Centre for Advanced Analytical Chemistry Energy Technology

COPPER SPECIATION AND TOXICITY IN THE FLY RIVER: A REVIEW

by N.J. Rogers, S.C. Apte, J.L. Stauber and A.W. Storey* * Wetland Research & Management Pty Ltd

Report No: ET/IR745R Prepared for Ok Tedi Mining Ltd

May 2005

Centre for Environmental Contaminants Research

i

EXECUTIVE SUMMARY Under the Environmental Regime, OTML is required to monitor copper bioavailability and aquatic toxicity at 4 specified locations on the Ok Tedi-Fly River system. The monitoring program adopted to date is based on the chemical measurement of copper speciation (labile copper and complexation capacity determination) and bioassays using copper-sensitive algae and bacteria. Both bioassays respond to concentrations of biologically-available copper in the low µg L-1 range and are intended to act as an early warning of copper toxicity to aquatic organisms resident in the Fly River system. The Regime monitoring results over the period 2003-2005 indicate a significant increase in algal and bacterial toxicity compared to measurements made in the 1990s. After almost 3 years of quarterly monitoring, it is now appropriate to critically review the monitoring data and to assess the potential effects of copper toxicity on higher aquatic organisms living in the Fly River system, and also to consider the effects on aquatic communities. The specific objectives of this study were as follows: •

To evaluate and report on all previous copper toxicity work carried out on the Fly River system.

•

To fully evaluate routine bioassay and copper speciation data collected by OTML since the commencement of the Environmental Regime in December 2001.

•

To review relevant recent literature information on copper toxicity to freshwater aquatic organisms.

•

To use the findings of the current food-web studies conducted by OTML to develop an understanding of the potential toxic impacts of copper on biological communities in the Fly River system. This includes the effect of copper toxicity on primary carbon production.

•

To recommend future studies required to assess the effects of bioavailable copper on aquatic organisms in the Fly River system.

The findings of the review were as follows: 1. The speciation monitoring data indicate that both dissolved copper concentrations and copper complexing capacities in the Upper and Middle Fly River have remained relatively constant over the survey period (1996 - 2004). However, in the Middle Fly, labile copper concentrations (a surrogate for biologically-available copper) appear to be increasing. The reason for this increase is unknown, but may be due either to inputs of more reactive copper from a change in mining throughput and/or the dredge stockpiles at Bige, or to vegetation dieback on the flood-plain and a consequent reduced organic matter input which influences copper speciation.

Centre for Environmental Contaminants Research

ii

2. The frequency of algal growth inhibition has increased over the duration of the quarterly speciation surveys. At sites on the Middle Fly River, the increase in the frequency of algal growth inhibition corresponds to an increase in labile copper concentrations. The magnitude of algal inhibition has not significantly increased over the duration of the quarterly surveys and there is no significant correlation between the amount of algal inhibition and labile copper concentrations in the river. Addition of the metal chelating agent EDTA to water samples consistently removed the observed toxicity in both the algal and bacterial bioassays. This indicated that the toxic effects were definitely related to the presence of metal contaminants. The combined toxic effect of several metals or antagonistic interactions between metals cannot at this stage be ruled out. 3. The effects of labile copper concentrations on other freshwater aquatic organisms was assessed using best practice toxicity assessment procedures. Cumulative frequency distributions of the copper sensitivity of freshwater species (species sensitivity distributions) were compiled from high quality published toxicity data. These data compilations indicate that some bacteria, algae and cladocerans (e.g. water fleas) are the most sensitive organisms affected by copper toxicity. 4. The impacts of copper toxicity were analysed in terms of acute and chronic exposure to copper. At the labile copper concentrations measured during the Regime monitoring program, acute copper toxicity is not an issue. At worst, acute toxicity may be observed in fewer than 10% of invertebrate species, with no acute effects on fish. However, at current labile copper concentrations, chronic effects of copper resulting from long-term exposure to elevated bioavailable copper concentrations may be expected in 50 to 80% of freshwater species. Any further increases in labile copper in the river may result in chronic toxicity to the majority of freshwater species. The uncertainties in these calculations and conservative nature of this ‘best practice’ assessment are recognised and discussed in the main body of the report. 5. Food web stable isotope studies clearly indicate the importance of microalgae, such as periphyton, in the food webs of the Fly River system. Recent work suggests that periphyton abundance is reduced in the river sections downstream of D’Albertis Junction, and the contribution of algal carbon to the aquatic food web in these sections is also reduced compared with areas upstream of the mine. The factors affecting periphyton abundance in the Fly River remain to be established. At this stage, it is not possible to deconvolute the effects of copper, turbidity (with respect to light inhibition or scouring) and loss of habitat. Nevertheless, the concentrations of labile copper observed in the river system are in the range that may cause growth reduction in algae. Loss of algal carbon food sources and consequent effects on higher-level consumers, irrespective of the cause, is an important potential impact. It is recommended that: 1. OTML should continue the monitoring surveys (labile copper, bacterial and algal bioassays) to further track the trends that have been observed to date. This

Centre for Environmental Contaminants Research

iii

is especially important at sites in the Middle and Lower Fly River, where labile copper and toxicity appear to be increasing. Increased monitoring of off-river water bodies may also be appropriate. The frequency and location of monitoring sites should be reviewed. A major consideration for any future monitoring program is the incorporation of a sampling frequency that adequately captures the appropriate temporal and spatial scales of variability. It is likely that the greatest copper toxicity is observed after events such as drying and wetting of the floodplain (e.g. the first rainfall after a prolonged dry period). It is highly recommended that event-based sampling is conducted in order to capture some of these events. 2. Given that food web studies showed that periphyton provided an important carbon source supporting aquatic food chains in the Fly River system, it is recommended that the effect of both increased bioavailable copper concentrations and increased turbidity on periphyton communities in the Fly River should be understood. As a first step, it is recommended that periphyton populations be collected from both the Middle Fly and Kiunga and species diversity and abundance, and sensitivity/tolerance to copper and turbidity, be compared. 3. There is a lack of basic ecotoxicological data on the effects of copper on primary consumers (algal grazers) such as small invertebrates and zooplankton, planktivorous fish and other key aquatic organisms in the Fly River system. This is an important knowledge gap. We recommend a program of laboratory-based toxicity testing to address this issue. If feasible, in situ toxicity testing may also be useful. 4. Recent work at CSIRO has led to the development of a robust Chelex column method for measuring copper speciation in natural waters which is superior to the previously utilised ASV-labile method. The Chelex method is relatively simple and does not require specialist equipment. It is recommended that this new method be set up at OTML and used routinely to screen copper speciation. This method would be particularly useful for investigating copper speciation during high copper spike events in the river system.

