Copper and Zinc Content of Periphyton from Two Rivers As a Function of Dissolved Metal Concentration

Aquat. Sci. 64 (2002) 300-306 1015-1621/02/030300-07 © EAWAG, Diibendorf, 2002 IAquatic Sciences

Research Article

Copper and zinc content of periphyton from two rivers as a function of dissolved metal concentration

Renata Behra *, Rolf Landwehrjoha nn, Katrin Vogel, Bettina Wagner and Laura Sigg Swiss Federal Institute of Environmental Science and Technology (EAWAG), Oberlandstrasse 133, CH-8600 Diibendorf, Switzerland

Received: 26 November 2001; revised manuscript accepted: 13 July 2002

Abstract. Measuring the metal content of periphyton is a from two sites of the Thur River did not reflect the dis- means of evaluating the state of metal contamination in solved metal concentrations in the water, probably due to the environment with respect to levels of metals and their speciation effects. Nonexchangeable (intracellular) metal bioavailability. The aim of this study was to identify rela- determined after washing of periphyton samples with tionships between the dissolved metal concentrations and EDTA was variable and unpredictable when considering the total or intracellular metal concentrations in the peri- total metal content only. The different relationships in the phyton from two rivers. Metal levels at non-contaminated two rivers between the metal contents in periphyton and sites were comparable in both rivers (0.07-0. 71 µmol Cu the dissolved concentrations indicate the influence of g-1 dw and 0.22-4.36 µmol Zn g-1 dw). The metal content speciation on adsorption and bioavailability. Results are of periphyton from three sites of the Birs River was gen- discussed in respect to the significance of metal content erally reflective of the dissolved Cu and Zn concentra- as a tool for evaluating metal bioavailability. tions in the water. In contrast, metal content in periphyton

Key words. Copper; zinc; bioavailability; periphyton; EDTA non-exchangeable metal.

