- Email: [email protected]
Natural variation of copper, zinc, cadmium and selenium concentrations in Bembicium nanum and their potential use as a biomonitor of trace metals
Water Research 37 (2003) 2173–2185
Natural variation of copper, zinc, cadmium and selenium concentrations in Bembicium nanum and their potential use as a biomonitor of trace metals D. Gay*, W. Maher Ecochemistry Laboratory, University of Canberra, University Drive, Canberra ACT 2601, Australia Received 17 May 2002; accepted 22 November 2002
Abstract Copper, zinc, cadmium and selenium were measured in the gastropod mollusc Bembicium nanum at two uncontaminated locations, Jervis Bay and Rosedale, NSW, to determine natural variability of metals associated with gender, mass, shore position and temporal variability. Trace metals were also measured in B. nanum at three industrialised locations to determine the accumulation of trace metals in contaminated environments. Copper, zinc, cadmium and selenium concentrations were not significantly different between male and female B. nanum. No significant relationships were found between zinc, cadmium and selenium concentrations and mass. There was a significant relationship between copper concentration and mass but only 19% of the variation was explained by mass. Generally inherent variability within samples had a greater influence than gender or variations in mass on trace metal concentrations. No trend was found in cadmium and selenium concentrations with variation in shoreline position. Copper and zinc concentrations increased further away from the low tide mark, with a decrease in metal concentrations at the furthest site from the water. Variability in metal concentrations is attributable to variations in food source, food availability and different immersion times. Copper, zinc, cadmium and selenium concentrations varied over a 12-month period. Copper, cadmium and selenium were taken up and lost over time, as metal body burden followed the same trend as metal concentrations. Zinc concentrations were influenced by mass. Copper and cadmium concentrations fluctuated throughout the 12-month period but with no clear seasonal trends. Selenium concentrations peaked in spring (October), with concentrations remaining uniform over the other months. These differences in mean concentrations between months were most likely due to inherent trace metal variability associated with differences in food availability and changes in metabolic rates associated with changes in temperature during the study period. Measurement of trace metals in B. nanum at contaminated sites showed that B. nanum accumulates metals in response to contamination. B. nanum meets most of the requirements to be a biomonitor of trace metal contamination as they are abundant, sedentary, easy to identify, provide sufficient tissue for analysis, tolerate high concentrations of pollutants and they accumulate trace metals in response to contamination. However, as trace metal concentration can vary with mass, shoreline position and temporally, care must be taken to collect individual organisms with similar mass from similar shoreline positions and times. r 2003 Elsevier Science Ltd. All rights reserved. Keywords: Trace metals; Biomonitors; Bembicium nanum; Natural variation; Accumulation
*Corresponding author. 0043-1354/03/$ - see front matter r 2003 Elsevier Science Ltd. All rights reserved. doi:10.1016/S0043-1354(02)00622-X
2174
D. Gay, W. Maher / Water Research 37 (2003) 2173–2185
1. Introduction Many of the world’s population reside along coastal areas. With human settlement comes the building of infrastructure, land reclamation for port and industrial development, habitat modification, tourism and recreational activities. Industry situated along the coast has better access to shipping ports and easier disposal of industrial waste, which may be released into marine waters. Furthermore, domestic sewage effluent is pumped from outfalls located along the near shore regions. Trace metal contamination has been identified as a concern in coastal environments, due to discharges from smelters and mining operations and from storm water and sewage disposal [1,2]. As trace metals can have a deleterious effect on aquatic ecosystems, it is important to know where bioavailable metals are entering aquatic systems and take remedial management action. Organisms that concentrate metals in their tissues are often used to indicate and quantify contaminant levels or bioavailability of contaminants in the environment [3–7]. Generally bivalves such as mussels are used for this purpose. To be used as biomonitors organisms should be: *
*
*
*
Sessile or sedentary thus being representative of the study area. Hardy, tolerating high concentrations of pollutants and large ranges in salinity, and permitting laboratory studies. Abundant in study areas, easy to identify and sample and should provide sufficient tissue for analysis of contaminants of interest. Have a simple correlation between the contaminant concentration found in their tissues and the average bioavailable pollutant concentration. This correlation should be the same for all study sites.
Many species of marine molluscs have been shown to reflect ambient metal concentrations (e.g. [8–10]). Yet many physiological and environmental factors can combine to influence the overall metal concentrations found in marine molluscs. Intrinsic factors such as mass and size, gender, reproductive state, accumulation/ regulation strategies, and diet and extrinsic factors such as temperature, salinity, supply of metals, food availability and metal–metal relationships can influence trace metal concentrations [6,7,11,12]. Along the Australian coast, rocky intertidal platforms are common. These platforms serve as habitat for gastropods and other species. A widely distributed and common gastropod present in this intertidal region is B. nanum. Few studies worldwide have used gastropods in monitoring studies [13–17,9,10]. B. nanum meets many of the requirements of a suitable biomonitor. However, before B. nanum can be used in monitoring studies, an
estimate of natural trace metal variability due to intrinsic and extrinsic factors need to be assessed [18]. The aim of this study was to examine the factors that are contributing to the natural variability in trace metal concentrations in B. nanum at non-contaminated sites by the determination of: *
*
*
Variation in metal concentrations attributable to gender and changes in mass. Variation in metal concentrations with shoreline position. Temporal metal concentration variation and the relationship with total metal body burden and mass.
As well, the accumulation of trace metals by B. nanum in contaminated environments was also examined.
2. Methods 2.1. Study areas 2.1.1. Uncontaminated locations Plantation Point: Jervis Bay is a large, sheltered bay on the south-east NSW coast approximately 180 km south of Sydney with a total catchment area of 400 km2. The area is predominantly sandstone [19]. Three main freshwater inflows drain into Jervis Bay. These waterways support mangrove and salt marsh communities, which filter the water inflows of silt and particulates [20,21]. Plantation Point is located directly opposite the entrance to Jervis Bay and is exposed to the greatest amount of wave action. Plantation Point is a flat, mudstone platform 200 m long and 100 m wide [22] and has a tertiary treated sewage outfall located at the eastern end of the platform. At Plantation Point, water temperature is a minimum of 15–161C in winter (August–September) and a maximum of 22–231C in summer (January) [23]. Salinity remains relatively constant throughout the year at around 35.5% with the exception of periods of heavy rainfall when freshwater input lowers salinity [23]. There is nearly a complete absence of heavy industrial activity in the catchment of Jervis Bay and residential developments are sewered. Jervis Bay is considered to be near pristine and a previous study of trace metals in gastropods has found no evidence of any trace metal contamination [22]. South Rosedale: Rosedale is located on the south-east coast of Australia approximately 250 km south of Sydney. The geology of the area is predominantly sedimentary (Bureau of Meteorology, Geology and Topography Maps). South Rosedale Beach has a rocky platform located at each end, with two smaller rocky
D. Gay, W. Maher / Water Research 37 (2003) 2173–2185
areas located within the beach. The average annual rainfall in the area is 931.9 mm with a winter average of 175.3 mm (Bureau of Meteorology). The rocky platform is reasonably exposed to weather changes as it is not located in an enclosed bay. Salinity is constant throughout the year at approximately 36%. Rosedale has no industrial activity and only residential developments with mostly off-site sewage disposal (pump outs). There are no storm water outlets within a kilometre of the sampling site. 2.1.2. Contaminated locations Caves Beach: Caves Beach is located at Swansea, approximately 100 km north of Sydney. Caves beach is situated near the mouth of Lake Macquarie, outside of the narrow entrance to the Lake. A domestic sewage outlet is located approximately 5 m off the beach. La Perouse—Botany Bay: Botany Bay is located in the Sydney region. The Bay is the site for numerous commercial operations and for recreational pursuits. Located within the bay are the Kingsford Smith Airport, Kurnell Oil Refinery, Chemical industries ICI Australia, Australian Paper Manufacturers Pty Ltd, commercial shellfish and fishing industries [24]. Recreational activities include fishing, boating and diving [24]. La Perouse is located on the northern shore of Botany Bay. Due to tidal circulation, high metal inputs have been found or indicated by oysters and sediment metal concentrations at this site [24]. Salinity at La Perouse remains high due to tidal flow from the South Pacific Ocean. Port Kembla: Port Kembla is in the Wollongong region, located 80 km south of Sydney. A sewage treatment plant located at Red Point, a rocky reef platform, discharges between 13.1 and 15.1 ML/day of primary treated effluent into approximately 8 m of water [25]. The primary industry located at Port Kembla is BHP Steel, with metal contaminants from the industry received via seawater and aerially from stacks [26]. 2.2. Sampling design Mass and gender analysis: Samples were collected in May 1997, at Plantation Point, Jervis Bay from the midlittoral zone, along a 30-m length of platform. Organisms were collected with shell widths (as determined across the largest whirl on the base of the organism) between 9 and 18 mm. Two hundred organisms were collected from which 50 females and 50 males were selected. Shoreline position: Samples were collected in March 1996 at Plantation Point, Jervis Bay along a 120 m transect at 20 m intervals, perpendicular to the shoreline (Site 1, 2, 3, 4 and 5 were 20, 40, 60, 80, 100 m, respectively from the low tide mark). Eighty samples were collected along a 50 m length of platform at each transect, from which 15 males and 15 females for each
2175
transect were selected. Organism shell widths collected were between 12 and 20 mm. Temporal variation: Samples were collected at Rosedale each month from June 1998 to May 1999 from the mid-littoral zone to assess any variation in metal concentrations over 12 months. Twenty-five gastropods were collected with shell widths (as determined across the largest whirl on the base of the organism) between 14 and 19 mm from which 20 organisms were selected for analysis. Response to contamination: Twenty-five gastropods were collected from the mid-littoral zone. Gastropods with shell widths of approximately 17 mm were collected to ensure organisms of similar mass/size and to allow enough tissue for analysis. 2.3. Sample preparation Samples were placed in seawater taken from the area, depurated for a minimum of 6 h and frozen until returned to the laboratory. After defrosting at room temperature, gastropod shells were cracked with a vice and removed from shells using stainless steel tweezers that had been rinsed with ethanol (Pronalys, AR grade, Selby Biolab). For gender analysis, the sex of the organism was determined by visual examination of the sex organs. The sexes were determined by the colouration of the gonads with females having pale grey or grey–green ovaries, the males having yellowish green, yellowish brown to creamy pink testis [27] and a penis located near the mantle [28]. Samples were rinsed with deionised water to remove residual shell from the soft tissue and placed in acid washed 20 mL polyethylene snap cap vials and frozen until digested. 2.4. Sample digestion and analysis Samples were freeze dried and digested as outlined in Baldwin et al. [29]. The individual dry samples were weighed (0.07 g) into acid washed 7 mL Teflon bombs to which 1 mL concentrated nitric acid (Aristar, BDH) was added. Twenty-two samples were digested in a 600 W microwave oven (MDS 81D, CEM, Indian Trail, NC, USA) using low volume microwave digestion. Samples were digested using the following program: 2 min at 600 W; 2 min at 0 W; 45 min at 450 W. Digested samples were transferred into individual, acid washed, 10 mL centrifuge tubes and diluted to 10 mL with deionised water. Samples for temporal study were again diluted 1/ 10 V/V in preparation for ICP-MS analysis. Mass, gender and shoreline position trace metal analyses were carried out using electrothermal atomic absorption spectroscopy (Perkin-Elmer 5100PC) with Zeeman background correction. A magnesium/ammonium dihydrogen phosphate matrix modifier was used for Cd analysis [30] with a palladium and magnesium
D. Gay, W. Maher / Water Research 37 (2003) 2173–2185
2176
matrix modifier used for Se analysis [31]. Copper and zinc analysis was carried out using flame atomic absorption spectroscopy with deuterium background correction. Temporal variation trace metal analyses (copper, zinc, cadmium and selenium) were carried out using the ICPMS [32]. Standard reference material NIST 1566a Oyster tissue was routinely run with sample digestions (n ¼ 20). Recoveries (in mg/g) for copper (6277), zinc (893784), cadmium (4.3570.45) and selenium (2.2470.15) were in general agreement with certified values (66.374.3; 830+57; 4.1570.38; 2.2170.24, respectively). Precision (RSD (%)) for measurement of copper, zinc, cadmium and zinc in Oyster tissue were 7%, 3%, 5% and 8% for a 20-mg sample and 5%, 3%, 5% and 6% or a 100-mg sample, respectively, indicating that precision did not vary greatly with the mass analysed. Analyses of replicated samples were within 8% for all trace metals.