Centre for Environmental Contaminants Research

iv

TABLE OF CONTENTS EXECUTIVE SUMMARY .................................................................................................ii 1

INTRODUCTION............................................................................... 1

2 2.1 2.2 2.3 2.4 2.5

COPPER SPECIATION AND TOXICITY IN THE FLY RIVER .......... 2 Overview of Copper Toxicity and Bioavailability............................................... 2 Measurement of Copper Speciation................................................................. 3 Biomonitoring in the Fly River .......................................................................... 3 Invertebrates and Fish...................................................................................... 3 Microalgal Studies ............................................................................................ 5

3.1 3.2 3.2.1 3.2.2 3.2.3 3.3

ROUTINE BIOASSAY AND COPPER SPECIATION DATA ............. 8 Algal and Bacterial Growth Inhibition Bioassays .............................................. 8 Routine Speciation Surveys ............................................................................. 9 Algal bioassay data..................................................................................... 11 Bacterial bioassay data............................................................................... 12 Causes of the observed toxicity.................................................................. 12 Trends in Copper Toxicity at Sampling Sites on the Fly River ....................... 15

3

4 REVIEW OF RECENT DATA FOR COPPER TOXICITY IN FRESHWATERS 23 4.1 Background to toxicity testing......................................................................... 23 4.1.1 Aquatic microorganisms ............................................................................. 23 4.1.2 Macrophytes ............................................................................................... 26 4.1.3 Freshwater invertebrates ............................................................................ 26 4.1.4 Fish ............................................................................................................. 27 4.1.5 Overview..................................................................................................... 29 5 5.1 5.2 6 RIVER 7 7.1 7.2

EVALUATION OF SPECIES SENSITIVITies to COPPER.............. 30 Acute Toxicity ................................................................................................. 30 Chronic Toxicity.............................................................................................. 31 IMPACT OF MINING ACTIVITIES ON FOOD WEBS IN THE FLY 40 CONCLUSIONS AND RECOMENDATIONS .................................. 42 Conclusions.................................................................................................... 42 Recommendations ......................................................................................... 43

Centre for Environmental Contaminants Research

v

8

REFERENCES................................................................................ 45

APPENDIX 1 Algal Bioassays on the Fly River System………………………………….55 APPENDIX 2 Complexation Capacities of Fly River (Stauber and Apte 1996) ……….56 APPENDIX 3 Speciation Survey September 1996 – June 1997………………………..57 APPENDIX 4 Routine Speciation Survey Data 2002 to 2004……………………………60 APPENDIX 5. Published acute species mean values and hardness corrected acute and chronic species mean values for copper sensitivity of freshwater species……….63

LIST OF TABLES Table 1 Acute toxicity of dissolved and particulate copper - Ranger Uranium Mine study 1991 (from Smith 1997)......................................................................... 4 Table 2 Dominant algae in 20 µm and 63 µm net samples from the Fly and Strickland Rivers (from Stauber and Apte 1996) ............................................................. 5 Table 3

Sites impacted by copper toxicity during the 2002–2004 routine speciation surveys.......................................................................................................... 11

Table 4

Correlations between copper speciation measurements and algal and bacterial bioassays (statistically significant correlations, p90% of total dissolved copper concentrations. In systems receiving elevated inputs of dissolved copper, the complexing capacity of natural organic matter may be exceeded, leading to an increase in the amount of inorganic copper in solution. This is important as it is well established that inorganic copper species, particularly the free copper ion, are the species that are most bioavailable to aquatic organisms. In most natural waters, copper toxicity is lower than predicted by the dissolved metal concentration owing to the complexation effects of dissolved organic matter. Solution pH will also affect copper speciation, with low pH giving rise to larger proportions of free copper ion. Carbonate concentrations also influence inorganic copper concentrations with copper-carbonate complexes dominating the inorganic copper pool under alkaline solution conditions. Work conducted by CSIRO for OTML in the 1990s showed conclusively that the speciation of copper in the Fly River system was dominated by copper complexes with natural organic matter. Despite elevated dissolved copper concentrations, toxicity to highly sensitive bacterial and algal bioassays was not observed. This was explained by the low bioavailability of copper-natural organic matter complexes. The binding of free copper ions to receptor sites within aquatic organisms is influenced by pH, and the concentration of competing cations such as calcium and magnesium (hardness) and sometimes other cations. Hardness correction of copper toxicity data is widely applied in water quality regulations. Unfortunately, it is not possible to generalise on the role of pH and hardness on metal bioavailability. Different organisms have different responses. It should be noted that pH can also affect solution speciation, increasing the proportion of free metal ion as pH decreases. Metal binding to receptor sites however, is generally lower under acidic pH conditions.

2

Since copper toxicity and bioavailability may vary markedly depending on the pH and hardness of waters, toxicity data are often normalised to common pH and hardness conditions before comparisons are made.

2.2

Measurement of Copper Speciation Copper speciation measurements give an important indication of copper bioavailability and toxicity to aquatic organisms. Two measurements are routinely performed as part of the regime monitoring program: (i) labile copper which is determined by anodic stripping voltammetry (Apte et al. 1995) and more recently by a Chelex column method (Bowles et al. 2005) and (ii) copper complexation capacity. The latter parameter is measured by anodic stripping voltammetry and gives a measure of the capacity of natural dissolved organic matter to complex copper. During the early stages of mine life, complexing capacity was in significant excess to dissolved copper and as a result, bioavailable copper concentrations were low. Labile copper is a measure of reactive copper and is the best available surrogate for biologically available copper. Labile copper concentrations are method dependent and comprise the free copper ion, inorganic copper complexes and weak copper-organic complexes which are easily dissociable.

2.3

Biomonitoring in the Fly River OTML have consistently monitored the abundance and species diversity of fish species in the Fly River system and this data has been invaluable in assessing the impacts of the mine on the river system. However, this form of in situ monitoring gives an indication of the impact of multiple stressors on the system, i.e. the combined effects of increased turbidity, particulate and dissolved copper, channel aggradation and forest die-back and associated habitat loss and so on. It is not possible to deconvolute individual stressors from these data and, for instance, assess the impact of dissolved copper toxicity alone. The most useful data for assessing the effects of dissolved copper come from laboratory-based toxicity tests conducted on aquatic species representative of those found in the Fly River. These data are reviewed in the ensuing sections.

2.4

Invertebrates and Fish Few data have been collected on the direct toxicity of copper on aquatic organisms indigenous to the Fly River system. Many of the currently available estimates of potential risks from direct copper exposure rely on extrapolation of data from the wider freshwater community. The two most studied groups of aquatic organisms in the Fly/Strickland system are algae and fish, with only limited data available on invertebrate species which make up a significant proportion of the aquatic fauna. Laboratory-based ecotoxicological tests of OTML mining wastes were reviewed by Smith (1997). Toxicity testing of fish, prawns, cladocerans and mayflies showed that these species had sensitivities to dissolved copper comparable to published sensitivities for Australian species. Smith et al. (1990) reported the toxic effects of particulate copper to the freshwater prawns Macrobrachium rosenbergii and M. handschini, and the catfish Neosilurus ater. The test medium was synthetic Fly River water reconstituted from Woronora River (NSW) water with the addition of sodium bicarbonate to control pH and alkalinity. If the dissolved copper