Introduction metal available for uptake by organisms, and hence their effects upon them. Contamination of the aquatic environment by heavy met- Studies on the metal content of periphyton (benthic als occurs as the result of various human activities. Water algal assemblages) also are being used to evaluate aquatic quality criteria are used for regulating environmental metal contamination and metal bioavailability. This ap- metal concentrations in order to protect living systems proach is based on the rationale that metal bioavailability from adverse effects of elevated ambient metal concen- affects the metal concentration adsorbed to and accumu- trations. Appropriate management tools are needed to lated in the organisms making up the community. Labo- provide information on metal concentrations, and their ratory studies have demonstrated that the bioavailability changes, in the environment. Usually chemical analyses of metals to algae is determined by the free metal ion ac- ofwater or sediment are employed to evaluate the state of tivity (Sunda and Guillard, 1976; Anderson et al., 1978). aquatic systems with respect to metal concentrations. Yet, In contrast to these controlled laboratory studies, studies this information does not give an indication on the vari- on periphyton have rarely examined the relationships ous forms of metals occurring in waters or the fraction of among metal content in benthic algal communities, the water concentration of the metals (Ramelow et al., 1987; Whitton et al., 1989; Gupta, 1996) and metal speciation. * Corresponding author phone: +41 1 823 5119; fax: +41 1 823 5315; e-mail: [email protected] Most studies only report on the metal content and the de- Published on Web: October 17, 2002 rived accumulation data are used directly to infer envi- Aquat. Sci. Vol. 64, 2002 Research Article 301 ronmental changes in metal concentrations. Moreover, Thur River, has background levels of metals. The “An- metal content measurements often rely on a single sam- delfingen” site at km 125 is a contaminated site affected ple despite the fact that data are subject to temporal vari- by agricultural and sewage inputs. Both sites are similar ability. Factors that may be important include the metal with respect to the geochemical background, consisting speciation in the water column, the periphyton composi- mostly of carbonate rocks. tion in terms of algal species and the quantity of metal oxides associated with the periphyton mat. These factors are all subject to temporal dynamics in the environment. The Sampling and analytical methods appropriateness of periphyton as an indicator of bioavail- Sampling of periphyton, sediments and water was coor- ability of metals thus depends on an understanding of the dinated. The Birs sites were sampled 3–5 times for deter- variability of the metal content in communities from un- mination of intracellular metal (see below) over nine polluted sites. Before using metal content data for man- months starting in June 1996. The Thur sites were sam- agement purposes, relationships between content and pled 4 times over six months starting in November 1998. metal concentrations should be made clear and variabil- The river bed of the Birs River was covered with stones. ity should be quantified. This is particularly important at Fine sediments were only present as a thin layer on the metal concentrations lying only slightly above corre- stones and were collected as a slurry. The slurry of sur- sponding water quality criteria limits. Previous studies on face sediments was collected from the Birs River bed by the long-term effects of metals on periphyton did show sucking with a hand pump producing a vacuum in a flask the sensitivity of some algal species towards very low connected to teflon tubing. This sediment suspension was metal concentrations (Gustavson and Wängberg, 1995; collected in 2-L polyethylene bottles. In the laboratory, Soldo and Behra, 2000). the sediments were allowed to settle, and then sieved se- In this study, we have measured the temporal and spa- quentially through 122 μm- and 63 μm-mesh nets. The tial distribution of copper and zinc in two rivers moder- fraction <63 μm was freeze-dried and stored for analysis. ately polluted by metals. Our aim was to quantitatively Aliquots (40–50 mg) were digested with concentrated evaluate the relationship between ambient dissolved HNO3 (4 mL) and H2O2 (1 mL) in a microwave digestion metal concentrations and the metal content of periphyton system (Microwave Laboratory Systems, mls 1200 by collecting water and periphyton simultaneously. mega). The digested samples were diluted to 50 mL un- Moreover, we examined and compared intra-site variabil- der a clean bench. The concentrations of Zn and Cu were ity of metal content in periphyton taken from stone sur- measured by ICP-OES (Perkin-Elmer Elan 5000). The faces and from synthetic glass substrates. The results are accuracy of the metal measurements in sediment samples discussed in respect to the utility of periphyton for show- was checked regularly using a reference sediment (NBS ing patterns of metal contamination in rivers. Buffalo River sediment 2704). Water samples for metal analysis were collected in clean, acid-washed polypropylene bottles and placed in Materials and methods plastic bags in the field. Plastic gloves were worn for sample handling. Samples were collected by hand from Study sites the Birs River and using a peristaltic pump from the Thur. The Birs River (Switzerland) flows over a length of 73 km Before measurement, samples from the Birs were filtered from the Jura region to the Rhine River through a catch- under a clean bench using an acid-cleaned polysulfone ment dominated by carbonate rocks. Its alkalinity ranges filtration unit and acid-cleaned cellulose nitrate filters from 2 to 4 mM and its pH from 8.0 to 8.5. Based on pre- (0.45-μm, Sartorius), and acidified to 0.01 M HNO3. De- vious metal analyses, the Birs is considered as moder- tails of the analytical procedure to determine dissolved ately contaminated (Jakob et al., 1994). Three sampling metal concentrations in the Thur samples have been re- sites were selected along the first 35 km of the river as ex- ported previously (Sigg et al., 2000). amples for various pollution levels. Site 1, which is un- Cu in whole and in filtered water samples was mea- contaminated by metals, is located 2.9 km from the river sured directly by graphite furnace AAS (Varian GTA-95) source. Site 2, at km 6.1, is located downstream from a for the Birs samples and by ICP-MS for the Thur samples. water treatment plant that receives the input of a metal- In the Birs samples, Zn was determined by flame AAS lurgical plant. Site 3 is located 9 km downstream from a after preconcentration as described in Sigg et al. (1996). metallurgical plant at km 34.1. Zn in the Thur samples was determined by ICP-MS. The The Thur River (Switzerland) flows over a length of detection limits were 0.2μg L–1 (3 nM) for Cu and Zn in 134 km from the Toggenburg region to the Rhine River. the Birs samples, and 0.01 μg L–1 (0.2 nM) for Cu and Zn in Its alkalinity ranges from 3.5 to 4.5 mM and its pH from the Thur samples, as determined from calibration curves. 8.0 to 8.5. Two sampling sites were selected. The upper- Field blanks (bottles treated in the same way as sam- most site (“Necker”), at the mouth of a tributary of the ples) were taken on a regular basis. Blank values were 302 R. Behra et al. Copper and zinc availability in two rivers