3. Results Total metal concentrations represented as median concentrations, mean concentrations standard deviations and ranges are shown in Table 1. Medians are
reported, as they are a better representation of results as the data were not normally distributed. Selenium concentrations in samples below 0.02 g dry mass were below instrumental quantitation limits (0.05 mg/g). This dry mass corresponded to a shell width below 12 mm. These samples were assigned a selenium concentration of 0.01 mg/g for further statistical analysis. Frequency distributions on raw data of copper, zinc, cadmium and selenium were all positively skewed (Fig. 1). The Shapiro–Wilks test shows deviations from normality. 3.1. Effect of gender Five samples of shell width below 12 mm were unable to be sexed due to immaturity of sexual organs. These samples were omitted for gender analysis. ANCOVA using log transformed data showed no significant difference data between male and female B. nanum species for copper (F ¼ 0:00; DF=1,92; p ¼ 0:988), zinc (F ¼ 1:17; DF=1,91; p ¼ 0:283), cadmium (F ¼ 0:40; DF=1,90; p ¼ 0:527) and selenium (F ¼ 0:20; DF=1,83; p ¼ 0:658). This was confirmed with shoreline data showing no significant difference between males and females for Cu, Zn and Cd at transects which had sufficient numbers of males and females for comparison (transects 1 and 2)(Cu: F ¼ 0:46; DF=1;
Table 1 Metal concentrations in B. nanum from Sites on the NSW Coast, Australia Site
Copper JB CB BB PK
N
(mg/g) 96 22 25 25
Zinc (mg/g) JB 95 CB 22 BB 25 PK 25
Mean
SD
Median
Range
Skewness
Kurtosis
Normality
Dry mass (g)
po(W)
Mean
SD
152 37 94 379
90 30 55 152
130 26 83 336
23–473 5–143 15–296 196–667
1.11 2.32 2.16 0.61
1.37 6.80 7.12 1.00
0.0001 0.0001 0.0001 0.0001
0.06 0.20 0.11 0.13
0.04 0.05 0.03 0.02
67 63 127 110
20 27 87 32
65 59 95 93
12–164 36–152 63–452 67–180
1.26 1.81 2.66 0.96
5.14 4.43 7.97 0.15
0.0001 0.0001 0.0001 0.0001
0.06 0.20 0.11 0.13
0.04 0.05 0.03 0.02
Cadmium (mg/g) JB 94 CB 22 BB 25 PK 25
1.07 3.47 0.67 0.78
0.51 1.01 0.24 0.29
1.10 3.43 0.60 0.74
0.20–3.06 1.79–5.84 0.35–1.40 0.38–1.47
0.53 0.46 1.24 0.89
1.22 0.05 1.73 0.23
0.0001 0.0001 0.0001 0.0001
0.06 0.20 0.11 0.13
0.04 0.05 0.03 0.02
Selenium (mg/g) JB 81 CB 22 BB 25 PK 25
0.86 0.86 0.72 0.73
0.53 0.16 0.13 0.11
0.83 0.83 0.70 0.70
0.11–2.94 0.59–1.33 0.56–1.08 0.58–1.04
1.69 1.28 1.35 1.06
4.20 2.61 2.25 1.19
0.0001 0.0001 0.0001 0.0001
0.06 0.20 0.11 0.13
0.04 0.05 0.03 0.02
JB=Jervis Bay; CB=Caves Beach; BB=Botany Bay; PK=Port Kembla; N=Number of Organisms; SD=Standard Deviation; W=Shapiro–Wilks statistic for normality.
D. Gay, W. Maher / Water Research 37 (2003) 2173–2185
2177
3.2. Effect of mass As there were no significant differences between female and male B. nanum, all samples were combined for mass regression analysis. Regression analyses for copper, zinc, cadmium and selenium are shown in Fig. 2. No significant regressions were found for log transformed zinc (r2 ¼ 0:013; p ¼ 0:277), cadmium (r2 ¼ 0:015; p ¼ 0:243) and selenium (r2 ¼ 0:003; p ¼ 0:593) concentrations with mass for B. nanum. Regression analysis showed a significant relationship for log transformed copper concentrations and mass in B. nanum species (r2 ¼ 0:19; po0:001). This states that 19% of the variation in B. nanum copper concentrations can be explained by mass, with organisms of greater mass having higher copper concentrations. Analysis of shoreline data confirmed this result. Regression analysis found a significant positive correlation of dry mass with copper concentration (F ¼ 13:9; DF=1,103; po0:0005). This relationship only explained 12% of the variation in copper concentrations. No significant regressions of dry mass with zinc (F ¼ 0:09; DF=1,103; p ¼ 0:76), cadmium (F ¼ 0:28; DF=1,103; p ¼ 0:60) and selenium (F ¼ 0:19; DF=1,66; p ¼ 0:67) concentrations were found. There was the possibility that analysing males and females together was masking underlying differences in gender associated with the mass of the species. When males and females were analysed separately a significant regression of cadmium concentration and mass was found only for males (r2 ¼ 0:213; po0:001). Generally, non-significant relationships between metal concentrations and mass for B. nanum were found at contaminated sites (Table 2). 3.3. Effect of sample size Results for determination of sample size required to detect specific differences are presented in Fig. 3. Calculations used a t statistic of t½0:05; a ¼ 1:96 with N degrees of freedom. Results varied for each metal with around 10–15 organisms required to detect a 30% change in the mean. 3.4. Effect of shoreline position Fig. 1. Frequency distributions of Cu, Zn, Cd and Se (mg/g) for B. nanum, Jervis Bay, NSW. x-axis is ranges.
po0:51;F ¼ 0:76; DF=1; po0:39; Zn: F ¼ 1:88; DF=1; po0:18; F ¼ 3:32; DF=1; po0:08; Cd: F ¼ 0:06; DF=1; po0:81; F ¼ 2:89; DF=1; po0:1), with only a small but significant difference in selenium concentrations found at transect 2 (F ¼ 8:01; DF=1; po0:01) but not at transect 1 (F ¼ 0:1; DF=1; po0:75).
As no differences that would greatly influence metal concentrations (Cu, Zn, Cd and Se) were found between males and females and mass, data was pooled to determine differences in metal concentrations in B. nanum, between transects. Significant differences were found between transects for copper (F ¼ 16:18; DF=4,00; po0:0001), zinc (F ¼ 4:95; DF=4,100; po0:001) and cadmium (F ¼ 2:77; DF=4,100; po0:05) concentrations. No significant differences were found between transects for selenium (F ¼ 1:43; DF=2,65; p ¼ 0:25) concentrations. Ranking from
D. Gay, W. Maher / Water Research 37 (2003) 2173–2185
2178
Copper
150 Sample Size
3.00 log (Cu µg/g)
2.50 2.00 1.50 1.00
Copper 135
100 50
34
0 10%
0.50 0.05
0.15
0.2
6 50%
Zinc 40
Zinc
2.50 log (Zn µg/g)
0.1 Dry mass (g)
Sample size
0
2.00 1.50
35
30 20 10
9 4
0 10%
1.00
20%
30%
2 50%
Percent change in mean
0.50 0
0.05
0.1
0.15
0.2
Dry mass (g) Cadmium 0.70 0.60 0.50 0.40 0.30 0.20 0.10 0.00
Sample size
0.00
100 80
Cadmium 88
60 40 22
20 0 10%
0.1
0.15
0.2
Dry mass (g)
Sample size
0.05
150
20% 30% Percent change in mean
4 50%
146
100 50
37
0 10%
Selenium
10
Selenium
200
0
20%
17 30%
6 50%
Percent change in mean
0.70 log (Se µg/g)
30%
Percent change in mean
0.00
log (Cd µg/g)
20%
15
0.60 0.50 0.40 0.30 0.20 0.10 0.00
Fig. 3. Sample sizes required to detect changes in mean trace metal concentrations for Cu, Zn, Cd and Se in Bembicium nanum from Jervis Bay, NSW. Note vertical scales are not equal.
0
0.05
0.1 Dry mass (g)
0.15
0.2
Fig. 2. Regression analysis for Cu, Zn, Cd and Se concentrations and mass for B. nanum.
highest metal concentrations to lowest by Tukey’s HSD test are (in mg/g): copper: 4=5=3>2=1; Zinc: 3=4X2X1=5, and cadmium: 3X4=5=1X2. Copper
concentrations were significantly higher at transects 3, 4 and 5 compared to the lower transects 1 and 2 (Fig. 4). Although Zinc concentrations between transects were found to be statistically significant visual examination of box whisker plots (Fig. 4) and ranking by Tukey’s HSD test, show no clear separation between transects. Cadmium concentrations were similar at all transects (Fig. 4), supported by the low significance level of the ANOVA. Data on selenium concentrations for sites 3 and 5 are missing due to experimental problems. Given the limited data for selenium at each transect, it is difficult to
D. Gay, W. Maher / Water Research 37 (2003) 2173–2185
2179
Table 2 Regression analysis for Cu, Zn, Cd and Se concentrations and mass for B. nanum at three contaminated NSW sites Copper 2
Caves Beach Botany Bay Port Kembla
Zinc 2
Cadmium 2
Selenium
r
p
r
p
r
p
r2
p
0.039 0.010 0.006
0.377 0.639 0.722
0.195 0.206 0.002
0.040* 0.023* 0.848
0.000 0.156 0.012
0.976 0.051 0.602
0.027 0.097 0.470
0.468 0.131 0.0002*
*Denotes significant relationships.
determine whether the non-significant result is due to no difference in selenium concentrations between transects 1, 2 and 4, or a reflection of a smaller sample size. 3.5. Temporal variation Significant between month differences for copper (F ¼ 5:05; DF=11,216; po0:0001) and cadmium (F ¼ 6:44; DF=11,215; po0:0001) concentrations in B. nanum were found using log transformed data. Untransformed data showed significant differences between months for zinc (F ¼ 6:32; DF=11,217; po0:0001) and selenium (F ¼ 9:05; DF=11,217; po0:0001) concentrations. Mean dry mass of selected B. nanum was constant across the 12-month period (Fig. 5). All months exhibited skewed distributions for metal concentrations, for example metal concentration frequency distributions of Cu, Zn, Cd and Se for May 1999 (Fig. 6). (i) Copper: Copper concentrations fluctuated throughout the 12-month period with peaks in September and March and lows in June, July and December. Highest mean concentration of copper was found in March. High copper concentrations in a few organisms from January to May greatly influenced mean concentrations for each of these months. Comparison of mean copper concentrations to median copper concentrations supports this. Copper body burdens followed a similar trend to copper concentrations (Fig. 5). (ii) Zinc: Zinc concentrations were found to be fairly similar for the 12-month period. Peaks were recorded for October and February, with lows in June, July, August and November, although clear separation of sites for zinc concentrations could not be determined using Tukey’s HSD test. Zinc body burdens followed zinc concentrations except for the months June, August and November, which showed higher zinc body burden. These months also had higher mean mass compared to other months. (iii) Cadmium: Cadmium concentrations fluctuated throughout the 12-month period, with peak con-
centrations in September, January and March, and low concentrations in August, November, December and February (Fig. 5). Again, greater variability in cadmium concentrations in the months from January to May influenced mean concentrations for these months. Cadmium body burdens closely followed cadmium concentrations over the 12 months (Fig. 5). (iv) Selenium: Selenium concentrations in B. nanum were lowest in June and July, increasing to an October peak. Concentrations then remained fairly constant for the rest of the 12-month period. The range of concentrations for each month also remained similar. Selenium body burdens followed selenium concentrations (Fig. 5). . 3.6. Accumulation of trace metals in contaminated environments B. nanum had increased trace metal concentrations at Port Kembla, Botany Bay and Caves Beach (Table 1), locations known to be contaminated with copper, zinc and cadmium [24–26]. Frequency distributions of these trace metals are also skewed to higher concentrations at these sites (Fig. 7). The higher concentrations measured in B. nanum at the known contaminated sites show that B. nanum accumulate Cu, Zn and Cd in response to exposure to these metals. B. nanum were also abundant at these sites indicating they are hardy and tolerate high concentrations of trace metals.