3

concentrations were kept below those likely to occur in the Fly River, little impact from particulate copper was observed at concentrations ranging from 14300 to 15100 mg Cu g-1, much higher concentrations than those expected to occur at sites downstream from the mine. Smith (1997) also reviewed work by the Ranger Uranium Mine (Ranger 1991) on the acute toxicity of dissolved and particulate copper to the fish Lates calcarifer (barramundi), Neosilurus ater, Mogurnda mogurnda (gudgeon), Melanotaenia splendida inornata (rainbow fish) and to the cladoceran Moinodaphnia macleayi. The test medium was Magela Creek (NT) water supplemented with sodium bicarbonate, magnesium sulfate, potassium chloride, and calcium chloride to achieve comparable cation and anion concentrations, pH and alkalinity with Fly River water. The results of these tests are summarised in Table 1. The cladoceran M. macleayi was the most sensitive species tested, with an LC50 comparable to that reported for other cladocerans, however, this species is not known to occur in the Fly River system. Juvenile barramundi were the most sensitive fish species to dissolved copper concentrations (LC50 of 79 µg L-1), but the juvenile forms used for these tests normally live in brackish or saline waters and may have been additionally stressed by the freshwater test medium used (Smith 1997). No adverse effects of particulate copper (16000 µg g-1) were observed for any of these tests. Table 1 Acute toxicity of dissolved and particulate copper - Ranger Uranium Mine study 1991 (from Smith 1997). Species

Test Type

Duration

LC50

NOEC

-1

(µg L ) Lates calcarifer

Dissolved Cu

96 h

Particulate Cu Neosiluris ater

Dissolved Cu

96 h

Particulate Cu Morgunda morgurnda

Dissolved Cu

96 h

Particulate Cu Melanotaenia splendida

Dissolved Cu

96 h

Particulate Cu Moinodaphnia macleayi

Dissolved Cu

24 h

Particulate Cu

79

15 µg L-1

-

16000 µg g

466

277 µg L

-

16000 µg g

189

72 µg L

-

16000 µg g

154

72 µg L

-

16000 µg g

27

-

-

16000 µg g

-1

-1 -1

-1 -1

-1 -1

-1

Recent studies of acute copper toxicity to the native barramundi species of the Fly River reported 96-h LC50 values between 410 µg L-1 (initial measured concentration) and 270 µg L-1 (final measured concentrations) dissolved copper at pH 7.0 - 7.6 and 20.5 mg L-1 CaCO3 (Australian Water Technologies 2002). The dilution water used for the test medium was carbon-filtered Melbourne mains water, which would not have been representative of the natural water in the Fly River system. The LC50 values reported for this study were considerably higher than those previously reported by Ranger Uranium Mine (Smith 1991) where the

4

mean 96-h LC50 value was 79 µg L-1. The difference in sensitivity to copper was most likely due to the life stage of the fish used in the testing procedures (juvenile fish with a mean length of 306 ± 21 mm in the Australian Water Technologies test would be less sensitive compared with 14-day old post-larval stages with mean length 24 ± 2 mm in the Ranger test). Differences in the source and husbandry of the fish would further confound the comparison. It should be noted that the data on copper toxicity to fish are limited to larger fish. There are no data available for planktivores such as Nematalosa spp.

2.5

Microalgal Studies Microalgae are particularly important in tropical aquatic ecosystems, being responsible for most of the primary production at the base of the food chain. Stauber and Apte (1996) reported a screening taxonomic study of microalgal presence and abundance in off-river water bodies connected to the Fly and Strickland Rivers. They collected tropical algae from eight sites in the Fly and Strickland River systems in January 1995. Two sites, Lake Daviumbu and Lake Pangua in the Fly River floodplain, were potentially impacted by copper, while six control sites in the Fly River oxbows near Kiunga or near the Strickland River (Lake Aesake and Oxbow Levamme) were not copper impacted. Good species diversity was found in all lakes at the time of this survey. Over sixty algal isolates were identified to genus level, with the dominant isolates at all sites being green algae (Chlorophyceae), diatoms, and blue-green algae (Cyanophyceae). There was also little difference in the genus composition or diversity between the impacted lakes in the Fly River system and the control lakes in the Strickland River system. Thirty-five genera of green algae were identified, of which nine were only present in non-impacted waters and two were present only in copper-impacted waters. The blue-green algae were distributed mainly in Lake Daviumbu, Lake Pangua and Lake Aesake and of the 14 genera of diatoms identified, only two were present exclusively in the Strickland River control sites. Dominant algal genera in phytoplankton net samples during this 1995 survey are shown in Table 2.

Table 2 Dominant algae in 20 µm and 63 µm net samples from the Fly and Strickland Rivers (from Stauber and Apte 1996)

5

Algae in Highest Abundance Site

Blue-Green algae

Diatoms

Green algae

-

Melosira Navicula Synedra

Scenedesmus

Dactyloccopsis Lyngbya

Navicula Synedra Diatoma

Scenedesmus Ankistrodesmus

Lyngbya Oscillatoria

-

-

Downstream Oxbow Kiunga

Dactyloccopsis

-

Phacus

UpstreamOxbow Kiunga

Dactyloccopsis

Navicula Synedra

-

-

-

Chlorella

Impacted Lake Pangua Non-Impacted Lake Aesake

Oxbow Levamme

Oxbow 4 Kiunga

Algal abundance was also determined by cell counts in bottle-collected samples. All samples from both impacted and non-impacted sites contained large numbers and diversity of blue-green algae, diatoms, dinoflagellates, green algae and one cryptomonad (1-640 cells mL-1). Stauber and Apte (1996) isolated twenty-one strains of microalgae, comprising 9 species, from control sites in the Strickland River system. Most of these were green algae from Lake Aesake, together with several diatoms. Unialgal cultures were established, identified and 9 isolates were also tested for copper sensitivity in screening bioassays. All species were sensitive to copper, with 72-h EC50 values ranging from 7-17 µg Cu L-1. Earlier bioassays to investigate the potential toxicity of copper in the waters of the Fly River used a temperate green algal species Chlorella protothecoides (CSIRO Division of Fisheries Culture Collection) (Stauber and Critelli 1993). These tests compared the growth of C. protothecoides in pristine laboratory water controls, matched to the same hardness, pH, nitrate and phosphate concentrations as Fly River water, to the growth of this alga in Fly River water at pH 7.9. Eight river water samples from Ok Tedi, Ok Mani and the Fly River, together with two off-river water bodies (Bosset Lagoon and Lake Daviumbu) were tested. No toxicity was observed for any of the 10 water samples from sites on the Fly River, despite the fact that they contained between 0.5 and 13 µg L-1 of dissolved copper, well above the lowest observed effect concentration (LOEC) of 2.5 µg Cu L-1 for this alga. Copper speciation measurements were not made on these early samples. The results of this study are summarised in Appendix 1. A copper-sensitive green alga, Chlorella sp. 12, was subsequently isolated from water samples taken from Lake Aesake on the Strickland River in 1995 and used in the development of a site-specific bioassay. A comparative study (Stauber and

6

Apte 1996) used C. protothecoides, Chlorella sp. 12, a bacterial bioassay (Apte et al. 1995, Davies et al. 1998), and anodic stripping voltammetry (ASV) to measure the complexation capacity of water samples from seven sites on the Fly River, which had been collected over the preceding twelve month period. All of the samples had copper complexation capacities (4-33 µg L-1) in excess of their respective dissolved copper concentrations (1.8-17 µg L-1), and again no toxicity was observed at any of the sites (Appendix 2). In terms of copper toxicology, microalgae represent the most studied group of aquatic organisms in the Fly River. It is clear that indigenous algal species are sensitive to relatively low concentrations of bioavailable copper.