<0.03 μg L–1 (0.4 nM) for Cu and < 0.3 μg L–1 (5 nM) for Statistical analysis Zn. The accuracy of the metal determinations by ICP-MS Metal content data were evaluated statistically in two was checked on a regular basis using SLRS-3 reference ways. First, Cu and Zn content in periphyton were re- water (National Research Council Canada). gressed against metal concentrations in the river water. Teflon racks, holding eight microscope slides (76 ¥ Second, differences in metal content data between sites 26 mm) on each side, were used for periphyton coloniza- were analyzed using one-way analysis of variance tion in the Birs River. Three racks were fixed within a 6 ¥ (ANOVA). Statistical significance of differences in metal 3 m area at each site with the slides sitting about 20 cm content between sites was examined using Tukey’s Hon- below the water surface, and their long axes parallel to the est Significant Difference Test for means with different current. Periphyton from each of the three racks was ex- sample sizes. The tests were performed on log10 trans- amined separately by sampling 2–4 slides from each formed data after having tested for normality and homo- rack. Periphyton from the Thur River was collected by geneity of variance using Kolmogorov’s and Levene’s scraping biotic and abiotic material from stones with test, respectively. All statistical analysis were computed glass slides. On each sampling date periphyton was sepa- with the software package STATISTICA (StatSoft Inc., rately sampled from 5 locations at each site of the two 1994). rivers. Periphyton samples were transported to the laboratory in bottles containing river water collected at the same site and analyzed separately. Results Periphyton samples for metal analyses were sus- pended in filtered river water (0.45 μm, Sartorius). Sam- Concentrations of Cu and Zn in water and periphyton ples were examined microscopically and were found to be from the Birs and Thur Rivers are presented in Tables 1, a homogenous suspension of algae and bacteria. Periphy- 2, 3 and in Figures 1 and 2. Each point represents the av- ton was separated from sediment particles (inorganic ma- erage of three to five samples collected separately during terial) by decantation and 2–3 washings with river water. one sampling campaign. In the Birs River, Cu and Zn Mineral particles were found embedded in periphyton content in periphyton generally increased with increasing collected from stones (Thur), whereas only minimal dissolved metal concentrations, although with a certain amounts of these particles were found on the glass slides variability (Fig. 1). Significant linear regressions were (Birs), presumably because the vertical orientation of the found between Cu concentrations in periphyton and wa- slides with respect to the water surface minimized parti- ter (r 2 = 0.25, P < 0.005). For Zn, the slope of the regres- cle entrapment. Three aliquots of each final suspension sion became highly significant (r2 = 0.79, P < 0.005) were filtered through 0.45 μm pore-sized acid-washed when the extreme low content value corresponding to cellulose nitrate filters (Sartorius). The filters were dried 563 nM Zn in water was omitted from the analysis. This for 15 h at 50°C and the dry weight of the periphyton was value was not considered since it was obtained from ana- measured. The filters were placed in Teflon digestion lyzing only one instead of three periphyton samples. The flasks containing 4 mL 65% HNO3 and 1 mL of 30% three sites of the Birs differed in their dissolved Cu and H2O2 . Periphyton was digested in this mixture for 13 min Zn concentrations (Table 1). Site 3 was mostly contami- in a microwave oven. The digestion solutions were trans- nated with Cu, and site 6 with Zn. Correspondingly, the ferred into graduated flasks, and the volume was adjusted copper content in periphyton at site 3 was significantly to 25 mL with nanopure water. higher than at site 1 (P = 0.0001) and 6 (P = 0.0001). Zn In order to discriminate between metal adsorbed to in periphyton was significantly higher at site 6 in com- abiotic or biotic material and intracellular metal, three parison to the background site 1 (P = 0.001). The increase aliquots of each sample were treated with 4 mM EDTA at in dissolved Cu and Zn concentrations also was clearly pH 8.0 for 10 min. These conditions were applied because visible in the sediment concentrations (Table 1). periphyton samples are predominantly composed of algae Cu and Zn concentrations in the Thur were lower than and bacteria. EDTA washings were carried out in homo- in the Birs, both for water and periphyton. Although dis- geneous suspensions. The intracellular accumulated solved Cu and Zn concentrations increased from the metal was operationally defined as the metal content de- Necker to the Andelfingen sites, the Cu and Zn contents termined after the EDTA wash (Knauer et al., 1997). Ad- in periphyton did not clearly differ between these sites sorbed metal was calculated as the difference between the (Table 3, Fig. 2). The concentrations of Cu and Zn in pe- metal content before and after washing with EDTA. riphyton collected from the unpolluted Necker site and Metal concentrations in periphyton from the Birs were from the site 1 of the Birs were in a similar range (Tables measured by furnace (Cu) and flame (Zn) atomic absorp- 1, 2, and 3). tion spectrometry (AAS, Perkin Elmer 5000). Metal con- The non-exchangeable Cu and Zn content in periphy- centrations in periphyton from the Thur were measured ton after the EDTA wash were generally similar to those by ICP-MS (Perkin Elmer 5000). determined without an EDTA wash (Tables 2 and 3). At Aquat Sci. Vol. 64, 2002 Research Article 303