4. Discussion 4.1. Factors contributing to natural variation in trace metal concentrations 4.1.1. Gender No significant difference of copper, zinc, cadmium and selenium were found between male and female B. nanum. Of the few studies that have looked at the effect of gender on trace metal differences in gastropods, gender has not been found to significantly affect metal
2180
D. Gay, W. Maher / Water Research 37 (2003) 2173–2185
Fig. 4. Concentrations of Cu, Zn, Cd and Se (mg/g) in B. nanum at each transect. Box whisker plots represent line: median; box: 25th– 75th percentile; whiskers: 5th–95th percentile.
concentrations in whole tissue analysis (L. littorea, [33] Bembicium auratum, Austrocochlea constricta, [34]). 4.1.2. Mass Copper, zinc, cadmium and selenium concentrations in B. nanum were generally independent of mass
(Table 2, Fig. 2). Although copper concentrations were significantly correlated with mass, it accounted for only 19% of the variation. The small amounts of variability attributable to mass are not significant compared to overall inherent variability (see next section).
14 12 10 8 6 4 2 0
Copper
180 160 140 120 100 80 60 40 20 0
2181
[Cu] (µg/g)
Dry Mass (x100) (g) and Total Cu (µg)
D. Gay, W. Maher / Water Research 37 (2003) 2173–2185
10
Zinc
8 6 4 2 0
140 120 100 80 60 40 20 0
[Zn] (µg/g)
Dry Mass (x100) (g) and Total Zn (µg)
Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Month
Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Month 7 6 5 4 3 2 1 0
[Cd] (µg/g)
1.6 1.4 1.2 1 0.8 0.6 0.4 0.2 0
Dry Mass (g) and [Se] (µg/g)
Dry Mass (g) and Total Cd (µg)
Cadmium 0.4 0.35 0.3 0.25 0.2 0.15 0.1 0.05 0 Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Month
Total Se (µg)
Selenium 0.14 0.12 0.1 0.08 0.06 0.04 0.02 0 Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Month
Fig. 5. Temporal variation of metal concentration, metal body burden and dry mass of Cu, Zn, Cd and Se in B. nanum. Each point is n ¼ 20; error bars are omitted for clarity but are typically x–y%. Legend: K Dry mass (g); ’ Metal Concentration (mg/g); m Metal Body Burden (mg).
In a number of gastropod species, metal concentrations have been found to be statistically independent of mass (Cu, Cd: Littorea. littorea, Cu, Zn, Cd: Scaphander lignarius, Credipula fornicata; [35] Se: B. nanum, A. constricta, Nerita atramentosa, [22] Cd: H. tuberculata,[36]) but often show an increase/decrease with mass/ size [7,37]). Boyden [35] has stated that it is difficult to determine mass influences on metal concentrations by regression analysis unless a species exhibits a tenfold increase in size range, due to the inherent variability shown by biological organisms. As B. nanum naturally exhibits a small mass range (0.01–0.165 g in this case) and organisms that can be easily analysed are in the range 0.05–0.165 g, significance relationships of metal concentration and mass are not expected. However, as
Fig. 6. Metal concentration frequency distributions for Cu, Zn, Cd and Se, Rosedale, NSW (May 1999). x-axis is ranges.
trace metals often show a trend with mass, similar mass organisms should be selected for analysis when comparing different locations. 4.1.3. Inherent variability All metal distributions in B. nanum were not normally distributed (Fig. 1). Copper, zinc, cadmium and selenium had positively skewed distributions. Most
D. Gay, W. Maher / Water Research 37 (2003) 2173–2185
Port Kembla
[Zn] µg/g
More
550
450
2.75
2.25
1.75
0.75
18 16 14 12 10 8 6 4 2 0 0.25
No. of Individuals More
180
140
100
60
350
Caves Beach
16 14 12 10 8 6 4 2 0 20
250
50
[Cu] µg/g
Botany Bay No. of individuals
150
No. of Individuals [Zn] µg/g
18 16 14 12 10 8 6 4 2 0
More
180
140
100
60
20
No. of Individuals
Port Kembla 16 14 12 10 8 6 4 2 0
1.25
2182
[Cd] µg/g
Fig. 7. Metal concentration frequency distributions at Port Kembla (Cu, Zn), Botany Bay (Zn) and Caves Beach (Cd). x-axis is ranges.
individuals had lower metal concentrations with only a small number showing higher concentrations, extending the tail of the distribution (Fig. 1). Skewness could not be explained by variations in mass or gender as the metals were mostly not correlated with mass and no significant differences were found between males and females in B. nanum. Trace element concentrations in marine molluscs have been found to have positively skewed distributions in non-contaminated environments [38,39,33]. Non-normal distributions have been found for zinc in M. edulis, for copper in L. littorea [33] and for selenium in B. nanum [22]. Non-normal distributions are generally attributed to inherent variability due to individual variations in uptake and excretion of trace elements by organisms (Lobel et al. [33]), caused by physiological and metabolic differences between organisms. Metal accumulation in B. nanum will be affected by microhabitat differences on rocky platforms due to rock pools, crevices and over hangs which cause differences in shelter (stress), food availability, food types, feeding periods and metabolism [40–42]. All these factors will contribute to trace metal concentration variability within gastropod samples. 4.1.4. Shoreline position Shoreline position did not greatly influence cadmium and selenium concentrations in B. nanum (Fig. 4).
Copper and zinc concentrations were influenced by shoreline position with gastropods collected from transects further away from the low tide mark having increased trace metal concentrations. B. nanum are grazers feeding on microscopic algae and bacteria [43]. Trends in copper and zinc concentrations could be explained by variations in food type, food availability and immersion that affect growth rates and trace metal intake. The observations for copper and zinc are similar to the trend found for zinc concentrations in the barnacle Elminius modestus [44]. Longer immersion times in E. modestus were found to result in increased growth due to longer feeding times and corresponding lower zinc concentrations. B. nanum only feed whilst immersed [43]. Gastropods located higher on the shore would experience shorter immersion times, resulting in greater desiccation and decreased growth altering metabolic rates [45] thus influencing metal concentrations. It is expected that variability exists in food type and amounts at different heights on the shore. Underwood and Atkinson [42] observed less splashing of seawater at Jervis Bay higher on shore due to lower waves than would be experienced at more exposed rocky platforms. This would produce drier conditions and alter algal cover higher on shore and thus affect food availability. Underwood [43] found that growth rates in B. nanum were only affected at the highest position on shore,
D. Gay, W. Maher / Water Research 37 (2003) 2173–2185
explained by a decrease in algal cover and thus a decrease in food supply. Feeding rates of individuals of different sizes of B. nanum have been found not to be significantly different or influenced by shore position [43,46]. 4.1.5. Temporal changes Copper, zinc, cadmium and selenium concentrations were found to vary temporally in B. nanum, yet no clear seasonal trends were evident (Fig. 5). Temporal variation in metal concentrations in molluscs has been found in previous studies [47–49]. These studies found seasonal variation in metal concentrations associated with changes in mass of the organism, with maximums in winter and early spring and minimums in late spring and summer. Temporal variation associated with changes in the mass of B. nanum is not likely as the mean mass of the gastropods sampled over the 12-month period were relatively constant (Fig. 5). A number of factors have been identified to influence temporal variation in metal concentrations. These are variation in pollution delivery to the environment, changes in salinity and water temperature, and an organism’s physiology, associated with reproductive cycles [4]. Salinity is constant in the areas studied so is not a factor influencing temporal variability of trace metal concentrations. No known point sources of metal contamination exist in the Rosedale area. Sewage at Rosedale, a possible source of trace metals, is mostly taken off-site for disposal. Median concentrations of zinc, cadmium and selenium were higher in B. nanum at Rosedale compared to metal concentrations found in B. nanum at Jervis Bay. Rosedale receives storm water from urban areas during flood events by overland transport that may result in higher trace metal concentrations in gastropods at this location than at Jervis Bay and cause some of the temporal variation of trace metal concentrations. Water temperature during this study ranged from 231C in summer to 131C in winter and could be influencing trace metal concentrations in B. nanum. Laboratory experiments have shown that temperature changes of this magnitude can result in the increase or decrease of trace metal concentrations because of changes in metabolic and excretion rates [50,47,51]. Changes in temperature can also trigger spawning in molluscs [52], which can change metal concentrations in molluscs through the loss of mature oocytes. B. nanum have been found to spawn during summer with a peak in spawning in January and finishing by February/March [53,54]. If metals were to be lost with mature oocytes, it is expected that decreases in mass and total body metal burden would occur. No significant mass losses occurred during summer, possibly due to ample food supply during summer thus organisms maintained body mass.
2183
Taylor [34] found that mature oocytes during spawning in a similar gastropod B. auratum were released slowly over an extended period, which could also explain the absence of mass losses in summer due to spawning activity. It is likely that reproductive cycles do not significantly influence trace metal body burdens in B. nanum. 4.2. Use of B. nanum as a biomonitor B. nanum meets the requirements of a suitable biomonitor species as summarised by Phillips [3–5] as it is * *
*
*
Sedentary and thus representative of a location. Hardy and tolerate high levels of pollutants as indicated by their high abundance at contaminated sites. Abundant on all rock platforms, easy to identify and provide enough tissue (0.05–0.1 g dry mass) for analysis. Accumulate trace metals in contaminated environments, although we have not established a simple relationship between contamination and bioavailable trace metal concentration.