7

3 ROUTINE BIOASSAY AND COPPER SPECIATION DATA 3.1

Algal and Bacterial Growth Inhibition Bioassays A number of studies have been conducted over the last 10 to 15 years at Ok Tedi Mining Limited (OTML) to investigate the speciation and toxicity of copper to aquatic organisms in the aquatic environments downstream of the Ok Tedi mine. These investigations have included the development and application of coppersensitive algal and bacterial bioassays (Stauber and Crittelli 1993, Stauber and Apte 1996, Apte et al.1997) and OTML HERA studies, (Parametrix 1999). Bioassays, or toxicity tests, are generic tests that use living organisms as indicators of contaminant bioavailability in aquatic systems. While chemical measurement techniques and geochemical speciation modelling may detect or predict the different forms of copper in aquatic systems, they do not provide direct data on adverse biological effects. Most microbial bioassays are acute (shortterm) tests and typically measure organism survival over 96 h or a sub-lethal effect such as respiration, bioluminescence or enzyme inhibition. Chronic tests, such as inhibition of growth of microalgae, determine toxicity over several generations of cells. Such tests may be of extended duration (weeks) or short term in the case of single-celled algae (that divide approximately once per day). Because different organisms have different sensitivities to toxicants such as copper, batteries of toxicity tests, using sensitive species from different trophic levels, are used. Total dissolved copper concentrations and copper speciation data, in conjunction with ecotoxicological bioassay data, have been collected by OTML and CSIRO since 1993. Early surveys in 1993 and 1995 used as the toxicity test species the temperate green algal species Chlorella protothecoides (CSIRO Division of Fisheries Culture Collection) and subsequently the tropical Lake Aesake algal isolate Chlorella sp 12, together with a bacterium isolated from the Upper Fly River (Isolate 37). No toxicity was observed at any sites in the Fly River, despite the fact that dissolved copper concentrations (0.5 - 17 µg L-1) exceeded the test species LOEC values for copper (Section 2 above). ` A more detailed study of copper speciation and copper toxicity was carried out over the period September 1996 - June 1997 using both electrochemical speciation techniques and copper-sensitive algal and bacterial growth bioassays (Apte et al. 1997). These seven surveys utilised both the Lake Aesake algal isolate Chlorella sp. 12 (Stauber and Apte 1996) and the bacterium Isolate 37, isolated from the Upper Fly River (Apte et al. 1995, Davies et al. 1998). The most sensitive bioassays available were used to provide an ‘early warning’ of potential copper toxicity. The EC50 values were 7.1 and 7.5 µg L-1 for the bacterial and algal bioassays respectively. A total of 48 river water samples were analysed for total dissolved copper, labile copper, copper complexation capacity and algal and bacterial growth inhibition. Total dissolved copper concentrations ranged from 5.4 to 42.6 µg L-1 at sites downstream from the mine and from 1.0 to 3.8 µg L-1 at the riverine control site Kiunga. Labile copper was detected in 26 of the 48 samples, but not at the control sites, and ranged from 0.5 to 8.3 µg L-1. In eight of the samples, labile copper was detected even though the copper complexation capacity had not been exceeded. The authors suggested that this was due to the

8

presence of weakly-bound copper-organic complexes which dissociated during the ASV analysis. Algal and/or bacterial growth inhibition was observed at 4 out of the 7 sites sampled (at Ningerum 6 times and, D’Albertis Junction, Nukumba and the Fly at Bosset each on one occasion) (Appendix 3). Samples from Ningerum were consistently toxic on 6 out of 8 occasions. For both the algal and bacterial bioassays, dissolved copper in the river water samples was less toxic than for a calibration bioassay carried out in laboratory water, illustrating the role of the natural organic matter present in the natural waters in ameliorating copper toxicity. There was no general relationship between growth inhibition and labile copper, but inhibition was only observed when labile copper was present. These data could be split into two populations: 1. Where no toxicity was observed in the presence of labile copper. For these samples, labile copper concentrations were generally below the lowest observable effect concentration (LOEC) for inorganic copper in synthetic water (about 5 µg Cu L-1) for Chlorella sp 12, i.e. there was insufficient labile copper to cause any growth inhibition. 2. Where there was a relationship between increasing labile copper concentrations and toxicity. In these samples, the concentrations of copper to cause growth inhibition were typically 5 µg Cu L-1 lower than the concentrations in the inorganic copper calibration bioassays. The most likely explanation for this was the presence of easily dissociable copperorganic complexes, or the presence of lipid-soluble copper complexes which have a greater toxicity than ionic copper (Florence and Stauber, 1986). Alternatively, other toxic metals may be present in the sample, i.e. the growth inhibition observed was not solely due to copper. In the final survey conducted in June 1997, consistent trends in copper speciation and algal and bacterial growth inhibition were observed. Labile copper was detected at three sites closest to the mine input (Ningerum, D’Albertis Junction and Nukumba) and both bioassays revealed growth inhibition at these sites. The linear relationship between algal growth inhibition and labile copper concentration (m = 12.8 ± 2.4, r2 = 0.967) was similar to that for ionic copper in the calibration bioassay (m = 16.3± 0.2, r2 = 0.999). Less labile copper was needed to elicit a similar degree of growth inhibition, the labile copper curve being displaced from the ionic copper calibration curve by about 5 µg L-1. Again, this suggested that a fraction of the copper-organic complexes were bioavailable to the algae. Finally, copper uptake experiments were performed using three of the samples. Cellular copper concentrations increased with increasing amounts of labile copper in the samples, and both the algae and bacteria showed the greatest amount of copper uptake from the Ningerum sample.

3.2

Routine Speciation Surveys Since 2002, CSIRO has been commissioned by OTML to undertake a program of routine copper speciation measurements and bioassays with waters taken quarterly from the Fly River System. The program utilises highly sensitive bioassays of copper toxicity in order to provide an ‘early warning’ of potential metal