s: 2.0 s: 6 "O "O ';' A ';' A Cl Cl 5 0 1.5 0 E :I. ~ 4 c f c ~ 1.0 ~3 j .s::. .s::. 0. ·c .g. 2 • (l.J (l.J 0. 0. + 0.5 .S .s 1 ) j :i' :i' _.# ... Q. Q. 0 J 0.0 t 0 50 100 150 200 0 10 20 30 40 Dissolved Cu concentration (nM) Dissolved Cu concentration (nM)

s: 50 s: 5 "O "O B ';' B '0> 40 . Cl 4 0 0 30 ~3 ~c c 0 ~ >. -a 20 :9. 2 iD ·~ 0. 10 1 I .s .s c • c tit f f f f !:::!. o~'~·JA-·~~~~....-~~~....-...... -~~....-~--1 !:::!. 0 • 0 200 400 600 0 10 20 30 40 50 60 Dissolved Zn concentration (nM) Dissolved Zn concentration (nM)

Figure I. Concentration of Cu (A) and Zn (B) (µmot g-1 dw) in Figure 2. Cu(A) and Zn (B) (µmot g- 1 dw) in periphyton from the periphyton from the River Birs as a function of dissolved Cu (nM) River Thur as a function of dissolved Cu (nM) and Zn (nM). Dia- and Zn (nM). Circles: site I; squares: site 3; triangles: site 6 (see monds: Necker site; squares: Andelfingen site (see text). Bars show text). Bars show mean± I S.D. (n = 3-6). mean ± I S.D. (n = 4).

Table I. Copper and zinc concentration in water, periphyton and sediment from the Birs River.

Site Sampling date Cu in water Cu in periphyton• Cu in sediments Zn in water Zn in periphyton• Zn in sediments nM µmot g- 1 dw µmot g- 1 dw nM µmot g- 1 dw µmolg- 1 dw

15.08.1996 16 0.55 ± 0.06 0.6 38 4.36 ± 0.09 3.7 19.09.1996 9 0.40 ± 0,07 0.6 14 1.15 ± O.Q2 2.6 25.09.1996 6 0,07 ± 0.03 0.8 nd nd nd 17.10.1996 14 0.26 ± 0.03 1.0 11 0.22 ± 0.01 3.3 31.10.1996 17 0.68 ± 0.04 1.3 nd nd nd

3 30.08.1996 154 2.46 ± 0.15 3.8 150 12.70 ± 2.86 13.3 19.09.1996 41 3.93 ± 1.56 4.7 95 16.33 ± 3.40 6.0 17.10.1996 69 3.91 ± 0.46 6.8 181 13.49 ± 2.45 13.6 31.10.1996 88 1.89 ± 0.19 7.3 nd nd nd

6 15.08.1996 42 0.30 ± 0.07 1.3 563 (5.64)b 8.1 30.08.1996 20 0.59 ± 0.04 135 16.57 ± 0.64 nd 19.09.1996 16 0.40 ± 0.03 0.7 nd nd nd 25.09.1996 19 I .I I ± 0.04 0.9 nd nd nd 17.10.1996 22 2.13 ± 1.57 0.9 280 16.58 ± 4.55 12.3 31.10.1996 16 1.27 ± 0,07 1.3 378 38.07 ± 3.22 14.3

• values are means ± SD, n = 3. b n =I. nd, not determined. 304 R. Behra et al. Copper and zinc availability in two rivers

Table 2. Total and non-exchangeable (intracellular) concentration of Cu and Zn in periphyton from the Birs River. Values are means ± SD, n = 3.