In addition, gender has been found not to significantly influence trace metal concentrations of copper, zinc, cadmium and selenium in B. nanum. This study has shown that trace metal concentrations can vary with gastropod mass, shoreline position and temporally. Thus care must be taken to collect individual gastropods of similar mass from similar shoreline positions and times. Sampling in the midlittoral zone will minimise influences due to shore position. The large natural variability in metal concentrations of copper, zinc, cadmium and selenium in B. nanum populations would make it unsuitable to monitor subtle changes in trace metal contamination of a given area but this species can be used to indicate areas of gross contamination. Sample size calculations based on the natural variability of trace metals in B. nanum (Fig. 3) indicate that around 10–15 organisms are enough to detect a 30% change in the mean of a sample for B. nanum. Sample size estimates for future analyses would be dependent on the percentage change in the mean required to meet the purpose of the study. To detect very low differences in means, i.e. less than 10%, would require large sample sizes (greater than 80 organisms) due to the inherent variability exhibited by this species. For financial reasons, sample sizes of this magnitude are rarely used for routine monitoring. More studies are required to demonstrate the nature of the response of B. nanum to metals in contaminated environments. However, generally the use of a biomo-
2184
D. Gay, W. Maher / Water Research 37 (2003) 2173–2185
nitor is to show that contamination exists in an area, not how much contamination there is.
References [1] Allan RJ, Campbell PGC, Forstner U, Lum K. Fate and effects of toxic elements in large rivers and their estuaries. Sci Total Environ 1990;97/98:1–872. [2] SOE. Australia: State of the Environment, Australia, Collingwood: CSIRO Publishing, 1996. [3] Phillips DJH. The use of biological indicator organisms to monitor trace metal pollution in marine and estuarine environments—a review. Environ Pollut 1977;13:281–317. [4] Phillips DJH. Quantitative aquatic biological indicators. London: Applied Science Publishers, 1980. [5] Phillips DJH. Use of macroalgae and invertebrates as monitors of metal levels in estuaries and coastal waters. In: Furness RW, Rainbow PS, editors. Heavy metals in the marine environment. Boca Raton, FL: CRC. Press Inc, 1990. p. 81–99. [6] Phillips DJH, Rainbow PS. The biomonitoring of trace metals and radionuclides. In: Phillips DJH, editor. Biomonitoring of trace aquatic contaminants. Amsterdam: Elsevier Science Publishers Ltd., 1993. p. 79–132. [7] Langston WJ, Bebianno MJ, Burt GR. Metal handling strategies in molluscs. In: Langstone WJ, Bebianno MJ, editors. Metal metabolism in aquatic environments. London: Chapman & Hall, 1998. p. 219–85 [Chapter 8]. [8] Nicolaidou A, Nott JA. Mediterranean pollution from a ferro-nickel smelter. Mar Pollut Bull 1990;21:137–43. [9] Nicolaidou A, Nott JA. Metals in sediment, seagrass and gastropods near a nickel smelter in Greece. Possible interactions. Mar Pollut Bull 1998;36:360–5. [10] Abdallah AT, Moustafa MA. Accumulation of lead and cadmium in the marine prosobranch Neriti saxtilis, chemical analysis, light and electron microscopy. Environ Pollut 2002;116:185–91. [11] Langston WJ, Spence SK. Biological factors involved in metal concentrations observed in aquatic organisms. In: Tessier A, Turner DR, editors. Metal speciation and bioavailability. New York: Wiley, 1995. p. 407–78. [12] Brown MT, Depledge MH. Determinants of trace metal concentrations in marine organisms. In: Langstone WJ, Bebianno MJ, editors. Metal metabolism in aquatic environments. London: Chapman & Hall, 1998. p. 185–217 [Chapter 7]. [13] Ireland MP, Wootton RJ. Distribution of lead, zinc, copper and manganese in the marine gastropods, Thais Lapillus and Littorina littorea, around the coast of Wales. Environ Pollut 1977;12:27–41. [14] Kang SG, Wright DA, Koh CH. Baseline metal concentration in the Asian periwinkle Littorina brevicula employed as a biomonitor to assess metal pollution in Korean coastal water. Sci Total Environ 2000;263:143–53. [15] Peerzada N, Eastbrook C, Guinea M. Heavy metal concentration in Telescopium from Darwin Harbour, N.T, Australia. Mar Pollut Bull 1990;21:307–8. [16] Blackmore G. Interspecific variation in heavy metal body concentrations in Hong Kong marine invertebrates. Environ Pollut 2001;114:303–11.
[17] Campanella L, Conti ME, Cubadda F, Sucapane C. Trace metals in seagrass, algae and molluscs from an uncontaminated area in the Mediterranean. Environ Pollut 2001;111:117–26. [18] Gorden M, Knauer GA, Martin JH. Mytilus californius as a bioindicator of trace metal pollution: variability and statistical considerations. Mar Pollut Bull 1980;10: 195–8. [19] Cho G. The Jervis Bay environment. In: Cho G, Georges A, Stoutjesdijk R, editors. Jervis Bay, A place of cultural, scientific and educational value. Canberra: ANPWS, 1995. p. 3–8. [20] Jones N, Abell R, Jacobson G. Hydrology. In: Cho G, Georges A, Stoutjesdijk R, editors. Jervis Bay a place of cultural scientific and educational values. Canberra: ANPSW, 1995. p. 53–64. [21] Taylor G, Abell R, Paterson I. Geology, geomorphology, soils and earth resources. In: Cho G, Georges A, Stoutjesdijk R, editors. Jervis Bay. A place of cultural, scientific and educational value. Canberra: ANPWS, 1995. p. 41–52. [22] Baldwin S, Maher W, Kleber E, Krikowa F. Selenium in marine organisms of seagrass habitats (Posidonia australis) of Jervis Bay, Australia. Mar Pollut Bull 1996; 32(3):310–6. [23] Holloway PE. Water circulation. In: Cho G, Georges A, Stoutjesdijk R, editors. Jervis Bay. A place of cultural, scientific and educational value. Canberra: ANPWS, 1995. p. 65–70. [24] Hayes WJ, Anderson IJ, Gaffoor MZ, Hurtado J. Trace metals in oysters and sediments of Botany Bay, Sydney. Sci Total Environ 1998;212:39–47. [25] Hall SJ, Miller LB, Gidding LA. Hard substrate benthic communities. In: Marine investigation for the Bellambi and Port Kembla STP Upgradings Annual Report. Environment Management Unit, Water Board, Sydney, 1991. [26] Moran PJ. Water quality control and its effects on the concentration of heavy metals in Port Kembla Harbour, NSW. Mar Pollut Bull 1984;15:294–7. [27] Anderson H. The gastropod genus Bembicium Phillipe. Aust J Mar Freshwater Resources 1958;9:547–73. [28] Galstoff PS. Physiology of reproduction in molluscs. Am Zool 1961;1:273–89. [29] Baldwin S, Deaker M, Maher W. Low-volume microwave digestion of marine biological tissues for the measurement of trace elements. Analyst 1994;119:1701–4. [30] Slavin W. Graphite Furnace AAS. A Source Book. United States of America: The Perkin Elmer Corporation, 1994. p. 83. [31] Deaker M, Maher W. Determination of selenium in seleno-compounds and marine biological tissues using stabilised temperature platform furnace atomic absorption spectrometry. J Anal At Spectrosc 1995;10:423–34. [32] Maher W, Forster S, Krikowa F, Snitch P, Chapple G, Craig P. Measurement of trace elements and phosphorus in marine animal and plant tissues by low-volume microwave digestion and ICP-MS. At Spectrosc 2001; 22(5):361–70. [33] Lobel PB, Mogie P, Wright DA, Wu BL. Metal accumulation in four molluscs. Mar Pollut Bull 1982;13:170–4.
D. Gay, W. Maher / Water Research 37 (2003) 2173–2185 [34] Taylor A. Zinc, copper and cadmium accumulation, detoxification and storage in the gastropod molluscs Austrocochlea constricta and Bembicium auratum and an assessment of their potential as biomonitors of trace metal pollution in estuarine environments. Thesis for the Degree of Master of Applied Science, University of Canberra, ACT, 1998. [35] Boyden CR. Effect of size upon metal content of shellfish. J Mar Biol Assoc UK 1977;57:675–714. [36] Bryan GW, Potts GW, Forster GR. Heavy metals in the gastropod mollusc Haliotis tuberculata (L). J Mar Biol Assoc UK 1977;57:379–90. [37] Leung KMY, Furness RW. Effects of animal size on concentrations of metallothionein and metals in periwinkles Littorina littorea collected from the Firth of Clyde, Scotland. Mar Pollut Bull 1999;39:126–36. [38] Lobel PB, Belkhode SP, Bajdik C, Jackson SE, Longerich HP. General characteristics of the frequency distributions of element concentrations and of interelemental correlations in aquatic organisms. Mar Environ Res 1992;33: 111–26. [39] Lobel PB, Bojdik CD, Belkhode SP, Jackson SE, Longerich HP. Improved protocol for collecting mussel watch specimens taking into account sex, size, condition, shell shape and chronological age. Arch Environ Contamin Toxicol 1991;21:409–14. [40] Fairweather PG, Underwood AJ, Moran MJ. Preliminary investigation of predation by the whelk Morula marginalba. Mar Ecol Prog Ser 1984;17:143–56. [41] Fletcher WJ. Interactions among subtidal Australian sea urchins, gastropods, and algae: effects of experimental removals. Ecol Monogr 1987;57:89–109. [42] Underwood AJ, Atkinson MH. In: Cho G, Georges A, Stoutjesdijk R, editors. Jervis Bay. A place of cultural, scientific and educational value. Canberra: ANPWS, 1995. p. 123–32. [43] Underwood A. Microalgal food and the growth of the intertidal gastropods Nerita atramentosa Reeve and Bembicium nanum (Lamarck) at four heights on a shore. J Exp Mar Biol Ecol 1984;79:277–91. [44] Al-Thaqafi K, White KN. Effect of shore position and environmental levels on body metal burdens in the
[45]
[46]
[47]
[48]
[49]
[50]
[51]
[52]
[53]
[54]
2185
barnacle, Elminius modestus. Environ Pollut 1991;69: 89–104. Barnes H, Barnes M, Finlayson DM. The metabolism during starvation of Balanus balanoides. J Mar Biol Assoc UK 1963;43:213–23. Petraitis PS. Effects of body size and water temperature on grazing rates of four intertidal gastropods. Aust J Ecol 1992;17:409–14. Bryan GW. Some aspects of heavy metal tolerance in aquatic organisms. In: Lockwood APM, editor. Effect of pollutants on aquatic organisms, vol. 2. Society for Experimental Biology Seminar Series. Cambridge University Press, 1976. p. 7–34. Phillips DJH. The common mussel Mytilus edulis as an indicator of pollution by zinc, cadmium, lead and copper. I. Effects of environmental variables on uptake of metals. Mar Biol 1976;38:59–69. Simpson CD. Uptake and loss of zinc and lead by mussels (Mytilus edulis) and relationships with body weight and reproductive cycle. Mar Pollut Bull 1979;10: 74–8. Fowler SW, Benayoun G. Influence of environmental factors on selenium flux in two marine invertebrates. Mar Biol 1976;37:59–68. Orren MJ, Eagle GA, Hennig HF-KO, Green A. Variations in trace metal content of the mussel Chromytilus meridionalis (Kr.) with season and sex. Mar Pollut Bull 1980;11:253–7. Hadfield AJ, Anderson DT. Reproductive cycles of the bivalve molluscs Anadara trapezia (Deshayes), Venerupis crenata Lamarck and Anomia descripta Inedale in the Sydney region. Aust J Mar Freshwater Res 1988;39: 649–60. Anderson DT. The reproduction and early life history of the gastropod B. nanum (Lamarck 1822) (Fam. LITTORINIDAE). The Proceedings of the Linnean Society of NSW, Sydney, vol. 1xxxvi(part 2) 1961. p. 203–6. Underwood AJ. The reproductive cycles and geographical distribution of some common Eastern Australian prosobranchs (Mollusca: gastropoda). Aust J Mar Freshwater Res 1974;25:63–88.
