its tox­icity,

but also due to its effect on the speciation, mobility, and bioavail-

ability of other tox­icants (Lopes et al., 1999). For instance, many

regions are severely impacted by acidic rain, acid mine drainage,

volcanic activities or by other types of anthropogenic activities

(Geller et al., 1998; Ribeiro et al., 2002), resulting in decreasing pH

values, followed by increasing metal solubility (Geller et al., 1998).

As a whole, acidifi­cation in the Brazilian ecosystems is little doc-

umented, although the impact caused on the biodiversity is very

similar to other regions (Jesus, 1996). Rodhe et al. (1988) identi-

fi­ed schematically areas in which acidifi­cation might represent a

potential threat on the basis of ex­pected emissions, population

density, and soil sensitivity for south-eastern Brazil. More recently,

Moreira-Nordemann et al. (1988); Ometto et al. (2005) and Mar-

. All rights reserved.

ax­: +55 71 3263 6511.

. Araújo).

tinelli et al. (2006) pointed out the risks involved in sugar-cane

burning to the acidifi­cation of the same region. An acidifi­ed lake,

and its ecotox­icity history, after rehabilitation, has been described

by da Silva et al. (1999a, 2000).

The tox­icity of the trace elements, especially metals, does not

only depend on their concentration, but also on their bioavailabil-

ity, which is higher under reduced pH conditions (Wren and Ste-

phenson, 1991; Renoux­ et al., 2001). Therefore, the degradation

of fi­sh populations is ex­plained primarily by the direct combined

influence of low pH and aluminum ions, which causes biochemi-

cal and physiological disturbances (Moiseenko and Sharova, 2006).

Besides, pH influences the ionic regulation of aquatic species, the

rate of organic matter decomposition, and primary production

(Abel, 1996; Geller et al., 1998). According to Driscoll et al. (2003),

few fi­sh species can survive in pH lower than 4.5. Indeed, most

of the acidifi­ed waters throughout the world are fi­sh-free environ-

ments (Van Sickle et al., 1996; Nix­dorf et al., 1998); however some

small South American fi­sh can be found in acidic blackwater rivers

(such as Rio Negro, Amazon, Brazil) whose pH can be around 3.5

(Mounier et al., 1999; Matsuo and Val, 2002; Aride et al., 2007).

Gonzalez et al. (1998) suggested that ex­ceptional acid tolerance is

a characteristic of fi­sh that inhabit acidic ecosystem. A tool adopted

for preliminary environmental evaluation of acidifi­cation effects

in aquatic ecosystems is the use of survival time as an endpoint,

allowing the establishment of the species recolonization potential

(Ribeiro et al., 2002). It is reasonable to suppose that, increasing

366 C.V.M. Araújo et al. / Chem­osphere 73 (2008) 365–370

pH values should improve water quality of acidic systems, leading

to previous or acceptable ecological conditions, and to reestablish-

ment of previous food webs, therefore it is ex­tremely relevant to

assess biological and chemical conditions of acidifi­ed ecosystems

(Tipping et al., 2002; Driscoll et al., 2003).

Poecilia reticulata (guppy) is a tropical species, very abundant

in shallow waters of canals, rivers, lagoons and reservoirs in South

America, and largely employed in ecotox­icological tests. The results

of several studies showed that P. reticulata is a very sensitive spe-

cies when ex­posed to tox­icants (Gallo et al., 1995; Miliou et al.,

1998; Polat et al., 2002; Yilmaz et al., 2004), and, therefore, ade-

quate for biomonitoring programs (Widianarko et al., 2000) and is

recommended as a standard ecotox­icological test organism (OECD,

1992; ABNT, 2002). The sensitivity of P. reticulata in comparison to

other fi­sh species, as Brachydan­io rerio (zebra fi­sh), Cyprin­us carpio

(common carp), Lepom­is m­acrochirus (bluegill), Pim­ephales prom­-

elas (fathead minnow), Salm­o gairdn­eri (rainbow trout), was stud-

ied by Vittozzi and De Angelis (1991). Additionally, P. reticulata has

been used with success in the biomonitoring of an acidic lake for

several years (da Silva et al., 1999a), including in in­ situ bioassays

(Araújo et al., 2006). This species has been chosen as a test organ-

ism for the Dunas Lake survey, because it was a common species in

the lake prior to the contamination episode.

According to da Silva et al. (2000), in the late 1980s large quanti-

ties (ca. 34 t) of both industrial and domestic solid waste including

sulphur, iron, titanium diox­ide and ilmenite residues were depos-

ited on the dunes adjacent to a small lake (Dunas Lake, Camaçari,

BA, Brazil). These wastes were leached out by rainwater percolat-

ing through the dunes, thus contaminating the ground- and sur-

face-water. There was a decrease in pH to 1.8 of the ground- and

surface-water and an increase in the concentrations of dissolved

iron and sulphate, causing the precipitation of humic acids, leading

to highly transparent waters with a concurrent disruption of the

biological communities in the lake (da Silva et al., 1999b, 2000). It

was not investigated if the tox­icity of the water of the Dunas Lake

was due to the direct effect of low pH value, about 3.0, or due to

the ability of low pH to increase the bioavailability of tox­ic metals.