Site Sampling date Cu total Cu Zn total Zn μmol g–1 dw non-exchangeable μmol g–1 dw non-exchangeable μmol g–1 dw μmol g–1 dw

1 14.11.1996 0.68 ± 0.04 0.55 ± 0.01 3.40 ± 0.47 3.19 ± 0.91 19.12.1996 0.23 ± 0.12 0.20 ± 0.01 1.89 ± 0.18 1.82 ± 0.25 16.01.1997 0.36 ± 0.23 0.31 ± 0.07 0.49 ± 0.02 0.38 ± 0.05 19.03.1997 0.23 ± 0.04 0.21 ± 0.02 2.58 ± 0.37 0.42 ± 0.02 09.04.1997 nd nd 2.76 ± 0.52 1.01 ± 0.44 3 14.11.1996 1.89 ± 0.19 1.41 ± 0.12 nd nd 19.12.1996 3.06 ± 0.39 1.95 ± 0.19 12.94 ± 0.69 5.05 ± 0.34 16.01.1997 1.49 ± 0.35 1.30 ± 0.19 30.65 ± 1.86 13.09 ± 1.73 19.03.1997 9.36 ± 1.31 6.49 ± 0.84 7.62 ± 3.07 2.14 ± 1.15 09.04.1997 1.08 ± 0.11 0.87 ± 0.11 6.04 ± 1.83 1.35 ± 0.44 6 14.11.1996 2.13 ± 1.57 0.59 ± 0.08 38.07 ± 3.22 22.45 ± 1.07 19.12.1996 0.58 ± 0.15 0.44 ± 0.04 30.40 ± 1.66 27.84 ± 0.04 19.03.1997 2.54 ± 0.09 1.69 ± 0.26 34.03 ± 2.28 24.74 ± 4.30 09.04.1997 0.68 ± 0.18 0.49 ± 0.10 12.3 ± 2.34 7.29 ± 0.64 nd, not determined.

Table 3. Copper and zinc concentration in water and periphyton (total and non-exchangeable) from the Thur River.

Site Sampling date Cu in Cu total Cu non- Zn in Zn total Zn non- water periphyton a exchangeable water periphyton a exchangeable nM μmol g–1 dw periphyton a nM μmol g–1 dw periphyton a μmol g–1 dw μmol g–1 dw

Necker 07.05.1998 3 0.21 ± 0.18 0.18 ± 0.04 1 0.58 ± 0.20 0.31 ± 0.08 26.06.1998 7 0.22 ± 0.06 0.26 ± 0.08 2 0.48 ± 0.12 0.51 ± 0.16 24.09.1998 12 0.23 ± 0.09 0.21 ± 0.05 25 0.66 ± 0.23 0.52 ± 0.09 20.11.1998 7 0.17 ± 0.04 0.18 ± 0.03 6 0.41 ± 0.09 0.38 ± 0.06 Andelfingen 07.05.1998 10 0.50 ± 0.23 0.61 ± 0.28 7 1.03 ± 0.44 1.00 ± 0.28 26.06.1998 24 0.67 ± 0.28 0.78 ± 0.32 18 2.26 ± 0.83 2.54 ± 1.31 24.09.1998 31 0.52 ± 0.27 0.55 ± 0.15 51 0.81 ± 0.43 0.79 ± 0.41 20.11.1998 23 0.25 ± 0.06 0.24 ± 0.06 33 0.76 ± 0.36 0.62 ± 0.25 a values are means ± SD, n = 5.

Table 4. Within-site variability of Cu and Zn concentrations in periphyton from the Birs and Thur Rivers without and after an EDTA wash. Values are means of coefficients of variation (%) across all sampling dates (range in parentheses).