Natural variation of copper, zinc, cadmium and selenium concentrations in Bembicium nanum and their potential use as a biomonitor of trace metals D. Gay*, W. Maher Ecochemistry Laboratory, University of Canberra, University Drive, Canberra ACT 2601, Australia Received 17 May 2002; accepted 22 November 2002
Abstract Copper, zinc, cadmium and selenium were measured in the gastropod mollusc Bembicium nanum at two uncontaminated locations, Jervis Bay and Rosedale, NSW, to determine natural variability of metals associated with gender, mass, shore position and temporal variability. Trace metals were also measured in B. nanum at three industrialised locations to determine the accumulation of trace metals in contaminated environments. Copper, zinc, cadmium and selenium concentrations were not significantly different between male and female B. nanum. No significant relationships were found between zinc, cadmium and selenium concentrations and mass. There was a significant relationship between copper concentration and mass but only 19% of the variation was explained by mass. Generally inherent variability within samples had a greater influence than gender or variations in mass on trace metal concentrations. No trend was found in cadmium and selenium concentrations with variation in shoreline position. Copper and zinc concentrations increased further away from the low tide mark, with a decrease in metal concentrations at the furthest site from the water. Variability in metal concentrations is attributable to variations in food source, food availability and different immersion times. Copper, zinc, cadmium and selenium concentrations varied over a 12-month period. Copper, cadmium and selenium were taken up and lost over time, as metal body burden followed the same trend as metal concentrations. Zinc concentrations were influenced by mass. Copper and cadmium concentrations fluctuated throughout the 12-month period but with no clear seasonal trends. Selenium concentrations peaked in spring (October), with concentrations remaining uniform over the other months. These differences in mean concentrations between months were most likely due to inherent trace metal variability associated with differences in food availability and changes in metabolic rates associated with changes in temperature during the study period. Measurement of trace metals in B. nanum at contaminated sites showed that B. nanum accumulates metals in response to contamination. B. nanum meets most of the requirements to be a biomonitor of trace metal contamination as they are abundant, sedentary, easy to identify, provide sufficient tissue for analysis, tolerate high concentrations of pollutants and they accumulate trace metals in response to contamination. However, as trace metal concentration can vary with mass, shoreline position and temporally, care must be taken to collect individual organisms with similar mass from similar shoreline positions and times. r 2003 Elsevier Science Ltd. All rights reserved. Keywords: Trace metals; Biomonitors; Bembicium nanum; Natural variation; Accumulation
*Corresponding author. 0043-1354/03/$ - see front matter r 2003 Elsevier Science Ltd. All rights reserved. doi:10.1016/S0043-1354(02)00622-X
2174
D. Gay, W. Maher / Water Research 37 (2003) 2173–2185
1. Introduction Many of the world’s population reside along coastal areas. With human settlement comes the building of infrastructure, land reclamation for port and industrial development, habitat modification, tourism and recreational activities. Industry situated along the coast has better access to shipping ports and easier disposal of industrial waste, which may be released into marine waters. Furthermore, domestic sewage effluent is pumped from outfalls located along the near shore regions. Trace metal contamination has been identified as a concern in coastal environments, due to discharges from smelters and mining operations and from storm water and sewage disposal [1,2]. As trace metals can have a deleterious effect on aquatic ecosystems, it is important to know where bioavailable metals are entering aquatic systems and take remedial management action. Organisms that concentrate metals in their tissues are often used to indicate and quantify contaminant levels or bioavailability of contaminants in the environment [3–7]. Generally bivalves such as mussels are used for this purpose. To be used as biomonitors organisms should be: *
*
*
*
Sessile or sedentary thus being representative of the study area. Hardy, tolerating high concentrations of pollutants and large ranges in salinity, and permitting laboratory studies. Abundant in study areas, easy to identify and sample and should provide sufficient tissue for analysis of contaminants of interest. Have a simple correlation between the contaminant concentration found in their tissues and the average bioavailable pollutant concentration. This correlation should be the same for all study sites.
Many species of marine molluscs have been shown to reflect ambient metal concentrations (e.g. [8–10]). Yet many physiological and environmental factors can combine to influence the overall metal concentrations found in marine molluscs. Intrinsic factors such as mass and size, gender, reproductive state, accumulation/ regulation strategies, and diet and extrinsic factors such as temperature, salinity, supply of metals, food availability and metal–metal relationships can influence trace metal concentrations [6,7,11,12]. Along the Australian coast, rocky intertidal platforms are common. These platforms serve as habitat for gastropods and other species. A widely distributed and common gastropod present in this intertidal region is B. nanum. Few studies worldwide have used gastropods in monitoring studies [13–17,9,10]. B. nanum meets many of the requirements of a suitable biomonitor. However, before B. nanum can be used in monitoring studies, an
estimate of natural trace metal variability due to intrinsic and extrinsic factors need to be assessed [18]. The aim of this study was to examine the factors that are contributing to the natural variability in trace metal concentrations in B. nanum at non-contaminated sites by the determination of: *
*
*
Variation in metal concentrations attributable to gender and changes in mass. Variation in metal concentrations with shoreline position. Temporal metal concentration variation and the relationship with total metal body burden and mass.
As well, the accumulation of trace metals by B. nanum in contaminated environments was also examined.
2. Methods 2.1. Study areas 2.1.1. Uncontaminated locations Plantation Point: Jervis Bay is a large, sheltered bay on the south-east NSW coast approximately 180 km south of Sydney with a total catchment area of 400 km2. The area is predominantly sandstone [19]. Three main freshwater inflows drain into Jervis Bay. These waterways support mangrove and salt marsh communities, which filter the water inflows of silt and particulates [20,21]. Plantation Point is located directly opposite the entrance to Jervis Bay and is exposed to the greatest amount of wave action. Plantation Point is a flat, mudstone platform 200 m long and 100 m wide [22] and has a tertiary treated sewage outfall located at the eastern end of the platform. At Plantation Point, water temperature is a minimum of 15–161C in winter (August–September) and a maximum of 22–231C in summer (January) [23]. Salinity remains relatively constant throughout the year at around 35.5% with the exception of periods of heavy rainfall when freshwater input lowers salinity [23]. There is nearly a complete absence of heavy industrial activity in the catchment of Jervis Bay and residential developments are sewered. Jervis Bay is considered to be near pristine and a previous study of trace metals in gastropods has found no evidence of any trace metal contamination [22]. South Rosedale: Rosedale is located on the south-east coast of Australia approximately 250 km south of Sydney. The geology of the area is predominantly sedimentary (Bureau of Meteorology, Geology and Topography Maps). South Rosedale Beach has a rocky platform located at each end, with two smaller rocky
D. Gay, W. Maher / Water Research 37 (2003) 2173–2185
areas located within the beach. The average annual rainfall in the area is 931.9 mm with a winter average of 175.3 mm (Bureau of Meteorology). The rocky platform is reasonably exposed to weather changes as it is not located in an enclosed bay. Salinity is constant throughout the year at approximately 36%. Rosedale has no industrial activity and only residential developments with mostly off-site sewage disposal (pump outs). There are no storm water outlets within a kilometre of the sampling site. 2.1.2. Contaminated locations Caves Beach: Caves Beach is located at Swansea, approximately 100 km north of Sydney. Caves beach is situated near the mouth of Lake Macquarie, outside of the narrow entrance to the Lake. A domestic sewage outlet is located approximately 5 m off the beach. La Perouse—Botany Bay: Botany Bay is located in the Sydney region. The Bay is the site for numerous commercial operations and for recreational pursuits. Located within the bay are the Kingsford Smith Airport, Kurnell Oil Refinery, Chemical industries ICI Australia, Australian Paper Manufacturers Pty Ltd, commercial shellfish and fishing industries [24]. Recreational activities include fishing, boating and diving [24]. La Perouse is located on the northern shore of Botany Bay. Due to tidal circulation, high metal inputs have been found or indicated by oysters and sediment metal concentrations at this site [24]. Salinity at La Perouse remains high due to tidal flow from the South Pacific Ocean. Port Kembla: Port Kembla is in the Wollongong region, located 80 km south of Sydney. A sewage treatment plant located at Red Point, a rocky reef platform, discharges between 13.1 and 15.1 ML/day of primary treated effluent into approximately 8 m of water [25]. The primary industry located at Port Kembla is BHP Steel, with metal contaminants from the industry received via seawater and aerially from stacks [26]. 2.2. Sampling design Mass and gender analysis: Samples were collected in May 1997, at Plantation Point, Jervis Bay from the midlittoral zone, along a 30-m length of platform. Organisms were collected with shell widths (as determined across the largest whirl on the base of the organism) between 9 and 18 mm. Two hundred organisms were collected from which 50 females and 50 males were selected. Shoreline position: Samples were collected in March 1996 at Plantation Point, Jervis Bay along a 120 m transect at 20 m intervals, perpendicular to the shoreline (Site 1, 2, 3, 4 and 5 were 20, 40, 60, 80, 100 m, respectively from the low tide mark). Eighty samples were collected along a 50 m length of platform at each transect, from which 15 males and 15 females for each
2175
transect were selected. Organism shell widths collected were between 12 and 20 mm. Temporal variation: Samples were collected at Rosedale each month from June 1998 to May 1999 from the mid-littoral zone to assess any variation in metal concentrations over 12 months. Twenty-five gastropods were collected with shell widths (as determined across the largest whirl on the base of the organism) between 14 and 19 mm from which 20 organisms were selected for analysis. Response to contamination: Twenty-five gastropods were collected from the mid-littoral zone. Gastropods with shell widths of approximately 17 mm were collected to ensure organisms of similar mass/size and to allow enough tissue for analysis. 2.3. Sample preparation Samples were placed in seawater taken from the area, depurated for a minimum of 6 h and frozen until returned to the laboratory. After defrosting at room temperature, gastropod shells were cracked with a vice and removed from shells using stainless steel tweezers that had been rinsed with ethanol (Pronalys, AR grade, Selby Biolab). For gender analysis, the sex of the organism was determined by visual examination of the sex organs. The sexes were determined by the colouration of the gonads with females having pale grey or grey–green ovaries, the males having yellowish green, yellowish brown to creamy pink testis [27] and a penis located near the mantle [28]. Samples were rinsed with deionised water to remove residual shell from the soft tissue and placed in acid washed 20 mL polyethylene snap cap vials and frozen until digested. 2.4. Sample digestion and analysis Samples were freeze dried and digested as outlined in Baldwin et al. [29]. The individual dry samples were weighed (0.07 g) into acid washed 7 mL Teflon bombs to which 1 mL concentrated nitric acid (Aristar, BDH) was added. Twenty-two samples were digested in a 600 W microwave oven (MDS 81D, CEM, Indian Trail, NC, USA) using low volume microwave digestion. Samples were digested using the following program: 2 min at 600 W; 2 min at 0 W; 45 min at 450 W. Digested samples were transferred into individual, acid washed, 10 mL centrifuge tubes and diluted to 10 mL with deionised water. Samples for temporal study were again diluted 1/ 10 V/V in preparation for ICP-MS analysis. Mass, gender and shoreline position trace metal analyses were carried out using electrothermal atomic absorption spectroscopy (Perkin-Elmer 5100PC) with Zeeman background correction. A magnesium/ammonium dihydrogen phosphate matrix modifier was used for Cd analysis [30] with a palladium and magnesium
D. Gay, W. Maher / Water Research 37 (2003) 2173–2185
2176
matrix modifier used for Se analysis [31]. Copper and zinc analysis was carried out using flame atomic absorption spectroscopy with deuterium background correction. Temporal variation trace metal analyses (copper, zinc, cadmium and selenium) were carried out using the ICPMS [32]. Standard reference material NIST 1566a Oyster tissue was routinely run with sample digestions (n ¼ 20). Recoveries (in mg/g) for copper (6277), zinc (893784), cadmium (4.3570.45) and selenium (2.2470.15) were in general agreement with certified values (66.374.3; 830+57; 4.1570.38; 2.2170.24, respectively). Precision (RSD (%)) for measurement of copper, zinc, cadmium and zinc in Oyster tissue were 7%, 3%, 5% and 8% for a 20-mg sample and 5%, 3%, 5% and 6% or a 100-mg sample, respectively, indicating that precision did not vary greatly with the mass analysed. Analyses of replicated samples were within 8% for all trace metals.