The methodologies to discriminate a tox­ic component using serial

dilutions or chelating agents can mask the ecotox­icity and alter

the characteristics of the samples (Lopes et al., 1999; Ribeiro et al.,

2002). In this contex­t, this work aimed to discriminate, through

bioassays with altered pH, the potential tox­icity of the acidity to

P. reticulata; identifying possible reduction in the tox­icity due to

increased pH; and to determine threshold survival for P. reticulata

in relation to Dunas Lake pH.

Fig. 1. Location o

2. Mate­rials and me­th­o­ds

2.1. Study site

The acidic lake (Dunas Lake) is located in Camaçari (BA, Brazil)

(Fig. 1) between geographic coordinates 12°489090 to 12°48912.30 S and 38°139090 to 38°139140 W, lying within a depression, forming

a narrow and shallow body of freshwater between dunes along the

Atlantic Ocean (da Silva et al., 2000). After contamination, a reha-

bilitation program was carried out (1992–1993) to recover ground-

and surface water quality and reduce contamination (Gomes, 1994;

da Silva et al., 1999b). Initially, the residues were partially removed

and the contaminated dune was sealed with impermeable layers

of clay and topsoil (hydraulically encapsulated). An additional

action was to pump the groundwater to reduce the contaminated

plume (Gomes, 1994; da Silva et al., 1999b, 2000). The results of

the established biomonitoring program, including bioassays with

guppy fi­ngerlings (P. reticulata), were described by da Silva et al.

(1999a, 2000) and Araújo et al. (2006).

2.2. Sam­plin­g

Monthly samplings (n­ = 19) of the Dunas Lake were carried out

from March 2003 to November 2004, ex­cept for August and Octo-

ber 2004. Samples were transported to the laboratory and kept at

4.0 ± 1.0 °C until the nex­t day, when the pH was measured. Imme-

diately the samples were distributed in 5 l beakers to alter the pH.

Samples were treated with NaOH 1 M to raise their pH, with the fol-

lowing treatments: original Dunas Lake pH (pH not altered – about

3.0), 3.5, 3.8, 4.0, 4.3, 4.6, 5.0, 5.5, 6.0, and 6.5. These treatments

aimed to assess the tox­icity reduction after increasing pH. To assess

the tox­ic effect of pH, the control water pH (i.e. dechlorinated tap

water) was reduced to the same pH value of the Dunas Lake, using

H2SO4 1 M, because this was the main acid in the acidifi­cation

process (da Silva et al., 1999b). A possible difference between the

Dunas Lake water in­ n­atura and the control sample with reduced

pH may be attributed to other factors potentially tox­ic present in

the Dunas Lake that could interfere in the tox­icity which may be

influenced by pH.

2.3. Physical–chem­ical an­alysis

Dissolved ox­ygen content (WTW, Inolab Ox­i Level 2), water hard-

ness (APHA, 1998), pH (Digi-Sense, Cole Parmer) and conductivity

(Hanna HI 9033) of the samples were evaluated at the beginning

f the study site.

C.V.M. Araújo et al. / Chem­osphere 73 (2008) 365–370 367

and end of each ex­periment. Mean pH was computed by calculat-

ing the average concentration of hydrogen ions and then calculat-

ing the respective pH. Bioassay results were validated when pH

variation was no higher than 0.5. All measurements were carried

out at 25 ± 1 °C.

Physical–chemical characterization of the Dunas Lake and

control water (dechlorinated tap water) was carried out and it is

summarized in Table 1. Metals were analyzed using inductively cou-

pled plasma atomic emission spectrometry (ICP-AES) and Cr(VI),

nitrate, nitrite, phosphate and sulphate were analyzed by ion chro-

matography (APHA, 1998). Data of control water was provided by

EMBASA (Water and Sanitation Company of the State of Bahia) that

is responsible for drinking water quality control.

2.4. Test organ­ism­ an­d acclim­ation­

Guppy (P. reticulata) fi­ngerlings, 10–15 days old (average length

of 1.0 ± 0.2 cm) were obtained from a local aquarist who kept the

fi­sh under standardised conditions. In spite of the recommenda-

tion by OECD (1992) and ABNT (2002) to use adult P. reticulata

in acute and static bioassays, fi­ngerlings in early life stages were

employed since the early life stages of fi­sh are in many cases the

most sensitive to adverse conditions (Farag et al., 1993; Petersen

and Kristensen, 1998), especially under pH changes (Vuorinen

et al., 2003).

The organisms were transported to the laboratory in glass flasks

with suf­fi­cient air, and acclimatized in 20 l glass aquaria containing

dechlorinated tap water for 24 ± 2 h prior to the ex­periments. The

fi­sh were not fed during this period or during the ex­periment. Accli-

mation and bioassays were performed at 25.0 ± 1.0 °C, in constant

temperature rooms with a photoperiod of 12:12 h (light and dark).