Site Cu, without EDTA Cu, after EDTA wash Zn, without EDTA Zn, after EDTA wash

Birs 1 25 (6 – 66) 10 (2 – 21) 7 (1 – 19) 21 (6 – 43) Birs 3 15 (6 – 40) 12 (8 – 15) 19 (1 – 19) 27 (7 – 54) Birs 6 20 (4 – 73) 15 (9 – 21) 10 (4 – 27) 8 (5 – 17) Necker 44 (25 – 83) 2 (16 – 32) 29 (22 – 35) 22 (16 – 31) Andelfingen 41 (23 – 53) 34 (22 – 46) 45 (37 – 53) 43 (28 – 52)

the non-contaminated site in both rivers, non-exchange- Within-site variability of metal content in periphyton able Cu represented 80 to around 100% of the total mea- was estimated for all study sites (Table 4). Coefficients of sured content. At other sites, non-exchangeable contents variation were generally lower for periphyton from the were as low as 28% of the total Cu. Non-exchangeable Zn Birs, which had developed on glass slides, than for peri- was generally high and ranged from 52 to around 100% in phyton from the Thur, which was collected from stones. samples from all sites in both rivers, though samples with Within-site variability was similar at non-contaminated only 16% intracellular bound Zn were found (Table 3). and contaminated sites. Aquat. Sci. Vol. 64, 2002 Research Article 305