3. Results Total metal concentrations represented as median concentrations, mean concentrations standard deviations and ranges are shown in Table 1. Medians are
reported, as they are a better representation of results as the data were not normally distributed. Selenium concentrations in samples below 0.02 g dry mass were below instrumental quantitation limits (0.05 mg/g). This dry mass corresponded to a shell width below 12 mm. These samples were assigned a selenium concentration of 0.01 mg/g for further statistical analysis. Frequency distributions on raw data of copper, zinc, cadmium and selenium were all positively skewed (Fig. 1). The Shapiro–Wilks test shows deviations from normality. 3.1. Effect of gender Five samples of shell width below 12 mm were unable to be sexed due to immaturity of sexual organs. These samples were omitted for gender analysis. ANCOVA using log transformed data showed no significant difference data between male and female B. nanum species for copper (F ¼ 0:00; DF=1,92; p ¼ 0:988), zinc (F ¼ 1:17; DF=1,91; p ¼ 0:283), cadmium (F ¼ 0:40; DF=1,90; p ¼ 0:527) and selenium (F ¼ 0:20; DF=1,83; p ¼ 0:658). This was confirmed with shoreline data showing no significant difference between males and females for Cu, Zn and Cd at transects which had sufficient numbers of males and females for comparison (transects 1 and 2)(Cu: F ¼ 0:46; DF=1;
Table 1 Metal concentrations in B. nanum from Sites on the NSW Coast, Australia Site
Copper JB CB BB PK
N
(mg/g) 96 22 25 25
Zinc (mg/g) JB 95 CB 22 BB 25 PK 25
Mean
SD
Median
Range
Skewness
Kurtosis
Normality
Dry mass (g)
po(W)
Mean
SD
152 37 94 379
90 30 55 152
130 26 83 336
23–473 5–143 15–296 196–667
1.11 2.32 2.16 0.61
1.37 6.80 7.12 1.00
0.0001 0.0001 0.0001 0.0001
0.06 0.20 0.11 0.13
0.04 0.05 0.03 0.02
67 63 127 110
20 27 87 32
65 59 95 93
12–164 36–152 63–452 67–180
1.26 1.81 2.66 0.96
5.14 4.43 7.97 0.15
0.0001 0.0001 0.0001 0.0001
0.06 0.20 0.11 0.13
0.04 0.05 0.03 0.02
Cadmium (mg/g) JB 94 CB 22 BB 25 PK 25
1.07 3.47 0.67 0.78
0.51 1.01 0.24 0.29
1.10 3.43 0.60 0.74
0.20–3.06 1.79–5.84 0.35–1.40 0.38–1.47
0.53 0.46 1.24 0.89
1.22 0.05 1.73 0.23
0.0001 0.0001 0.0001 0.0001
0.06 0.20 0.11 0.13
0.04 0.05 0.03 0.02
Selenium (mg/g) JB 81 CB 22 BB 25 PK 25
0.86 0.86 0.72 0.73
0.53 0.16 0.13 0.11
0.83 0.83 0.70 0.70
0.11–2.94 0.59–1.33 0.56–1.08 0.58–1.04
1.69 1.28 1.35 1.06
4.20 2.61 2.25 1.19
0.0001 0.0001 0.0001 0.0001
0.06 0.20 0.11 0.13
0.04 0.05 0.03 0.02
JB=Jervis Bay; CB=Caves Beach; BB=Botany Bay; PK=Port Kembla; N=Number of Organisms; SD=Standard Deviation; W=Shapiro–Wilks statistic for normality.
D. Gay, W. Maher / Water Research 37 (2003) 2173–2185
2177
3.2. Effect of mass As there were no significant differences between female and male B. nanum, all samples were combined for mass regression analysis. Regression analyses for copper, zinc, cadmium and selenium are shown in Fig. 2. No significant regressions were found for log transformed zinc (r2 ¼ 0:013; p ¼ 0:277), cadmium (r2 ¼ 0:015; p ¼ 0:243) and selenium (r2 ¼ 0:003; p ¼ 0:593) concentrations with mass for B. nanum. Regression analysis showed a significant relationship for log transformed copper concentrations and mass in B. nanum species (r2 ¼ 0:19; po0:001). This states that 19% of the variation in B. nanum copper concentrations can be explained by mass, with organisms of greater mass having higher copper concentrations. Analysis of shoreline data confirmed this result. Regression analysis found a significant positive correlation of dry mass with copper concentration (F ¼ 13:9; DF=1,103; po0:0005). This relationship only explained 12% of the variation in copper concentrations. No significant regressions of dry mass with zinc (F ¼ 0:09; DF=1,103; p ¼ 0:76), cadmium (F ¼ 0:28; DF=1,103; p ¼ 0:60) and selenium (F ¼ 0:19; DF=1,66; p ¼ 0:67) concentrations were found. There was the possibility that analysing males and females together was masking underlying differences in gender associated with the mass of the species. When males and females were analysed separately a significant regression of cadmium concentration and mass was found only for males (r2 ¼ 0:213; po0:001). Generally, non-significant relationships between metal concentrations and mass for B. nanum were found at contaminated sites (Table 2). 3.3. Effect of sample size Results for determination of sample size required to detect specific differences are presented in Fig. 3. Calculations used a t statistic of t½0:05; a ¼ 1:96 with N degrees of freedom. Results varied for each metal with around 10–15 organisms required to detect a 30% change in the mean. 3.4. Effect of shoreline position Fig. 1. Frequency distributions of Cu, Zn, Cd and Se (mg/g) for B. nanum, Jervis Bay, NSW. x-axis is ranges.
po0:51;F ¼ 0:76; DF=1; po0:39; Zn: F ¼ 1:88; DF=1; po0:18; F ¼ 3:32; DF=1; po0:08; Cd: F ¼ 0:06; DF=1; po0:81; F ¼ 2:89; DF=1; po0:1), with only a small but significant difference in selenium concentrations found at transect 2 (F ¼ 8:01; DF=1; po0:01) but not at transect 1 (F ¼ 0:1; DF=1; po0:75).
As no differences that would greatly influence metal concentrations (Cu, Zn, Cd and Se) were found between males and females and mass, data was pooled to determine differences in metal concentrations in B. nanum, between transects. Significant differences were found between transects for copper (F ¼ 16:18; DF=4,00; po0:0001), zinc (F ¼ 4:95; DF=4,100; po0:001) and cadmium (F ¼ 2:77; DF=4,100; po0:05) concentrations. No significant differences were found between transects for selenium (F ¼ 1:43; DF=2,65; p ¼ 0:25) concentrations. Ranking from
D. Gay, W. Maher / Water Research 37 (2003) 2173–2185
2178
Copper
150 Sample Size
3.00 log (Cu µg/g)
2.50 2.00 1.50 1.00
Copper 135
100 50
34
0 10%
0.50 0.05
0.15
0.2
6 50%
Zinc 40
Zinc
2.50 log (Zn µg/g)
0.1 Dry mass (g)
Sample size
0
2.00 1.50
35
30 20 10
9 4
0 10%
1.00
20%
30%
2 50%
Percent change in mean
0.50 0
0.05
0.1
0.15
0.2
Dry mass (g) Cadmium 0.70 0.60 0.50 0.40 0.30 0.20 0.10 0.00
Sample size
0.00
100 80
Cadmium 88
60 40 22
20 0 10%
0.1
0.15
0.2
Dry mass (g)
Sample size
0.05
150
20% 30% Percent change in mean
4 50%
146
100 50
37
0 10%
Selenium
10
Selenium
200
0
20%
17 30%
6 50%
Percent change in mean
0.70 log (Se µg/g)
30%
Percent change in mean
0.00
log (Cd µg/g)
20%
15
0.60 0.50 0.40 0.30 0.20 0.10 0.00
Fig. 3. Sample sizes required to detect changes in mean trace metal concentrations for Cu, Zn, Cd and Se in Bembicium nanum from Jervis Bay, NSW. Note vertical scales are not equal.
0
0.05
0.1 Dry mass (g)
0.15
0.2
Fig. 2. Regression analysis for Cu, Zn, Cd and Se concentrations and mass for B. nanum.
highest metal concentrations to lowest by Tukey’s HSD test are (in mg/g): copper: 4=5=3>2=1; Zinc: 3=4X2X1=5, and cadmium: 3X4=5=1X2. Copper
concentrations were significantly higher at transects 3, 4 and 5 compared to the lower transects 1 and 2 (Fig. 4). Although Zinc concentrations between transects were found to be statistically significant visual examination of box whisker plots (Fig. 4) and ranking by Tukey’s HSD test, show no clear separation between transects. Cadmium concentrations were similar at all transects (Fig. 4), supported by the low significance level of the ANOVA. Data on selenium concentrations for sites 3 and 5 are missing due to experimental problems. Given the limited data for selenium at each transect, it is difficult to
D. Gay, W. Maher / Water Research 37 (2003) 2173–2185
2179
Table 2 Regression analysis for Cu, Zn, Cd and Se concentrations and mass for B. nanum at three contaminated NSW sites Copper 2
Caves Beach Botany Bay Port Kembla
Zinc 2
Cadmium 2
Selenium
r
p
r
p
r
p
r2
p
0.039 0.010 0.006
0.377 0.639 0.722
0.195 0.206 0.002
0.040* 0.023* 0.848
0.000 0.156 0.012
0.976 0.051 0.602
0.027 0.097 0.470
0.468 0.131 0.0002*
*Denotes significant relationships.
determine whether the non-significant result is due to no difference in selenium concentrations between transects 1, 2 and 4, or a reflection of a smaller sample size. 3.5. Temporal variation Significant between month differences for copper (F ¼ 5:05; DF=11,216; po0:0001) and cadmium (F ¼ 6:44; DF=11,215; po0:0001) concentrations in B. nanum were found using log transformed data. Untransformed data showed significant differences between months for zinc (F ¼ 6:32; DF=11,217; po0:0001) and selenium (F ¼ 9:05; DF=11,217; po0:0001) concentrations. Mean dry mass of selected B. nanum was constant across the 12-month period (Fig. 5). All months exhibited skewed distributions for metal concentrations, for example metal concentration frequency distributions of Cu, Zn, Cd and Se for May 1999 (Fig. 6). (i) Copper: Copper concentrations fluctuated throughout the 12-month period with peaks in September and March and lows in June, July and December. Highest mean concentration of copper was found in March. High copper concentrations in a few organisms from January to May greatly influenced mean concentrations for each of these months. Comparison of mean copper concentrations to median copper concentrations supports this. Copper body burdens followed a similar trend to copper concentrations (Fig. 5). (ii) Zinc: Zinc concentrations were found to be fairly similar for the 12-month period. Peaks were recorded for October and February, with lows in June, July, August and November, although clear separation of sites for zinc concentrations could not be determined using Tukey’s HSD test. Zinc body burdens followed zinc concentrations except for the months June, August and November, which showed higher zinc body burden. These months also had higher mean mass compared to other months. (iii) Cadmium: Cadmium concentrations fluctuated throughout the 12-month period, with peak con-
centrations in September, January and March, and low concentrations in August, November, December and February (Fig. 5). Again, greater variability in cadmium concentrations in the months from January to May influenced mean concentrations for these months. Cadmium body burdens closely followed cadmium concentrations over the 12 months (Fig. 5). (iv) Selenium: Selenium concentrations in B. nanum were lowest in June and July, increasing to an October peak. Concentrations then remained fairly constant for the rest of the 12-month period. The range of concentrations for each month also remained similar. Selenium body burdens followed selenium concentrations (Fig. 5). . 3.6. Accumulation of trace metals in contaminated environments B. nanum had increased trace metal concentrations at Port Kembla, Botany Bay and Caves Beach (Table 1), locations known to be contaminated with copper, zinc and cadmium [24–26]. Frequency distributions of these trace metals are also skewed to higher concentrations at these sites (Fig. 7). The higher concentrations measured in B. nanum at the known contaminated sites show that B. nanum accumulate Cu, Zn and Cd in response to exposure to these metals. B. nanum were also abundant at these sites indicating they are hardy and tolerate high concentrations of trace metals.