9

toxicity problems that may affect resident organisms in the Fly River. The wisdom of this approach was justified in the early studies of copper toxicity in the Fly River (Stauber and Critelli 1993), which clearly demonstrated low toxicity to a highly sensitive algal bioassay despite the presence of dissolved copper concentrations which were above regulatory limits applying in most countries. These studies have used the same alga, Chlorella sp. 12, as used in the earlier studies, and a temperate bacterial isolate Erwinia persicinius that is extremely sensitive to copper (Rogers et al. 2005). The bacterial strain, Isolate 37, that was used in the previous study and which was indigenous to the Fly River system, was no longer viable for use in these studies owing to changes in copper sensitivity of the laboratory cultures maintained at CSIRO. The methodology for the algal bioassay was described by Apte et al. (1997) but extensive changes were made to the bacterial bioassay protocol to that used in the 1996-1997 surveys. A radiochemical assay was utilised which measured the inhibition of bacterial respiration rather than growth inhibition (Rogers et al. 2005). The radiochemical bacterial bioassay has the advantage of not requiring the addition of growth stimulating nutrients to the test waters and employs a very short test duration. These factors minimise the potential for speciation changes in the test waters during the course of the test and thus enable more representative environmental conditions to be maintained (Davies et al. 1998, Rogers et al. 2005). Matrix-matched controls were prepared by the addition of the chelating agent EDTA to one subsample of each of the test waters (Apte et al. 1997). EDTA complexes metals such as copper, rendering them non-toxic. Microalgal growth or bacterial respiration in the EDTA-amended samples was compared to growth/respiration in the non-amended samples. In this way, each water sample served as its own control. These sensitive bioassays were used in an attempt to provide an early warning of potential copper toxicity. Since early 2004, the regime monitoring surveys have analysed a total of 53 water samples to date (Appendix 4). Of these, toxicity has been detected by either or both the algal or bacterial bioassays in 39 samples (74%). This suggests that the number of potentially toxic samples has increased compared to the 1996-1997 surveys. Nine different sites were tested and toxicity was observed at 7 sites. In general, toxicity was observed consistently at sites on the Upper Fly River and only intermittently at sites on the lower river or at off-river water bodies. No toxicity was observed at the control site Kiunga (2 samples) or at Oxbow 4 (2 samples). The sites and frequency with which toxicity was observed are shown in Table 3. Good agreement was observed between the algal and bacterial bioassays during this period, with differences occurring only for 7 of the 53 samples. Although there was again no relationship between dissolved copper or labile copper and toxicity, labile copper was generally present when toxicity was observed. The only exception to this was at Oxbow 5. This site exhibited only slight toxicity to the algal bioassay (4% inhibition) in February 2004, which is atypical of the response previously recorded at off-river bodies. Although 4% inhibition may be statistically significant due to high test precision, it may not be biologically meaningful.

10

Table 3 Sites impacted by copper toxicity during the 2002–2004 routine speciation surveys Site

3.2.1

Total number of analyses

Number of times toxic

Number of times labile copper present when toxic

Ningerum

10

10

10

Nukumba

10

10

10

Obo

10

10

10

Ogwa

4

4

4

Burei Junction

3

2

2

Lewada

9

2

2

Oxbow 5

3

1

0

Algal bioassay data No clear relationship was discernable between dissolved copper and algal growth inhibition (Figure 1a). This plot also includes the algal bioassay response to inorganic copper in laboratory synthetic freshwater, represented by the inorganic copper calibration curve (Apte et al 1997). As in the previous surveys, the concentration of copper in the test waters required to elicit an inhibitory response was much higher than for inorganic copper (i.e. to the right of the inorganic copper calibration curve). Eight test waters exhibited toxicity in the region of the calibration curve suggesting that there was little ameliorative effect of dissolved organic matter in these eight waters. The relationship between algal growth inhibition and labile copper was measured by two different methods, and is shown in Figure 1b (by ASV) and Figure 1c (by Chelex column) (Bowles et al., 2005 submitted). The results of these two analyses are broadly similar, but the Chelex column data provide a marginally better correlation (Table 4) with algal growth inhibition (R = -0.605) than the ASV data (R = -0.574). In each case, four regions are discernable. Firstly, concentrations of labile copper in the samples which were not inhibitory to algal growth were generally at or below 6 µg L-1, the threshold lowest observable effect concentration for ionic copper for this bioassay. Therefore insufficient labile copper was present to cause an inhibitory effect. Secondly, some of the waters tested exhibited toxicity in the region of the inorganic copper calibration curve suggesting that the measured labile copper was responsible for the observed toxicity. The rest, and majority, of the data fall either to the right of the calibration curve indicating that the measured labile copper was less toxic than predicted, or to the left of the curve indicating greater than predicted toxicity. This may be due to the presence of additional metals or other toxicants in the water samples. These may be

11

independently toxic, or have an additive, antagonistic or synergistic effect on copper toxicity to Chlorella sp. 12. 3.2.2

Bacterial bioassay data Increasing dissolved copper concentrations generally increased inhibition of the bacterial bioassay (Figure 2a). This plot also includes the bacterial bioassay response to inorganic copper in laboratory synthetic freshwater, represented by the inorganic copper calibration curve (Rogers et al 2005). In many of the test waters, no inhibitory response was observed despite the presence of dissolved copper concentrations up to 10 µg L-1, illustrating the role of dissolved natural organic matter in ameliorating bacterial copper toxicity. As for the algal bioassay, however, the concentration of copper in the test waters required to elicit an inhibitory response was much higher than for ionic copper (i.e. to the right of the inorganic calibration curve). The relationship between bacterial inhibition and labile copper is shown in Figure 2b (by ASV) and Figure 2c (by Chelex column). Increasing concentrations of labile copper measured by either ASV or the Chelex column method are well correlated (Table 4) with inhibition of the bacterial bioassay (R = -0.703 and R = -0.806 respectively). Despite the presence of ASV-labile copper concentrations between 1.3 and 3.5 µg L-1 in six of the test waters, no bacterial toxicity was observed. This may indicate a tendency of the ASV technique to measure higher labile concentrations than actually present; the corresponding Chelex-labile copper concentrations for four of these waters ranged from 0.8 -1.9 µg L-1. The data for Chelex-labile copper are displaced to the right of the inorganic calibration curve by about 1 µg L-1 indicating that these samples were only slightly less toxic than that predicted by the ionic copper calibration curve. This suggests that some of the easily dissociable complexes measured by the Chelex method are not bioavailable to the bacteria. A proportion of the copper present as these complexes may reassociate with organic matter in the test waters rather than reacting with the bacterial cell membrane over the duration of the bioassay.

3.2.3

Causes of the observed toxicity While the bacterial and algal species used for biomonitoring were selected on the grounds of their sensitivity to copper, the effects of other toxicants which may be present in the Fly River cannot be neglected. Addition of the metal chelating agent EDTA to water samples consistently removed the observed toxicity in both the algal and bacterial bioassays. This is important information as it indicates that the toxic effects are definitely related to the presence of metal contaminants. The combined toxic effect of several metals or antagonistic interactions between metals cannot at this stage be ruled out.

12

(a) Algal Response (% EDTA amended control)

120 100 80 60 40 20 0 0

5

10

15

20

25

30

35

Dissolved copper µg L-1

(b) Algal Response (% EDTA amended control)

120 100 80 60 40 20 0 0

2

4

6

8

10

12

-1

ASV-labile copper µg L

(c) Algal Response (% EDTA amended control)

120 100 80 60 40 20 0 0

2

4

6

8

10

12

Chelex-labile copper µg L-1

Figure 1 Relationship between algal growth inhibition and (a) dissolved copper, (b) ASVlabile copper and (c) Chelex-labile copper. (▲) no inhibition, (■) inhibition. The solid line represents the inorganic copper calibration curve in synthetic water (Apte et al. 1997).