Static acute tox­icity tests were carried out according to OECD

(1992) and ABNT (2002). Dechlorinated tap water, in which the

fi­sh had been reared, was used as a control.

2.5. Bioassays

Glass aquaria of 1.2 l capacity, containing 900–1000 ml of water

sample, were used as test vessels. Five replicates of each treat-

ment, containing eight to ten fi­sh in each replicate were tested,

totalling 40–50 organisms ex­posed to each treatment. For each

treatment 19 bioassays (March 2003 to November 2004, ex­cept

for August and October 2004) were carried out. The organisms

were randomly distributed in the test vessels and fi­sh mortality

was checked at reduced time intervals, 10 min in the fi­rst 3 h of

the tests, 30 min of the third until the tenth hour, and at longer

time intervals in the following hours. Time was the independent

variable, and samples were not diluted. Dead fi­sh were counted

and removed immediately to avoid adverse effects due to decom-

Table­ 1

Physical and chemical characterisation of the Dunas Lake and dechlorinated tap

water

Parameters Dunas

Lake

Control

water

Parameters Dunas

Lake

Control

water

SO4 (mg l¡1) 113.7 – Total-Fe (mg l¡1) 0.65 0.26

Cr(VI) (mg l¡1) <0.01 <0.01 Dissolved-Fe (mg l¡1) 0.64 –

NO2 (mg l¡1) <0.1 <0.1 Total-K (mg l¡1) 0.78 –

NO3 (mg l¡1) <0.1 <0.1 Total-Mg (mg l¡1) 2.6 –

Phosphate

(mg PO4 l¡1)

<1.0 <1.0 Total-Mn (mg l¡1) 0.5 <0.005

Total-Al (mg l¡1) 1.5 0.16 Total-Na (mg l¡1) 6.7 –

Total-Ca (mg l¡1) 19.0 – Total-Ni (mg l¡1) <0.016 <0.016

Total-Cd (mg l¡1) <0.005 <0.005 Total-P (mg l¡1) <0.33 <0.33

Total-Co (mg l¡1) <0.01 <0.01 Total-Pb (mg l¡1) <0.14 <0.14

Total-Cr (mg l¡1) <0.01 <0.01 Total-Ti (mg l¡1) 0.05 –

Total-Cu (mg l¡1) 0.014 <0.005 Total-Zn (mg l¡1) <0.01 <0.01

position of the organisms. Behavioural changes such as loss of equi-

librium and swimming disorders were recorded. Organisms were

only considered dead when operculum and gill movements had

ceased and there was no swimming response after stimulation

with a plastic Pasteur pipette.

2.6. Data an­alysis

Median lethal time (LT50: time at which 50% of the test animals

were dead) of the treatments was determined by Probit Analysis,

considering the 5 replicates. Mean LT50 values were based on the

19 monthly values. To compare mean values the analysis of vari-

ance (one-way ANOVA) was used followed by Tukey multiple com-

parison test (Zar, 1996). Differences were considered signifi­cant at

P < 0.05 (Zar, 1996). All values are given as means ± standard devia-

tion. A logarithmic regression model based on pH values against

raw data of the Dunas Lake samples LT50 was plotted to evaluate

the proportion of the median lethal time that is predictable from

pH values.

3. Re­sults and disc­ussio­n

Survival in all control assays was greater than 90%, thus accom-

plishing the recommendations from OECD (1992) and ABNT (2002).

Dissolved ox­ygen concentrations were always above 6.5 mg l¡1 in

the Dunas Lake treatments and above 8.0 mg l¡1 in the control. The

Dunas Lake and control water presented a pH mean value of 3.08

(±0.4), and 7.34 (±0.15), respectively (measurement at 25 ± 1 °C).

Conductivity and hardness total values of all samples are sum-

marized in Table 2. There was no signifi­cant difference in conduc-

tivity and hardness mean values of Dunas Lake treatments after

changed pH. Regarding the Dunas Lake treatments, it could be

observed that the highest mean conductivity value measured

in the in­ n­atura sample was 343.80 lS cm¡1, and the lowest was

246.50 lS cm¡1 in the samples with pH 5.5 (Table 2). Therefore,

with a pH increase there was a correspondent decreasing in con-

ductivity. In the control samples, with no pH change, the mean con-

ductivity value was 418.55 lS cm¡1, corroborating that this param-

eter was not responsible for mortality during the ex­periments.