4. Discussion 4.1. Factors contributing to natural variation in trace metal concentrations 4.1.1. Gender No significant difference of copper, zinc, cadmium and selenium were found between male and female B. nanum. Of the few studies that have looked at the effect of gender on trace metal differences in gastropods, gender has not been found to significantly affect metal
2180
D. Gay, W. Maher / Water Research 37 (2003) 2173–2185
Fig. 4. Concentrations of Cu, Zn, Cd and Se (mg/g) in B. nanum at each transect. Box whisker plots represent line: median; box: 25th– 75th percentile; whiskers: 5th–95th percentile.
concentrations in whole tissue analysis (L. littorea, [33] Bembicium auratum, Austrocochlea constricta, [34]). 4.1.2. Mass Copper, zinc, cadmium and selenium concentrations in B. nanum were generally independent of mass
(Table 2, Fig. 2). Although copper concentrations were significantly correlated with mass, it accounted for only 19% of the variation. The small amounts of variability attributable to mass are not significant compared to overall inherent variability (see next section).
14 12 10 8 6 4 2 0
Copper
180 160 140 120 100 80 60 40 20 0
2181
[Cu] (µg/g)
Dry Mass (x100) (g) and Total Cu (µg)
D. Gay, W. Maher / Water Research 37 (2003) 2173–2185
10
Zinc
8 6 4 2 0
140 120 100 80 60 40 20 0
[Zn] (µg/g)
Dry Mass (x100) (g) and Total Zn (µg)
Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Month
Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Month 7 6 5 4 3 2 1 0
[Cd] (µg/g)
1.6 1.4 1.2 1 0.8 0.6 0.4 0.2 0
Dry Mass (g) and [Se] (µg/g)
Dry Mass (g) and Total Cd (µg)
Cadmium 0.4 0.35 0.3 0.25 0.2 0.15 0.1 0.05 0 Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Month
Total Se (µg)
Selenium 0.14 0.12 0.1 0.08 0.06 0.04 0.02 0 Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Month
Fig. 5. Temporal variation of metal concentration, metal body burden and dry mass of Cu, Zn, Cd and Se in B. nanum. Each point is n ¼ 20; error bars are omitted for clarity but are typically x–y%. Legend: K Dry mass (g); ’ Metal Concentration (mg/g); m Metal Body Burden (mg).
In a number of gastropod species, metal concentrations have been found to be statistically independent of mass (Cu, Cd: Littorea. littorea, Cu, Zn, Cd: Scaphander lignarius, Credipula fornicata; [35] Se: B. nanum, A. constricta, Nerita atramentosa, [22] Cd: H. tuberculata,[36]) but often show an increase/decrease with mass/ size [7,37]). Boyden [35] has stated that it is difficult to determine mass influences on metal concentrations by regression analysis unless a species exhibits a tenfold increase in size range, due to the inherent variability shown by biological organisms. As B. nanum naturally exhibits a small mass range (0.01–0.165 g in this case) and organisms that can be easily analysed are in the range 0.05–0.165 g, significance relationships of metal concentration and mass are not expected. However, as
Fig. 6. Metal concentration frequency distributions for Cu, Zn, Cd and Se, Rosedale, NSW (May 1999). x-axis is ranges.
trace metals often show a trend with mass, similar mass organisms should be selected for analysis when comparing different locations. 4.1.3. Inherent variability All metal distributions in B. nanum were not normally distributed (Fig. 1). Copper, zinc, cadmium and selenium had positively skewed distributions. Most
D. Gay, W. Maher / Water Research 37 (2003) 2173–2185
Port Kembla
[Zn] µg/g
More
550
450
2.75
2.25
1.75
0.75
18 16 14 12 10 8 6 4 2 0 0.25
No. of Individuals More
180
140
100
60
350
Caves Beach
16 14 12 10 8 6 4 2 0 20
250
50
[Cu] µg/g
Botany Bay No. of individuals
150
No. of Individuals [Zn] µg/g
18 16 14 12 10 8 6 4 2 0
More
180
140
100
60
20
No. of Individuals
Port Kembla 16 14 12 10 8 6 4 2 0
1.25
2182
[Cd] µg/g
Fig. 7. Metal concentration frequency distributions at Port Kembla (Cu, Zn), Botany Bay (Zn) and Caves Beach (Cd). x-axis is ranges.
individuals had lower metal concentrations with only a small number showing higher concentrations, extending the tail of the distribution (Fig. 1). Skewness could not be explained by variations in mass or gender as the metals were mostly not correlated with mass and no significant differences were found between males and females in B. nanum. Trace element concentrations in marine molluscs have been found to have positively skewed distributions in non-contaminated environments [38,39,33]. Non-normal distributions have been found for zinc in M. edulis, for copper in L. littorea [33] and for selenium in B. nanum [22]. Non-normal distributions are generally attributed to inherent variability due to individual variations in uptake and excretion of trace elements by organisms (Lobel et al. [33]), caused by physiological and metabolic differences between organisms. Metal accumulation in B. nanum will be affected by microhabitat differences on rocky platforms due to rock pools, crevices and over hangs which cause differences in shelter (stress), food availability, food types, feeding periods and metabolism [40–42]. All these factors will contribute to trace metal concentration variability within gastropod samples. 4.1.4. Shoreline position Shoreline position did not greatly influence cadmium and selenium concentrations in B. nanum (Fig. 4).
Copper and zinc concentrations were influenced by shoreline position with gastropods collected from transects further away from the low tide mark having increased trace metal concentrations. B. nanum are grazers feeding on microscopic algae and bacteria [43]. Trends in copper and zinc concentrations could be explained by variations in food type, food availability and immersion that affect growth rates and trace metal intake. The observations for copper and zinc are similar to the trend found for zinc concentrations in the barnacle Elminius modestus [44]. Longer immersion times in E. modestus were found to result in increased growth due to longer feeding times and corresponding lower zinc concentrations. B. nanum only feed whilst immersed [43]. Gastropods located higher on the shore would experience shorter immersion times, resulting in greater desiccation and decreased growth altering metabolic rates [45] thus influencing metal concentrations. It is expected that variability exists in food type and amounts at different heights on the shore. Underwood and Atkinson [42] observed less splashing of seawater at Jervis Bay higher on shore due to lower waves than would be experienced at more exposed rocky platforms. This would produce drier conditions and alter algal cover higher on shore and thus affect food availability. Underwood [43] found that growth rates in B. nanum were only affected at the highest position on shore,
D. Gay, W. Maher / Water Research 37 (2003) 2173–2185
explained by a decrease in algal cover and thus a decrease in food supply. Feeding rates of individuals of different sizes of B. nanum have been found not to be significantly different or influenced by shore position [43,46]. 4.1.5. Temporal changes Copper, zinc, cadmium and selenium concentrations were found to vary temporally in B. nanum, yet no clear seasonal trends were evident (Fig. 5). Temporal variation in metal concentrations in molluscs has been found in previous studies [47–49]. These studies found seasonal variation in metal concentrations associated with changes in mass of the organism, with maximums in winter and early spring and minimums in late spring and summer. Temporal variation associated with changes in the mass of B. nanum is not likely as the mean mass of the gastropods sampled over the 12-month period were relatively constant (Fig. 5). A number of factors have been identified to influence temporal variation in metal concentrations. These are variation in pollution delivery to the environment, changes in salinity and water temperature, and an organism’s physiology, associated with reproductive cycles [4]. Salinity is constant in the areas studied so is not a factor influencing temporal variability of trace metal concentrations. No known point sources of metal contamination exist in the Rosedale area. Sewage at Rosedale, a possible source of trace metals, is mostly taken off-site for disposal. Median concentrations of zinc, cadmium and selenium were higher in B. nanum at Rosedale compared to metal concentrations found in B. nanum at Jervis Bay. Rosedale receives storm water from urban areas during flood events by overland transport that may result in higher trace metal concentrations in gastropods at this location than at Jervis Bay and cause some of the temporal variation of trace metal concentrations. Water temperature during this study ranged from 231C in summer to 131C in winter and could be influencing trace metal concentrations in B. nanum. Laboratory experiments have shown that temperature changes of this magnitude can result in the increase or decrease of trace metal concentrations because of changes in metabolic and excretion rates [50,47,51]. Changes in temperature can also trigger spawning in molluscs [52], which can change metal concentrations in molluscs through the loss of mature oocytes. B. nanum have been found to spawn during summer with a peak in spawning in January and finishing by February/March [53,54]. If metals were to be lost with mature oocytes, it is expected that decreases in mass and total body metal burden would occur. No significant mass losses occurred during summer, possibly due to ample food supply during summer thus organisms maintained body mass.
2183
Taylor [34] found that mature oocytes during spawning in a similar gastropod B. auratum were released slowly over an extended period, which could also explain the absence of mass losses in summer due to spawning activity. It is likely that reproductive cycles do not significantly influence trace metal body burdens in B. nanum. 4.2. Use of B. nanum as a biomonitor B. nanum meets the requirements of a suitable biomonitor species as summarised by Phillips [3–5] as it is * *
*
*
Sedentary and thus representative of a location. Hardy and tolerate high levels of pollutants as indicated by their high abundance at contaminated sites. Abundant on all rock platforms, easy to identify and provide enough tissue (0.05–0.1 g dry mass) for analysis. Accumulate trace metals in contaminated environments, although we have not established a simple relationship between contamination and bioavailable trace metal concentration.
In addition, gender has been found not to significantly influence trace metal concentrations of copper, zinc, cadmium and selenium in B. nanum. This study has shown that trace metal concentrations can vary with gastropod mass, shoreline position and temporally. Thus care must be taken to collect individual gastropods of similar mass from similar shoreline positions and times. Sampling in the midlittoral zone will minimise influences due to shore position. The large natural variability in metal concentrations of copper, zinc, cadmium and selenium in B. nanum populations would make it unsuitable to monitor subtle changes in trace metal contamination of a given area but this species can be used to indicate areas of gross contamination. Sample size calculations based on the natural variability of trace metals in B. nanum (Fig. 3) indicate that around 10–15 organisms are enough to detect a 30% change in the mean of a sample for B. nanum. Sample size estimates for future analyses would be dependent on the percentage change in the mean required to meet the purpose of the study. To detect very low differences in means, i.e. less than 10%, would require large sample sizes (greater than 80 organisms) due to the inherent variability exhibited by this species. For financial reasons, sample sizes of this magnitude are rarely used for routine monitoring. More studies are required to demonstrate the nature of the response of B. nanum to metals in contaminated environments. However, generally the use of a biomo-
2184
D. Gay, W. Maher / Water Research 37 (2003) 2173–2185
nitor is to show that contamination exists in an area, not how much contamination there is.