13

Bacterial Response (% EDTA amended control)

120

(a)

100 80 60 40 20 0 0

5

10

15

20

25

30

35

-1

Dissolved copper µg L

120 Bacterial R esponse (% ED TA amended control)

(b)

100 80 60 40 20 0 0

2

4

6

ASV-labile copper µg L

(c)

8

10

12

10

12

-1

Bacterial Response (% ED TA amended control)

120 100 80 60 40 20 0 0

2

4

6

8

Chelex-labile copper µg L-1

Figure 2 Relationship between inhibition of bacterial respiration and (a) dissolved copper, (b) ASV-labile copper and (c) Chelex-labile copper. (▲) no inhibition, (■) inhibition, The solid line represents the inorganic copper calibration curve in synthetic water (Rogers et al. 2005).

14

Table 4 Correlations between copper speciation measurements and algal and bacterial bioassays (statistically significant correlations, p 80% of cladocerans) and 40% of algal species may not be protected from chronic copper toxicity. The chronic toxicity curve for tropical fish species was very steep and lacked the sigmoid shape of the acute toxicity curve, however approximately 50% of tropical fish species may potentially be chronically affected by labile copper concentrations around 10 µg L-1. Insects were in general less sensitive and cladocerans were more sensitive to copper than fish. Chronic toxicity studies are extremely difficult, time consuming and expensive to conduct, especially on higher organisms such as fish. Additionally, the required duration for chronic toxicity studies is poorly understood. The

35

ANZECC/ARMCANZ (2000) guidelines considered mortality measured over exposure periods of 7 days or longer to be an acceptable chronic endpoint. Whilst this may be appropriate for some invertebrate species it clearly represents a very minor proportion of the life cycle of a large fish species. In the case of large fish species, chronic toxicity studies lasting several months may still be unrepresentative of chronic toxicity. An exception to this may be EC50 values derived for single-celled organisms such as algae or bacteria. In these organisms a period of 48-72 h represents multigenerational exposure and hence may be considered a chronic study. A number of ≥ 72 -h copper EC50 values were used in the 1999 HERA (Parametrix 1999) or have been recently published for algae (Appendix 5). The recent data include values for algal copper toxicity generated by Stauber and co-workers using a number of sensitive algal species including Chlorella sp 12 (Franklin et al. 2001a, 2002a) isolated from the Strickland River. These EC50 values were used to derive a chronic algal species sensitivity distribution which could be used to predict copper concentrations at which effects would occur. This chronic algal distribution is shown in Figure 14. Over the range of labile copper concentrations typically found in the Ok Tedi and Middle Fly (4-9 µg L-1), the distribution curve indicates that between 1-25% of algal species may be affected by chronic copper toxicity. The toxicity data presented for cladocerans (Table 6) indicate that these organisms are also likely to be significantly affected by chronic copper toxicity over the concentration range labile copper found in the Ok Tedi and Middle Fly. 5.2.1 Uncertainties in estimating the toxic effects of dissolved copper There are many uncertainties in the analysis of the copper toxicity data presented here. A semi-quantitative analysis of errors is presented below. Use of NOEC data - Water quality guidelines and chronic species sensitivity distributions based on NOEC values are useful in deriving values for toxicants below which it is predicted no effects on the ecosystem will be observed. This approach has been criticised as being overly conservative. The obtained NOEC values highly dependent on both toxicity test design and precision (Chapman 1996). A more useful measure of chronic toxicity would be an LC/EC50 value (or EC15, EC25 etc. depending on the level of protection required). However, very few chronic studies which supply these data are available from the literature. Species sensitivity distributions - The variability in laboratory toxicity tests from which species sensitivity distributions are derived are typically 20-50% relative standard deviation. Confidence limits for species sensitivity distribution curves may be estimated by advanced statistical procedures (e.g. Bossuyt et al. 2005, Verdonck et al. 2001, 2003, Duboudin et al. 2004). The computation of such confidence limits was beyond the scope of this review as the computational and statistical resources were not available. Measurement of copper bioavailability - As noted earlier, labile copper is a surrogate measurement of bioavailable copper. For some aquatic organisms, it is possible that labile copper is an overestimate of the biologically-available fraction. Labile copper is unlikely to be an underestimate of biologically-available copper.

36

Extrapolation from laboratory to the field -The use of laboratory-derived toxicity data and laboratory cultured test species may overestimate risk. This is because laboratory organisms may be more sensitive to contaminant exposure than field populations that may have acclimated or evolved tolerances to various metals. Sampling Regime – the quarterly sampling regime is likely to underestimate maximum labile copper concentrations. As noted previously, the highest concentrations of bioavailable copper are likely to occur in first flush events following extended dry spells. The labile copper data from the monitoring campaigns are presented as a cumulative plot in Figure 10 and 11. Using these data in combination with the cumulative effect distributions for copper toxicity presented earlier, it is possible to statistically assess the percentage of species affected under different scenarios and also build in uncertainty estimates. It is recommended that this approach is adopted in future ecological risk assessments.

37

100

Cumulative Frequency (%)

90 80 70 60 50 40 30 20 10 0 1

10

100

Dissolved Copper (µg L-1)

Figure 12 Chronic copper sensitivities for freshwater species at water hardness 30 mg L-1 CaCO3 (ANZECC/ARMCANZ 2000)

Figure 13 Chronic copper sensitivities for freshwater species at water hardness 50 mg L-1 CaCO3 (Parametrix 1999)

38

100

Percent species affected

90 80 70 60 50 40 30 20 10 0 0.1

1

10

100

1000

Dissolved Copper µg L-1

Figure 14 Cumulative frequency distribution of species sensitivities to copper for freshwater algae using chronic EC50 values. Hardness normalised to 50 mg L-1 CaCO3.

Table 7 Maximum and mean measured labile copper concentrations at sites on the Fly River

Maximum Site

Ningerum

Nukumba

Obo

Ogwa

Labile Cu µgL

Mean Labile -1

Cu (SD) -1

(Date)

µg L

9.2

4.0 (2.3)

(April 2002)

n =17

9.2

4.1 (2.4)

(Feb 02)

n =15

8.2

3.6 (2.4)

(Feb 02)

n = 17

4.2

1.4 (1.2)

(Dec 03)