Control samples with reduced pH were signifi­cantly different from

all other samples (F11,216 = 48.179; P < 0.05) (Table 2). Mean hardness

of Dunas Lake treatments showed signifi­cant difference in relation

to the values registered in the control samples and control with

reduced pH (F11,216 = 26.371; P < 0.05) (Table 2). The mean value of

water hardness in the control samples was 126.25 mg CaCO3 l¡1,

while in the Dunas Lake was 79.00 mg CaCO3 l¡1 (Table 2). The

influence of the hardness on the bioavailability of some chemical

Table­ 2

Mean values and confi­dence intervals (95%) of conductivity and hardness of Dunas

Lake samples and control at the start of the ex­periment

Samples Parameters

Conductivity (lS cm¡1) Hardness (mg CaCO3 l¡1)

Control 418.55 (±67.32) A 126.25 (±20.49) A

Control (reduced pH

to Dunas Lake)

548.86 (±66.14) B 129.56 (±22.95) A

Dunas Lake (in­ n­atura) 343.80 (±12.14) AC 79.00 (±4.07) B

Dunas Lake (pH 3.5) 276.80 (±41.28) C 76.87 (±9.47) B

Dunas Lake (pH 3.8) 265.20 (±24.29) C 76.88 (±12.20) B

Dunas Lake (pH 4.0) 289.09 (±23.95) C 77.05 (±6.17) B

Dunas Lake (pH 4.3) 263.57 (±16.92) C 79.38 (±6.33) B

Dunas Lake (pH 4.6) 249.00 (±24.53) C 79.30 (±10.43) B

Dunas Lake (pH 5.0) 251.00 (±29.39) C 75.30 (±7.18) B

Dunas Lake (pH 5.5) 246.50 (±43.17) C 81.77 (±17.69) B

Dunas Lake (pH 6.0) 258.00 (±8.52) C 75.10 (±7.38) B

Dunas Lake (pH 6.5) 251.50 (±15.23) C 79.90 (±7.64) B

368 C.V.M. Araújo et al. / Chem­osphere 73 (2008) 365–370

compounds and the tox­icity is unarguable (Akkanen and Kukko-

nen, 2001), especially due to its capacity to modulate the pH effect

(Ribeiro et al., 2002). In our study, however, water hardness values

for the Dunas Lake did not change with pH increase, therefore the

changed pH of samples with NaOH did not show any influence in

this variable.

Mean survival time of the organisms in the Dunas Lake with

no altered pH was 1.36 (±0.48) h (Fig. 2). In the study of da Silva

et al. (1999a) with P. reticulata, the LT50 of Dunas Lake was <1 h

(pH lower than 3.0), with the majority of the values lower than

30 min. The results of the present study show a better tox­icologi-

cal condition of the lake, in comparison to the former study, due

to increased pH, but ecologically there has been no improvement

(i.e. the tested organisms are still not able to survive in the waters

of the Dunas Lake). In relation to other treatments of the Dunas

Lake there was a signifi­cant statistical difference, ex­cept for sam-

ples with pH 3.5 (Fig. 2). With gradual pH increase there was a

tox­icity reduction, resulting in higher LT50, although no signifi­cant

statistical difference was noted between treatments with pH 4.3 to

pH 5.5. The pH variation (0.5 units) for the bioassays may ex­plain

some of the lack of signifi­cance between these pH groups. The high-

est LT50 mean was registered in the samples with pH 5.5 (74.30 h).

Samples with pH 6.0 and 6.5 did not show any mortality during

96 h, with results being ex­pressed as >96 h. The relative increase

in survival time due to the pH increase can be considered a tox­ic-

ity measure that allows only the prediction of the ecotox­icological

effects to guppy fi­ngerlings and to assess the possibility of recoloni-

zation of this species during the rehabilitation of the Dunas Lake.

On the other hand, sublethal levels of acidity can be detrimental

affecting the weight and length of the fi­sh, oocyte development

earlier during vitellogenesis and delay of spermatogenesis (Vuori-

nen et al., 2003). When fi­sh are subjected to acid stress, blood

pH decreases possibly as the result of flux­ of H+ ions across gill

membranes into the blood and an acidemia (lowering of the pH

of the blood) may occur, decreasing the capacity of hemoglobin

to transport ox­ygen (Fromm, 1980). Moreover, the ability of fi­sh

uptake ox­ygen decreases with a decrease in water pH due to gill

mucus coagulation, and this may represent the primary cause of

death of fi­shes (Fromm, 1980). In a study with guppy, Dunson et

al. (1977) found mortality in all organisms after 11 days at pH 4.75,

and it was associated with loss of large amounts of body sodium.

Fig. 2. LT50 mean values to P. reticulata, confi­dence intervals (95%) of the Dunas Lake

(DL) samples and control with reduced pH (CR). Mean values followed by the same

letter in the same column do not differ signifi­cantly by Tukey’s test (F8,162 = 109.89;

P < 0.0001).

Undoubtedly, the tox­ic action of increased ambient H+ ions concen-

tration on freshwater fi­sh is variable (Fromm, 1980).

Dunas Lake in­ n­atura samples were compared to the control sam-

ples with reduced pH to assess whether the observed effect (LT50)

by action of the pH were similar to the observed effect in the Dunas

Lake samples. Differences in survival time between these two sam-

ples could show an additional tox­icity of Dunas Lake, unex­plained

by the pH, that will allow to discriminate the tox­icity due to pH

and due to other elements (Lopes et al., 1999). However, there was

no signifi­cant statistical difference between the LT50 mean values

of these two samples. Dunas Lake presented a mean LT50 of 1.36

(±0.48) h, while control water with reduced pH showed a mean

LT50 of 1.03 (±0.50) (Fig. 2). According to these results, it is possible

to attest that pH is the main responsible for tox­icity results from

Dunas Lake and that the fi­sh survival will increase with pH value

increases to 6.5.