Discussion metal uptake by periphyton at higher Cu concentrations. A similar effect is conceivable for Zn, but the Zn specia- The metal concentrations in periphyton from the River tion was not examined in these samples. The free ion ac- Birs differed among sites and temporally within sites. tivity of competing or interacting metal ions may also Generally, increases in dissolved Cu and Zn concentra- have influenced metal uptake. Culture experiments with tions in the water were reflected by increased metal con- Chlamydomonas sp. have indicated complex antagonistic centrations in periphyton (Fig. 1) and sediments. Deter- interactions between Cu, Zn and Mn in controlling Zn mination of concentration ranges for each site are clearly and Mn uptake (Sunda and Huntsman, 1998). indicative of differences in ambient metal concentrations The metal measured in periphyton samples includes among the sites (Tables 1 and 2). The Cu content of peri- the fractions taken up by organisms and those interacting phyton from the non-contaminated site in the River Birs with abiotic components of the periphyton assemblages. varied ten-fold from 0.07 to 0.70 μmol Cu g–1 dw, with a The latter includes a heterogeneous organic polysaccha- three-fold variation in water concentration (6–17 nM). A ride matrix that varies in composition according to the bi- similar but narrower range (0.17 to 0.23 μmol Cu g–1 dw) otic species composition, as well as inorganic material was found in periphyton from the non-contaminated site that may be variably rich in manganese and iron oxides of the Thur. Though comparative field data from non- (Newman and McIntosh, 1982). In this study, biouptake contaminated sites are scarce, Ramelow et al. (1987) and was estimated by measuring metal content in periphyton Gupta (1996) reported values for Cu contents in periphy- after washing the samples with EDTA. Due to its strong ton in the range found in this study. Similar values also binding of metals, EDTA is expected to complex any were reported for Cladophora, a filamentous algal metal adsorbed to the abiotic part of periphyton or to or- species previously suggested to be suitable for monitor- ganism surfaces, whereas metal bound in particles may ing heavy metals in waters (Whitton et al., 1981). In this not be removed. Comparisons of content data before and alga, water concentrations of 5 nM were associated with after the EDTA wash indicated that a variable fraction of a Cu content of 0.4 μmol g–1 dw (Oertel, 1991). In labo- the total metal content could be removed by EDTA. At the ratory studies the green alga Scenedesmus subspicatus non-contaminated Birs site and at both sites of the Thur, was found to accumulate 0.03 μmol g–1 dw at 10–14 M free the major part of total Cu could not be removed. At the Cu2+ concentrations (Knauer et al., 1997). To compare the two other sites, and at all sites in both streams for Zn, Cu content in periphyton from various origins, informa- non-exchangeable metal ranged from 16 to about 100% tion about the Cu speciation in water is needed. of the total content despite the fact that samples were se- At the non-contaminated site in the River Birs, dis- lected for their thick algal mat. Moreover, a major part of solved Zn ranged from 11 to 38 nM and Zn in the peri- fine inorganic sediments were removed from the samples phyton ranged from 0.2 to 4.4 μmol g–1 dw. Comparable through washing before metal analyses. The presence of values have been found in periphyton (Ivorra et al., 1999; Mn and Fe oxides also could influence the sorption of Lehmann et al., 1999) and in Cladophora (Whitton et al., metals to periphyton, as shown for nickel (Gray et al., 1989) at non-contaminated river sites. The comparison of 2001). Material of biological origin, such as the organic the Cu and Zn content data in this study with data from polysaccharide matrix in which periphytic organisms are the literature shows that values are similar. This is an in- embedded (Genter, 1996), also may influence the specia- teresting point because the data concern periphyton as- tion locally and thus the availability of metals to the or- semblages that differ in their biotic composition or even ganisms within the periphytic mat. This depends on single algal species, and they suggest a similar relation- whether the free ion concentration at the surface of algal ship between Cu or Zn content and the corresponding and bacterial cells within the periphyton assemblage is at aqueous metal concentration. equilibrium with that in bulk river water. In contrast to what was found in the Birs, increased Coefficients of variation for metal content were com- dissolved Cu and Zn concentrations in the River Thur pared between the Thur and the Birs in order to evaluate were not clearly reflected in an increased Cu content in the utility of collecting periphyton on artificial substrates periphyton (Fig. 2). This may be due to speciation effects to reduce within-site variability in metal content. Indeed, of Cu and Zn. In the case of copper, Cu2+ concentrations Cu and Zn content variability was lower for periphyton were within the same narrow range (log [Cu2+] = –15.1 to growing on artificial substrates than on natural substrates –15.6) at both sites (Sigg et al., 2000). According to the (Table 4). The high intra-site variability in the Thur sam- free-ion activity model, the free-metal ion activity deter- ples may be explained by differences in geometry and age mines the uptake and effects of metals in aquatic organ- of the organisms making up the periphyton mat. Thick isms (Sunda, 1988–1989; Campbell, 1995). In the River mats collected from stones include dead organisms lo- Thur, the increase in dissolved Cu is matched by an in- cated in deeper, packed layers that were observed to be al- crease in ligand concentration and stability that keeps most dry. Metal diffusion to these layers may be negligi- Cu2+ at low levels (Sigg et al., 2000), thereby limiting ble, and thus not respond to changes in ambient metal 306 R. Behra et al. Copper and zinc availability in two rivers concentration. However, upper layers include living or- Gustavson, K. and S.