n = 11

39

6 IMPACT OF MINING ACTIVITIES ON FOOD WEBS IN THE FLY RIVER The previous sections of this report have focussed on the direct toxicity of copper to living organisms such as algae, invertebrates and fish. As noted earlier, unicellular algae have a particular sensitivity to dissolved copper. Copper toxicity may also affect the viability of the Fly River fishery by affecting carbon flow through food webs. For instance labile copper concentrations may be inhibitory to primary producers and ‘starve’ higher trophic levels of carbon. Other effects such as loss of keystone species may also be manifest. In order to better understand these effects, it is necessary to characterise the aquatic food web of the river system. The importance of algal carbon sources and the role of copper toxicity on food web carbon flow, were recognised as important data gaps during the OTML HERA. As a result, an initial study was conducted by Bunn et al in 1999 and focussed on elucidating the food web for the Upper Fly River at Kawok (above D’Albertis Junction), and a forested oxbow at ARM 345 (below D’Albertis Junction). This study showed that microalgae are of considerable importance as a component of the food web in riverine and floodplain habitats and of great significance to the fishery. It was conservatively estimated that carbon derived from algae supported at least one third of the total fish biomass within the river channel and up to two thirds of fish biomass on the floodplain. It was noted that in the mine-impacted reaches of the river, algal carbon productivity might be affected by factors such as high turbidity, smothering by sediment and concentrations of labile copper that are inhibitory to growth. A recent follow up study conducted by OTML (Storey et al. in preparation) focussed on comparing the food webs at sites on the Fly River above and below D’Albertis Junction (Kiunga- Kawok and Kuambit respectively). A number of biological specimens were collected at each site and analysed for stable carbon and nitrogen isotope ratios. Two-point mixing models were utilised on data from each site, using the riparian and algal delta C (carbon) signatures at each site as end points to calculate the percent of algal carbon in each individual consumer. This effectively puts the individual consumers at each site on the same scale, allowing direct statistical comparison between sites. The main objective of the study was to compare and test for differences in the contribution of algal carbon to the food webs at both locations. Preliminary findings indicate that there are two main obvious primary sources of carbon at each location, being riparian vegetation and algae (in the absence of obvious periphyton layers, presumed to be mostly filamentous algae growing on the banks and on mud deposits). There were marked differences in algal abundance between sites. Collection of algae at Kuambit, the mine-impacted site, was problematic as there was little to collect, unlike Kiunga-Kawok, where it was plentiful.

40

Some marked differences were found in algal carbon signatures in fish between sites. Individual fish at Kuambit predominantly had a riparian signature (many had 100% riparian carbon) and the opposite applied at Kiunga-Kawok, with many fish with a predominantly algal carbon signature Figure 15). It may be postulated that those fish at Kuambit are there because they can survive on riparian inputs or on prawns (which depend on riparian carbon), whilst species at Kiunga-Kawok can access algal carbon, which is reflected in their diet. Unfortunately few aquatic invertebrates could be collected from either site. Therefore, no meaningful comparison of aquatic invertebrates was possible. It is possible that the KiungaKawok species cannot survive at Kuambit because they are more specialised in their dietary requirements and there are insufficient algae to support their food web. Although it is not possible to determine the reasons for a reduction in algal signature at Kuambit (toxicity from dissolved copper, smothering by aggradation, abrasion by TSS or reduced light penetration), it certainly appears that the food web at Kuambit has shifted away from algal to riparian carbon. The sensitivity of periphytonic algal species at these sites needs to be established in order to deconvolute the effects of these superimposed stressors.

12

del 15 N

10

8

6

4 0

0.2

0.4

0.6

0.8

P roportion of c arbon in higher c ons um ers derived from algae Fis h - A RM 450

Fis h - FLY10

P rawns - A RM 450

P rawns - FLY 10

Figure 15. Plot of higher consumers at Kiunga-Kawok (ARM450) versus Kuambit (FLY10) grouped by broad categories, showing delta N signature (trophic position) on the y-axis, and proportion of carbon derived from algae along the x-axis. “0” = no algal carbon and “1.0” = 100 % algal carbon. Each point corresponds to an individual animal (i.e fish/prawn). Fish with no algal carbon at Kawok have been off-set (+0.01) for clarity.

41

1

7 CONCLUSIONS AND RECOMENDATIONS 7.1

Conclusions 1. The speciation monitoring data indicate that both dissolved copper concentrations and copper complexing capacity in the Upper and Middle Fly River have remained relatively constant over the survey period (1996 - 2004). However, in the Middle Fly, labile copper concentrations (a surrogate for biologically-available copper) appear to be increasing. The reason for this increase is unknown, but may be due either to inputs of more reactive copper from a change in mining throughput and/or the dredge stockpiles at Bige, or to vegetation dieback on the flood-plain and consequent reduced organic matter input which influences copper speciation. 2. The frequency of algal growth inhibition has increased over the duration of the quarterly speciation surveys. At sites on the Middle Fly River, the increase in the frequency of algal growth inhibition corresponds to an increase in labile copper concentrations. The magnitude of algal inhibition has not significantly increased over the duration of the quarterly surveys and there is no significant correlation between the amount of algal inhibition and labile copper concentrations in the river. Addition of the metal chelating agent EDTA to water samples consistently removed the observed toxicity in both the algal and bacterial bioassays. This indicated that the toxic effects are definitely related to the presence of metal contaminants. The combined toxic effect of several metals or antagonistic interactions between metals cannot at this stage be ruled out. 3. The effects of labile copper concentrations on other freshwater aquatic organisms was assessed using best practice toxicity assessment procedures. Cumulative frequency distributions of the copper sensitivity of freshwater species (species sensitivity distributions) were compiled from high quality published toxicity data. These data compilations indicate that some bacteria, algae and cladocerans (e.g. water fleas) are the most sensitive organisms affected by copper toxicity. 4. The impacts of copper toxicity were analysed in terms of acute and chronic exposure to copper. At the labile copper concentrations measured during the regime monitoring program, acute copper toxicity is not an issue. At worst, acute toxicity may be observed in fewer than 10% of invertebrate species, with no acute effects on fish. However, at current labile copper concentrations, chronic effects of copper resulting from long-term exposure to elevated bioavailable copper concentrations may be expected in 50 to 80% of freshwater species. Any further increases in labile copper in the river may result in chronic toxicity to the majority of freshwater species. The uncertainties in these calculations and conservative nature of this ‘best practice’ assessment are recognised and discussed in the main body of the report.

42

5. Food web stable isotope studies clearly indicate the importance of microalgae, such as periphyton, in the food webs of the Fly River system. Recent work suggests that periphyton abundance is reduced in the river sections downstream of D’Albertis Junction, and the contribution of algal carbon to the aquatic food web in these sections is also reduced compared with areas upstream of the mine. The factors affecting periphyton abundance in the Fly River remain to be established. At this stage, it is not possible to deconvolute the effects of copper, turbidity (with respect to light production or scouring) and loss of habitat. Nevertheless, the concentrations of labile copper observed in the river system are in the range that may cause growth reduction in algae. Loss of algal carbon food sources and consequent effects on higher-level consumers, irrespective of the cause, is an important potential impact.

7.2

Recommendations 1. OTML should continue the monitoring surveys (labile copper, bacterial and algal bioassays) to further track the trends that have been observed to date. This is especially important at sites in the Middle and Lower Fly River, where labile copper and toxicity appear to be increasing. Increased monitoring of off-river water bodies may also be appropriate. The frequency and location of monitoring sites should be reviewed. A major consideration for any future monitoring program is the incorporation of a sampling frequency that adequately captures the appropriate temporal and spatial scales of variability. It is likely that the greatest copper toxicity is observed after events such as drying and wetting of the floodplain (e.g. the first rainfall after a prolonged dry period). It is highly recommended that event-based sampling is conducted in order to capture some of these events. 2. Given that food web studies showed that periphyton provided an important carbon source supporting aquatic food chains in the Fly River system, it is recommended that the effect of both increased bioavailable copper concentrations and increased turbidity on periphyton communities in the Fly River should be understood. As a first step, it is recommended that periphyton populations be collected from both the Middle Fly and Kiunga and species diversity and abundance, and sensitivity/tolerance to copper and turbidity, be compared. 3. There is a lack of basic ecotoxicological data on the effects of copper on primary consumers (algal grazers) such as small invertebrates and zooplankton, planktivorous fish and other key aquatic organisms in the Fly River system. This is an important knowledge gap. We recommend a program of laboratory-based toxicity testing to address this issue. If feasible, the use of in situ toxicity testing may also be useful. 4. Recent work at CSIRO has led to the development of a robust Chelex column method for measuring copper speciation in natural waters, which is superior to the previously utilised ASV-labile method. The Chelex method is relatively simple and does not require specialist equipment. It is recommended that this new

43

method be set up at OTML and used routinely to screen copper speciation. This method would be particularly useful for investigating copper speciation during high copper spike events in the river system.