References [1] Allan RJ, Campbell PGC, Forstner U, Lum K. Fate and effects of toxic elements in large rivers and their estuaries. Sci Total Environ 1990;97/98:1–872. [2] SOE. Australia: State of the Environment, Australia, Collingwood: CSIRO Publishing, 1996. [3] Phillips DJH. The use of biological indicator organisms to monitor trace metal pollution in marine and estuarine environments—a review. Environ Pollut 1977;13:281–317. [4] Phillips DJH. Quantitative aquatic biological indicators. London: Applied Science Publishers, 1980. [5] Phillips DJH. Use of macroalgae and invertebrates as monitors of metal levels in estuaries and coastal waters. In: Furness RW, Rainbow PS, editors. Heavy metals in the marine environment. Boca Raton, FL: CRC. Press Inc, 1990. p. 81–99. [6] Phillips DJH, Rainbow PS. The biomonitoring of trace metals and radionuclides. In: Phillips DJH, editor. Biomonitoring of trace aquatic contaminants. Amsterdam: Elsevier Science Publishers Ltd., 1993. p. 79–132. [7] Langston WJ, Bebianno MJ, Burt GR. Metal handling strategies in molluscs. In: Langstone WJ, Bebianno MJ, editors. Metal metabolism in aquatic environments. London: Chapman & Hall, 1998. p. 219–85 [Chapter 8]. [8] Nicolaidou A, Nott JA. Mediterranean pollution from a ferro-nickel smelter. Mar Pollut Bull 1990;21:137–43. [9] Nicolaidou A, Nott JA. Metals in sediment, seagrass and gastropods near a nickel smelter in Greece. Possible interactions. Mar Pollut Bull 1998;36:360–5. [10] Abdallah AT, Moustafa MA. Accumulation of lead and cadmium in the marine prosobranch Neriti saxtilis, chemical analysis, light and electron microscopy. Environ Pollut 2002;116:185–91. [11] Langston WJ, Spence SK. Biological factors involved in metal concentrations observed in aquatic organisms. In: Tessier A, Turner DR, editors. Metal speciation and bioavailability. New York: Wiley, 1995. p. 407–78. [12] Brown MT, Depledge MH. Determinants of trace metal concentrations in marine organisms. In: Langstone WJ, Bebianno MJ, editors. Metal metabolism in aquatic environments. London: Chapman & Hall, 1998. p. 185–217 [Chapter 7]. [13] Ireland MP, Wootton RJ. Distribution of lead, zinc, copper and manganese in the marine gastropods, Thais Lapillus and Littorina littorea, around the coast of Wales. Environ Pollut 1977;12:27–41. [14] Kang SG, Wright DA, Koh CH. Baseline metal concentration in the Asian periwinkle Littorina brevicula employed as a biomonitor to assess metal pollution in Korean coastal water. Sci Total Environ 2000;263:143–53. [15] Peerzada N, Eastbrook C, Guinea M. Heavy metal concentration in Telescopium from Darwin Harbour, N.T, Australia. Mar Pollut Bull 1990;21:307–8. [16] Blackmore G. Interspecific variation in heavy metal body concentrations in Hong Kong marine invertebrates. Environ Pollut 2001;114:303–11.
[17] Campanella L, Conti ME, Cubadda F, Sucapane C. Trace metals in seagrass, algae and molluscs from an uncontaminated area in the Mediterranean. Environ Pollut 2001;111:117–26. [18] Gorden M, Knauer GA, Martin JH. Mytilus californius as a bioindicator of trace metal pollution: variability and statistical considerations. Mar Pollut Bull 1980;10: 195–8. [19] Cho G. The Jervis Bay environment. In: Cho G, Georges A, Stoutjesdijk R, editors. Jervis Bay, A place of cultural, scientific and educational value. Canberra: ANPWS, 1995. p. 3–8. [20] Jones N, Abell R, Jacobson G. Hydrology. In: Cho G, Georges A, Stoutjesdijk R, editors. Jervis Bay a place of cultural scientific and educational values. Canberra: ANPSW, 1995. p. 53–64. [21] Taylor G, Abell R, Paterson I. Geology, geomorphology, soils and earth resources. In: Cho G, Georges A, Stoutjesdijk R, editors. Jervis Bay. A place of cultural, scientific and educational value. Canberra: ANPWS, 1995. p. 41–52. [22] Baldwin S, Maher W, Kleber E, Krikowa F. Selenium in marine organisms of seagrass habitats (Posidonia australis) of Jervis Bay, Australia. Mar Pollut Bull 1996; 32(3):310–6. [23] Holloway PE. Water circulation. In: Cho G, Georges A, Stoutjesdijk R, editors. Jervis Bay. A place of cultural, scientific and educational value. Canberra: ANPWS, 1995. p. 65–70. [24] Hayes WJ, Anderson IJ, Gaffoor MZ, Hurtado J. Trace metals in oysters and sediments of Botany Bay, Sydney. Sci Total Environ 1998;212:39–47. [25] Hall SJ, Miller LB, Gidding LA. Hard substrate benthic communities. In: Marine investigation for the Bellambi and Port Kembla STP Upgradings Annual Report. Environment Management Unit, Water Board, Sydney, 1991. [26] Moran PJ. Water quality control and its effects on the concentration of heavy metals in Port Kembla Harbour, NSW. Mar Pollut Bull 1984;15:294–7. [27] Anderson H. The gastropod genus Bembicium Phillipe. Aust J Mar Freshwater Resources 1958;9:547–73. [28] Galstoff PS. Physiology of reproduction in molluscs. Am Zool 1961;1:273–89. [29] Baldwin S, Deaker M, Maher W. Low-volume microwave digestion of marine biological tissues for the measurement of trace elements. Analyst 1994;119:1701–4. [30] Slavin W. Graphite Furnace AAS. A Source Book. United States of America: The Perkin Elmer Corporation, 1994. p. 83. [31] Deaker M, Maher W. Determination of selenium in seleno-compounds and marine biological tissues using stabilised temperature platform furnace atomic absorption spectrometry. J Anal At Spectrosc 1995;10:423–34. [32] Maher W, Forster S, Krikowa F, Snitch P, Chapple G, Craig P. Measurement of trace elements and phosphorus in marine animal and plant tissues by low-volume microwave digestion and ICP-MS. At Spectrosc 2001; 22(5):361–70. [33] Lobel PB, Mogie P, Wright DA, Wu BL. Metal accumulation in four molluscs. Mar Pollut Bull 1982;13:170–4.
D. Gay, W. Maher / Water Research 37 (2003) 2173–2185 [34] Taylor A. Zinc, copper and cadmium accumulation, detoxification and storage in the gastropod molluscs Austrocochlea constricta and Bembicium auratum and an assessment of their potential as biomonitors of trace metal pollution in estuarine environments. Thesis for the Degree of Master of Applied Science, University of Canberra, ACT, 1998. [35] Boyden CR. Effect of size upon metal content of shellfish. J Mar Biol Assoc UK 1977;57:675–714. [36] Bryan GW, Potts GW, Forster GR. Heavy metals in the gastropod mollusc Haliotis tuberculata (L). J Mar Biol Assoc UK 1977;57:379–90. [37] Leung KMY, Furness RW. Effects of animal size on concentrations of metallothionein and metals in periwinkles Littorina littorea collected from the Firth of Clyde, Scotland. Mar Pollut Bull 1999;39:126–36. [38] Lobel PB, Belkhode SP, Bajdik C, Jackson SE, Longerich HP. General characteristics of the frequency distributions of element concentrations and of interelemental correlations in aquatic organisms. Mar Environ Res 1992;33: 111–26. [39] Lobel PB, Bojdik CD, Belkhode SP, Jackson SE, Longerich HP. Improved protocol for collecting mussel watch specimens taking into account sex, size, condition, shell shape and chronological age. Arch Environ Contamin Toxicol 1991;21:409–14. [40] Fairweather PG, Underwood AJ, Moran MJ. Preliminary investigation of predation by the whelk Morula marginalba. Mar Ecol Prog Ser 1984;17:143–56. [41] Fletcher WJ. Interactions among subtidal Australian sea urchins, gastropods, and algae: effects of experimental removals. Ecol Monogr 1987;57:89–109. [42] Underwood AJ, Atkinson MH. In: Cho G, Georges A, Stoutjesdijk R, editors. Jervis Bay. A place of cultural, scientific and educational value. Canberra: ANPWS, 1995. p. 123–32. [43] Underwood A. Microalgal food and the growth of the intertidal gastropods Nerita atramentosa Reeve and Bembicium nanum (Lamarck) at four heights on a shore. J Exp Mar Biol Ecol 1984;79:277–91. [44] Al-Thaqafi K, White KN. Effect of shore position and environmental levels on body metal burdens in the
[45]
[46]
[47]
[48]
[49]
[50]
[51]
[52]
[53]
[54]
2185
barnacle, Elminius modestus. Environ Pollut 1991;69: 89–104. Barnes H, Barnes M, Finlayson DM. The metabolism during starvation of Balanus balanoides. J Mar Biol Assoc UK 1963;43:213–23. Petraitis PS. Effects of body size and water temperature on grazing rates of four intertidal gastropods. Aust J Ecol 1992;17:409–14. Bryan GW. Some aspects of heavy metal tolerance in aquatic organisms. In: Lockwood APM, editor. Effect of pollutants on aquatic organisms, vol. 2. Society for Experimental Biology Seminar Series. Cambridge University Press, 1976. p. 7–34. Phillips DJH. The common mussel Mytilus edulis as an indicator of pollution by zinc, cadmium, lead and copper. I. Effects of environmental variables on uptake of metals. Mar Biol 1976;38:59–69. Simpson CD. Uptake and loss of zinc and lead by mussels (Mytilus edulis) and relationships with body weight and reproductive cycle. Mar Pollut Bull 1979;10: 74–8. Fowler SW, Benayoun G. Influence of environmental factors on selenium flux in two marine invertebrates. Mar Biol 1976;37:59–68. Orren MJ, Eagle GA, Hennig HF-KO, Green A. Variations in trace metal content of the mussel Chromytilus meridionalis (Kr.) with season and sex. Mar Pollut Bull 1980;11:253–7. Hadfield AJ, Anderson DT. Reproductive cycles of the bivalve molluscs Anadara trapezia (Deshayes), Venerupis crenata Lamarck and Anomia descripta Inedale in the Sydney region. Aust J Mar Freshwater Res 1988;39: 649–60. Anderson DT. The reproduction and early life history of the gastropod B. nanum (Lamarck 1822) (Fam. LITTORINIDAE). The Proceedings of the Linnean Society of NSW, Sydney, vol. 1xxxvi(part 2) 1961. p. 203–6. Underwood AJ. The reproductive cycles and geographical distribution of some common Eastern Australian prosobranchs (Mollusca: gastropoda). Aust J Mar Freshwater Res 1974;25:63–88.