The ex­ponential regression shows that the model is highly

signifi­cant as evident from the P-value (P < 0.0001) and from the

coef­fi­cient of determination (r2 = 0.80), indicating that pH is able

to ex­plain ca. 80% of the tox­icity observed. Similar data was also

obtained when tox­icity tests results using natural lake water,

between 2005 and 2007 (P < 0.0001; r2 = 0.83), were integrated in

the original data (Fig. 3). In this model, as well as in Fig. 2, there is

a high variation of the LT50 results of the tox­ic moderately samples

(e.g.: pH 3.8–5.0). According to Salvadó et al. (1995), the relation

between some parameters, such as pH, and a species does not fol-

low a linear model, because in general there is an optimal range,

in which below and above of this range the species can be affected

with density reduction or disappearance. For instance, the range of

pH considered safer to fi­sh is between 5 and 9, and ex­treme values

as 3.0–3.5 were considered unlikely to survive of any fi­sh (EIFAC,

1969). According to Fromm (1980), it appears that the pH depres-

sion no-effect level for successful reproduction of fi­sh is around

6.5; however, an avoidance behavior could be observed at a pH of

5.5. Thus, the highest pH value tested was 6.5, because this is the

max­imum pH value registered in the water bodies of this region

(da Silva et al., 1999b, 2000), moreover, values above the neutral

pH can be tox­ic due to the alkalinity (Aride et al., 2007). When

the data generated in this work was plotted together with P. reticu-

lata tox­icity data (monthly results, between the years 2005–2007),

collected from natural Dunes Lake water (Chastinet, unpublished

results) (black circle in Fig. 3), a good agreement of the results was

pH values

2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0

LT50

(h)

0

12

24

36

48

60

72

84

96

Black and white circles (n=195)

+

=

0.282

3.898x

1

81.12y

e

+

=

0.243

3.913x

1

79.73y

e

0.832 =r

0.802 =r

White circles (n=171)

Fig. 3. Relationship between pH and lethal tox­icity for fi­ngerlings of P. reticulata.

White circles and continuous line represent results from the Dunas Lake modifi­ed

pH (this work). Black circles represent the same kind of test, using natural Dunas

Lake water between years 2005–2007 (Chastinet, unpublished results). The dotted

line represents the integration of black and white circles, as data from the two stud-

ies did not differ statistically (Mann-Whitney test: P < 0.05; U9 = 18400).

C.V.M. Araújo et al. / Chem­osphere 73 (2008) 365–370 369

obtained, confi­rming the predictability of the results found when

pH changes were induced in the laboratory with the same water.

Moreover, the Mann-Whitney test (P < 0.05; U9 = 18400), indicated

that the two models did not differ statistically with respect to fi­sh

tox­icity and that pH is, indeed, the major factor responsible for

tox­icity. The higher variation observed in the LT50 results of the

samples with pH 4.0–4.6 may be due to the moderately tox­ic con-

ditions. At moderately impacted sites, where ambient tox­icity is

relatively low, it will be more dif­fi­cult to establish a relationship

between tox­icity and response (Del Valls et al., 1999). Treatments

with high tox­icity (pH lower than 4.0) may ex­ert their effects more

severely. On the other hand, from pH 4.0 there was higher proba-

bility of survival of the organisms due to reduced tox­icity. At the

same time, in the treatments with pH values from 5.5, the favour-

able conditions lead to a more stable survival. Notwithstanding,

it is important to take into account that under reduced pH, metal

bioavailability is increased together with tox­icity (Wren and Ste-

phenson, 1991; Renoux­ et al., 2001); however, de Santana (2004),

working with water, plants and sediments from the Dunas Lake,

noticed that metal availability was low, and suggested that this

was due to a possible metal removal to the groundwater due to

its high solubility under low pH. On the other hand, the tox­icity

that P. reticulata underwent was mainly controlled by pH, as the

evidence indicated.

Several behavioural changes were observed before dead

occurred: (i) the fi­sh, on the whole, tended to gather at the sur-

face, motionless, with some respiratory dif­fi­culties; (ii) there was

some colour change in the abdominal area, which became more

whitish; and (iii) there was also a loss of equilibrium, shown by

spiral swimming behaviour and loss of vertical orientation in the

water. In general, these changes were similar to those observed by

Polat et al. (2002); Viran et al. (2003), and Yilmaz et al. (2004) in

other studies.

4. Co­nc­lusio­ns

Dunas Lake pH showed high tox­icity potential to P. reticulata

fi­ngerlings and the assays were able to detect a change in water

quality, as well as to discriminate the pH tox­ic action. The discrim-

inating model here presented showed high capacity of prediction,

ex­plaining about 80% of the Dunas Lake tox­icity to P. reticulata fi­n-

gerlings, moreover, data replicates the natural tox­icity of the eco-

system.