-A. Wängberg, 1995. Tolerance induction and ganisms associated within a more fluffy layer similar to succession in microalgae communities exposed to copper and atrazine. Aquat. Toxicol. 32: 283–302. that found on synthetic substrates. These algal layers are Hill, W. R., A. T. Bednarek and L. Larsen, 2000. Cadmium sorption exposed to the water column, which may allow metals to and toxicity in autotrophic biofilms. Can. J. Fish. Aquat. Sci. diffuse into the layer, thus better reflecting changes in 57: 530–537. ambient water concentration. Ivorra, N., J. Hettelaar, G. M. J. Tubbing, M. H. S. Kraak and W. Admiraal, 1999. Translocation of microbenthic algal assem- Periphyton has frequently been used for the identifi- blages used for in situ analysis of metal pollution in rivers. cation of metal-contaminated sites (Ramelow et al., 1992; Arch. Environ. Contam. Toxicol. 37: 19–28. Kiffney and Clements, 1993; Beltman et al., 1999; Jakob, A., J. Zobrist, J. Davis, P. Liechti and L. Sigg, 1994. Lehmann et al., 1999; Hill et al., 2000), but to our knowl- NADUF-Langzeitbeobachtung des chemisch-physikalischen Gewässerzustandes. Gas, Wasser, Abwasser 74: 171–186. edge ranges in metal content that reflect background sol- Kiffney, P. M. and W. H. Clements, 1993. Bioaccumulation of heavy uble metal concentrations have not been determined be- metals by benthic invertebrates at the Arkansas river, Colorado. fore. This information is necessary in order to distinguish Environ. Toxicol. Chem. 12: 1507–1517. 2+ between background contents and those occurring at ele- Knauer, K., R. Behra and L. Sigg, 1997. Effects of free Cu and Zn2+ ions on growth and metal accumulation in freshwater vated metal concentrations. The results of this study con- algae. Environ. Toxicol. Chem. 16: 220–229. firm the appropriateness of using periphyton collected Lehmann, V., G. Tubbing and W. Admiraal, 1999. Induced metal from artificial substrates for monitoring ambient metal tolerance in microbenthic communities from three lowland rivers with different metal loads. Arch. Environ. Contam. Toxi- concentrations, also when dissolved concentrations are col. 36: 384–391. only slightly higher than corresponding background con- Newman, M. C. and A. W. McIntosh, 1982. Slow accumulation of centrations. The different dependence in the two rivers of lead from contaminated food sources by the freshwater gas- the metal contents in periphyton from the dissolved con- tropods, Physa integra and Campeloma decisum. Arch. Envi- ron. Contam. Toxicol. 12: 685–692. centrations suggests an influence of speciation on ad- Oertel, N., 1991. Heavy-metal accumulation in Cladophora glom- sorption and bioavailability. Further investigations are erata (L.) Kütz in the River Danube. Ambio 20: 264–268. needed to better understand the relationships between Ramelow, G. J., S. L. Biven, Y. Zhang, J. N. Beck, J. C. Young, J. D. metal speciation and metal content in periphyton. Callahan and M. F. Marcon, 1992. The identification of point sources of heavy metals in an industrially impacted waterway by periphyton and surface sediment monitoring. Water, Air, Soil Poll. 65: 175–190. Acknowledgments Ramelow, G. J., R. S. Maples, R. L. Thomson, C. S. Mueller, C. Webre and J. N. Beck, 1987. Periphyton as monitors for heavy Our thanks go to G. P. Genoni for his comments and revi- metal pollution in the Calcasieu river estuary. Environ. Poll. 43: sion work. We are also grateful to D. Kistler for assistance 247–261. Sigg, L., D. Kistler and M. Ulrich, 1996. Seasonal variations of zinc in the field. in a eutrophic lake. Aquat. Geochem. 1: 313–328. Sigg, L., H. Xue, D. Kistler and R. Schönenberger, 2000. Size fractionation (dissolved, colloidal and particulate) of trace References metals in the Thur river, Switzerland. Aquat. Geochem. 6: 413–434. Anderson, M. A., F. M. M. Morel and R. R. L. Guillard, 1978. Soldo, D. and R. Behra, 2000. Long-term effects of copper on the Growth limitation of a coastal diatom by low zinc ion activity. structure of freshwater periphyton communities and their Nature 276: 70–72. tolerance to copper, zinc, nickel and silver. Aquat. Toxicol. 47: Beltman, D. J., W. H. Clements, J. Lipton and D. Cacela, 1999. 181–189. Benthic invertebrate metals exposure, accumulation and com- Sunda, W. G., 1988–1989. Trace metal interactions with marine munity-level effects downstream from a hard-rock mine site. phytoplankton. Biol. Ocean. 6: 411–442. Environ. Toxicol. Chem. 18: 299–307. Sunda, W. G. and R. R. L. Guillard, 1976. The relationship between Campbell, P. G. C., 1995. Interactions between trace metals and cupric ion activity and the toxicity of copper to phytoplankton. aquatic organisms: a critique of the free-ion activity model. In J. Mar. Res. 34: 511–529. Metal speciation and bioavailability in aquatic systems. A. Sunda, W. G. and S. Huntsman, 1998. Interactions among Cu2+, Tessier and D. A. Turner. Wiley, New York, pp. 45–102. Zn2+, and Mn2+ in controlling cellular Mn, Zn, and growth rate Genter, R. B., 1996. Ecotoxicology of inorganic chemical stress to in the coastal alga Chlamydomonas. Limnol. Oceanogr. 43: algae. In Algal ecology: freshwater benthic systems. R. J. 1055–1064. Stevenson, M. L. Bothwell and R. L. Lowe. Academic Press, Whitton, B. A., I. G. Burrows and M. G. Kelly, 1989. Use of San Diego, CA, pp. 403–468. Cladophora glomerata to monitor heavy metals in rivers. J. Gray, B. R., W. R. Hill and A. J. Stewart, 2001. Effects of develop- App. Phycol. 1: 293–299. mental time, biomass and ferromanganese oxides on nickel Whitton, B. A., P. J. Say and J. D. Wehr, 1981. Use of plants to sorption by stream periphyton. Environ. Poll. 112: 61–71. monitor heavy metals in rivers. In Heavy metals in Northern Gupta, A., 1996. Heavy metals in water, periphytonic algae, detri- England: environmental and biological aspects. P. J. Say and B. tus, and insects from two streams in Shillong, northeastern A. Whitton. Durham, England, Deartment of Botany, Univer- India. Environ. Mon. Ass. 40: 215–223. sity of Durham, pp. 135–145.