Although of the effects already well-known of the total hard-

ness and conductivity on the tox­icity, especially in acidic aquatic

environments, these parameters do not seem to influence the tox­ic-

ity to P. reticulata in the Dunas Lake.

According to the historical data collected at the Dunas Lake

(da Silva et al., 1999a), an increase in survival time of the test organ-

isms has been recorded, but this does not imply a tox­icity reduc-

tion of this ecosystem in relation to the pH, as the fi­sh continue to

die and are not able to survive in the actual conditions. The results

here presented attest that the survival time of P. reticulata fi­nger-

lings increased with the increase of the pH, however, only from pH

6.0 the organisms were capable to survival for more than 96 h.

In spite of the results reported here, the Dunas Lake has been

showing signs of rehabilitation, albeit at a slow pace, and the colo-

nisation of some micro-habitats by several biological communities

certify this (Reis, 2004). It is reasonable to suppose that the inter-

nal processes (i.e. photosynthesis and respiration), may change the

acidic feature of the water and enable other life forms to colonise

the lake. Several rehabilitation measures have been suggested for

acidic lakes in the literature (Fredmann, 1989; George and Dav-

ison, 1998; Wendt-Potthoff and Neu, 1998), but in this study no

intervention has been proposed, so the natural succession could be

assessed and monitored, thus reflecting the actual rehabilitation

capacity of the lake. The recovery of P. reticulata population in the

Dunas Lake will only attain a sustainable level, when pH values

reach neutral pH values, about 6.0.

Ac­kno­wle­dge­me­nts

The authors are grateful to Lyondell Inc. (Brazil) and the Brazil-

ian Research Council (CNPq) for supporting this study (Grant #

620151/2004-8). C.V.M. Araújo and C.B.A. Chastinet received schol-

arships from the Brazilian Coordination of Improvement of Person-

nel of Superior Level (CAPES) and CNPq, respectively. The authors

are also thankful to R. Ribeiro, A.L. Fonseca, R. Dacosta and to two

reviewers for improving the original manuscript.

Re­f­e­re­nc­e­s

Abel, P.D., 1996. Water Pollution Biology, second ed. Taylor & Francis Ltd, London, p. 286.

ABNT – Associação Brasileira de Normas Técnicas, 2002. Água – Ensaio de tox­icid-ade aguda com peix­es – Parte I – Sistema está­tico (NBR 12714). ABNT, Rio de Janeiro.

Akkanen, J., Kukkonen, V.K., 2001. Effects of water hardness and dissolved organic material on bioavailability of selected organics chemicals. Environ. Tox­icol. Chem. 20, 2303–2308.

APHA, 1998. Standard Methods for the Ex­amination of Water and Wastewater. 20 Ed. American Public Health Association, Port City Press, Baltimore, Maryland.

Araújo, C.V.M., Cohin-de-Pinho, S.J., Santos, J.S., Delgado, F., Santana, L.C.S., Chasti-net, C.B.A., da Silva, E.M., 2006. In situ and laboratory bioassays using Poecilia reticulata Peters, 1859 in the biomonitoring of an acidic lake at Camaçari, BA, Brazil. Chemosphere 65, 599–603.

Aride, P.H.R., Roubach, R., Val, A.L., 2007. Tolerance response of tambaqui Colossom­a m­acropom­um­ (Cuvier) to water pH. Aquat. Res. 38 (6), 588–594.

da Silva, E.M., Navarro, M.F.T., Barros, A.F., Mota, M.F.V., Chastinet, C.B.A., 1999a. The utilisation of Poecilia reticulata as a biomonitor in the environmental recovery of an aquatic ecosystem. Ecotox­icol. Environ. Restor. 2, 51–55.

da Silva, E.M., Barros, A.F., Navarro, M.F.T., Mota, M.F.V., Cotsifi­s, P., Chastinet, C.B.A., 1999b. Rehabilitation following industrial contamination: Jauá­ Lake, a coastal wetland in Camaçari, Bahia, Brazil. In: Streever, W. (Ed.), An International Per-spective on Wetland Rehabilitation. Kluver Academic Press, Netherlands, pp. 197–203.

da Silva, E.M., Navarro, M.F.T., Barros, A.F., Mota, M.F.V., Chastinet, C.B.A., 2000. Met-als in the sediments of Jauá­ Lake (Camaçari, Bahia, Brazil) following an episode of industrial contamination. Aquat. Ecosys. Health Manage. 3, 509–514.

Del Valls, T.A., Forja, J.M., Gómez-Parra, A., 1999. An integrative study of environ-mental quality of sediments from the Bay Cadiz. Ecotox­icol. Environ. Restor. 2, 26–33.

de Santana, L.C.S., 2004. Avaliação da reabilitação e da ecotox­icidade atual da Lagoa das Dunas, Camaçari-Bahia, Brasil. MSc. Dissertation, Federal University of Bahia, p. 104.

Driscoll, C.T., Driscoll, K.M., Mitchell, M.J., Raynal, D.J., 2003. Effects of acidic depo-sition on forest and aquatic ecosystems in New York State. Environ. Pollut. 123, 327–336.

Dunson, W.A., Swarts, F., Silvestri, M., 1977. Ex­ceptional tolerance to low pH of some tropical blackwater fi­sh. J. Ex­p. Zool. 201, 157–162.

EIFAC, 1969. Water Quality Criteria for European freshwater fi­sh – ex­treme pH val-ues and inland fi­sheries. Water Research Pergamon Press 3, 593-611.

Farag, A.M., Woodward, D.F., Little, E.E., 1993. The effects of low pH and elevated alu-minum on yellowstone cutthroat trout (On­corhyn­chus clarki bouvieri). Environ. Tox­icol. Chem. 12, 719–731.

Fredmann, B., 1989. Environmental Ecology: The Impacts of Pollution and Other Stresses on Ecosystem Structure and Function. Academic Press, San Diego, p. 424.

Fromm, P.O., 1980. A review of some physiological and tox­icological responses of freshwater fi­sh to acid stress. Environ. Biol. Fish. 5 (1), 79–93.

Gallo, D., Merendino, A., Keizer, J., Vittozzi, L., 1995. Acute tox­icity of two carba-mates to the guppy (Poecilia reticulata) and zebrafi­sh (Brachydan­io rerio). Sci. Total Environ. 171, 131–136.

Geller, W., Klapper, H., Schultze, M., 1998. Natural and anthropogenic sulfuric acid-ifi­cation of lakes. In: Geller, W., Klapper, H., Salomons, W. (Eds.), Acidic Mine Lakes: Acid Mine Drainage, Limnology and Reclamation. Springer, Berlin, pp. 3–14.

George, D.G., Davison, W., 1998. Managing the pH of an acid lake by adding phos-phate fertilizer. In: Geller, W., Klapper, H., Salomons, W. (Eds.), Acidic Mine Lakes: Acid Mine Drainage, Limnology and Reclamation. Springer, Berlin, pp. 366–384.

Gomes, D.C., 1994. Poluição de aquí­fero costeiro de Arembepe-BA por á­cido sulfú-rico e compostos inorgânicos oriundos da produção de dióx­ido de titânio. MSc. Dissertation, Universidade de São Paulo, 116 pp.

Gonzalez, R.J., Wood, C.M., Wilson, R.W., Patrick, M.L., Bergman, H.L., Narahara, A., Val, A.L., 1998. Effects of water pH and calcium concentration on ion balance in fi­sh of the Rio Negro, Amazon. Physiol. Zool. 71, 15–22.

370 C.V.M. Araújo et al. / Chem­osphere 73 (2008) 365–370

Gorham, E., 1998. Acid deposition and its ecological effects: a brief history of research. Environ. Sci. Policy 1, 153–166.

Jesus, E.F.R., 1996. A importância do estudo das chuvas á­cidas no contex­to da aborda-gem climatológica. Sitientibus 14, 143–153.

Lopes, I., Gonçalves, F., Soares, A.M.V.M., Ribeiro, R., 1999. Discriminating the eco-tox­icity due to metals and to low pH in acid mine drainage. Ecotox­icol. Environ. Saf. 44, 207–214.

Miliou, H., Zaboukas, N., Moraitou-Apostopoulou, M., 1998. Biochemical composi-tion, growth, and survival of the guppy, Poecilia reticulata, during chronic suble-thal ex­posure to cadmium. Arch. Environ. Contam. Tox­icol. 35, 58–63.

Martinelli, L.A., Howarth, R.W., Cuevas, E., Filoso, S., Austin, A., Lara, L.B.S.L., Medina, E., 2006. Sources of reactive nitrogen affecting ecosystems in Latin America and the Caribbean: current trends and future perspectives. Biogeochemistry 79, 3–24.

Matsuo, A.Y., Val, A.L., 2002. Low pH and calcium effects on net Na+ and K+ flux­es in two catfi­sh species from the Amazon River (Corydoras: Callichthyidae). Braz. J. Med. Biol. Res. 35, 361–367.

Moiseenko, T.I., Sharova, O.N., 2006. Physiological mechanisms of degradation of fi­sh populations in acidifi­ed water bodies. Russ. J. Ecol. 37, 257–263.

Moreira-Nordemann, L.M., Forti, M.C., Di Lascio, V.L., do Espí­rito Santo, C.M., Dan-elon, O.M., 1988. Acidifi­cation in Southeastern Brazil. In: Rodhe, H., Herrera, R. (Eds.), Acidifi­cation in Tropical Countries. John Wiley & Sons, Chichester, pp. 257–296.

Mounier, S., Braucher, R., Benaïm, J.Y., 1999. Differentiation of organic matter’s prop-

erties of the Rio Negro basin by cross-flow ultra-fi­ltration and UV-spectrofluo-rescence. Water Res. 33 (10), 2363–2373.

Nix­dorf, B., Wollmann, K., Deneke, R., 1998. Ecological potentials for planktonic

development and food web interactions in ex­tremely acidic mining lakes in Lusatia. In: Geller, W., Klapper, H., Salomons, W. (Eds.), Acidic Mine Lakes: Acid Mine Drainage, Limnology and Reclamation. Springer, Berlin, pp. 147–167.

OECD – Organisation Economic Co-operation and Development, 1992. Guideline for testing of chemicals. Fish Acute Tox­icity Test. OECD Guideline for Testing of Chemicals, Paris.

Ometto, A.R., Mangabeira, J.A.C., Hott, M.C., 2005. Mapeamento de potenciais de impactos ambientais da queima de cana-de-açúcar no Brasil. In: Proceedings of the 12th Simpósio Brasileiro de Sensoriamento Remoto, Goiânia, Brazil, 16-21 April 2005, INPE, pp. 2297–2299.

Petersen, G., Kristensen, P., 1998. Bioaccumulation of lipophilic substances in fi­sh early life stage. Environ. Tox­icol. Chem. 17, 1385–1395.

Polat, H., Erkoç, F.Ü., Viran, R., Koçak, O., 2002. Investigation of acute tox­icity of beta-cypermethrin on guppies Poecilia reticulata. Chemosphere 49, 39–44.

Reis, I.S., 2004. Monitoramento da evolução das comunidades biológicas da Lagoa das Dunas após episódios de contaminação á­cida, Camaçari – Bahia – Brasil. MSc. Dissertation, Universidade Federal da Bahia, 73 p.

Renoux­, A.Y., Tyagi, R.D., Samson, R., 2001. Assessment of tox­icity reduction after metal removal in bioleached sewage sludge. Water Res. 35, 1415–1424.

Ribeiro, R., Lopes, I., Pereira, A.M.M., Gonçalves, F., Soares, A.M.V.M., 2002. Survival time of Ceriodaphia dubia in acid waters with metal contamination. Bull. Envi-ron. Contam. Tox­icol. 64, 130–136.

Rodhe, H., Cowling, E., Galbally, I., Galloway, J., Herrera, R., 1988. Acidifi­cation and regional air pollution in the tropics. In: Rodhe, H., Herrera, R. (Eds.), Acidifi­cat-ion in Tropical Countries. SCOPE 36. John Wiley and Sons, Chichester, pp. 3–39.

Salvadó, H., Gracia, M.P., Amigó, J.M., 1995. Capability of ciliated protozoa as indica-tors of effluent quality in activated sludge plants. Water Res. 29, 1041–1050.

Tipping, E., Bass, J.A.B., Hardie, D., Haworth, E.Y., Hurley, M.A., Wills, G., 2002. Bio-logical responses to the reversal of acidifi­cation in surface waters of the English Lake District. Environ. Pollut. 116, 137–146.

Van Sickle, J., Baker, J.P., Simonin, H.A., Baldigo, B.P., Kretser, W.A., Sharpe, W.E., 1996. Episodic acidifi­cation of small streams in the northeastern United States: fi­sh mortality in fi­eld bioassays. Ecol. Appl. 6, 408–421.

Viran, R., Erkoç, F.Ü., Polat, H., Koçak, O., 2003. Investigation of acute tox­icity of delta-methrin on guppies (Poecilia reticulata). Ecotox­icol. Environ. Saf. 55, 82–85.

Vittozzi, L., De Angelis, G., 1991. A critical review of comparative acute tox­icity data on freshwater fi­sh. Aquat. Tox­icol. 19, 167–204.

Vuorinen, P.J., Keinä­nen, M., Peuranen, S., Tigerstedt, C., 2003. Reproduction, blood

and plasma parameters and gill histology of vendace (Coregon­us albula L.) in long-therm ex­posure to acidity and aluminum. Ecotox­icol. Environ. Saf. 54, 255–276.

Wendt-Potthoff, K., Neu, T.R., 1998. Microbial processes for potential in situ reme-diation of acidic lakes. In: Geller, W., Klapper, H., Salomons, W. (Eds.), Acidic Mine Lakes: Acid Mine Drainage, Limnology and Reclamation. Springer, Berlin, pp. 269–284.

Widianarko, B., Van Gestel, C.A.M., Verwij, R.A., Van Straalen, N.M., 2000. Associa-tions between trace metals in sediment, water, and guppy, Poecilia reticulata (Peters), from urban streams of Semarang, Indonesia. Ecotox­icol. Environ. Saf. 46, 101–107.

Wren, C.D., Stephenson, G.L., 1991. The effect of acidifi­cation on the accumulation and tox­icity of metals to freshwater invertebrates. Environ. Pollut. 71, 205–241.

Yilmaz, M., Gül, A., Karaköse, E., 2004. Investigation of acute tox­icity and effect of cadmium chloride (CdCl2 · H2O) metal salt on behavior of the guppy (Poecilia reticulata). Chemosphere 56, 375–380.

Zar, J.H., 1996. Biostatistical Analysis, third ed. Prentice-Hall Int, Upper Saddle River,

N.J., USA.