44

8 REFERENCES ANZECC/ARMCANZ (2000). Australia and New Zealand Environment and Conservation Council Australian and New Zealand guidelines for fresh and marine water quality. Canberra Australia Albright LJ, Wentworth JW, Wilson EM (1972). Technique for measuring metallic salt effects upon the indigenous heterotrophic microflora of a natural water. Water Research, 6,1589-1596 Apte, S.C., Day, G.M. (1993). Organic complexation and partitioning of copper in river waters. A review. OK Tedi Mining Ltd./CSIRO Joint Report, 49 pages. Apte, S.C., Batley, G.E., Davies, C.M., Stauber, J.L. (1995). Assessing the fate, transport and toxicity of mine derived contaminants. Proceedings of the 20th Annual Environment Workshop Managing Environmental Impacts – Policy and Practice. Minerals Council of Australia, Darwin 2-6 October 1995, 377 – 384. Apte, S.C., Stauber, J.L., Davies, C.M., Hales, L.T. (1997). Bioavailability of dissolved copper in waters downstream of the Ok Tedi mine CSIRO Investigation Report CET/IR 27R prepared for Ok Tedi Mining Ltd., 58 pages. Australian Water Technologies (2002). Testing of the toxicity of copper to barramundi AWT Report Number: 1999/0059. Barry, K.L., Grout, J.A., Nidle, B.H., Piercey, G.E. (2000). . Impacts of acid mine drainage on juvenile salmonids in an estuary near Britannia Beach in Howe Sound, British Columbia. Canadian Journal of Fisheries and Aquatic Sciences, 57, 20322043. Blinova, I. (2004). Use of freshwater algae and duckweeds for phytotoxicity testing. Environmental Toxicology, 19, 425 – 428. Bossuyt, B.T.A., Janssen, C.R. (2003) . Acclimation of Daphnia magna to environmentally realistic copper concentrations. Comparative Biochemistry and Physiology C- Toxicology & Pharmacology, 136, 253 – 264. Bossuyt, B.T.A., Janssen, C.R. (2004). Long-term acclimation of Pseudokirchneriella subcapitata (Korshikov) Hindak to different copper concentrations: changes in tolerance and physiology. Aquatic Toxicology, 68, 6174.

45

Bossuyt, B.T.A., Muyssen, B.T.A., Janssen, C.R. (2005). Relevance of generic and site-specific species sensitivity distributions in the current risk assessment procedures for copper and zinc. Environmental Toxicology and Chemistry, 24, 470-478. Bowles, K.C., Apte, S.C., Batley, G.E., Rogers, N.J. (2005). A rapid chelex column method for the deteramination of metal speciation in natural waters. Analytical Chimica Acta submitted. Brix, K.V., DeForest, D.K., Adams, W.J. (2001). Assessing acute and chronic copper risks to freshwater aquatic life using species sensitivity distributions for different taxonomic groups. Environmental Toxicology and Chemistry, 20, 18461856. Bunn, S., Tenakanai, C., Storey, A.W. (1999). Energy sources supporting Fly River fish communities. Ok Tedi Mining Limited Report, 44 pages. Cairns, J., McCormick, P.V., Belanger, S.E. (1992). Ecotoxicological testing: small is reliable. Journal of Environmental Pathology Toxicology and Oncology, 11, 247-263. Chapman, P.M. (1996). A warning: NOECs are inappropriate for regulatory use. Environmental Toxicology and Chemistry, 15, 77-79. Clement, B., Zaid, S. (2004). A new protocol to measure the effects of toxicants on daphnid-algae interactions. Chemosphere, 55, 1429-1438. Conners, D.E,, Black, M.C. (2004). Evaluation of lethality and genotoxicity in the freshwater mussel Utterbackia imbecillis (Bivalvia : Unionidae) exposed singly and in combination to chemicals used in lawn care. Archives of Environmental Contamination and Toxicology, 64, 362-371. Davies, C.M., Apte, S.C., Johnstone, A.L. (1998). A bacterial bioassay for the assessment of copper bioavailability in freshwaters. Environmental Toxicology and Water Quality, 13, 263-271. de Oliveira-Filho, E.C., Lopes, R.M., Paumgartten, F.J.R. (2004). Comparative study on the susceptibility of freshwater species to copper-based pesticides. Chemosphere, 56, 369-374.

46

Duboidin C., Ciffroy, P., Magaud, H. (2004). Effects of data manipulation and statistical methods on species sensitivity distributions. Environmental Toxicology and Chemistry, 23, 489-499. Florence, T.M., Stauber, J.L. (1986). Toxicity of copper complexes to the marine diatom N. closterium. Aquatic Toxicology, 8, 11-26. Franklin, N.M,, Stauber, J.L., Markich, S.J., Lim, R.P. (2000). pH-dependent toxicity of copper and uranium to a tropical freshwater alga (Chlorella sp.). Aquatic Toxicology, 48, 275-289. Franklin, N.M., Stauber, J.L., Lim, R.P. (2001a). Development of flow cytometrybased algal bioassays for assessing toxicity of copper in natural waters. Environmental Toxicology and Chemistry, 20, 160-170. Franklin, N.M., Adams, M.S., Stauber, J.L., Lim, R.P. (2001b). Development of an improved rapid enzyme inhibition bioassay with marine and freshwater microalgae using flow cytometry. Archives of Environmental Contamination and Toxicology, 40, 469-480. Franklin, N.M., Stauber, J.L., Apte, S.C., Lim, R.P. (2002a). Effect of initial cell density on the bioavailability and toxicity of copper in microalgal bioassays. Environmental Toxicology and Chemistry, 21, 742-751. Franklin, N.M., Stauber, J.L., Lim, R.P., Petocz, P. (2002b). Toxicity of metal mixtures to a tropical freshwater alga (Chlorella sp): The effect of interactions between copper, cadmium, and zinc on metal cell binding and uptake. Environmental Toxicology and Chemistry, 21, 2412-2422. Franklin, N.M., Stauber, J.L., Lim, R.P. (2004). Development of multispecies algal bioassays using flow cytometry. Environmental Toxicology and Chemistry, 26, 1452-1462. Hyne, R.V., Pablo, F., Julli, M., Markich, S.J. (2005). Influence of water chemistry on the acute toxicity of copper and zinc to the cladoceran Ceriodaphnia cf dubia. Environmental Toxicology and Chemistry, 24, In press. Karntanut, W., Pascoe, D. (2000). A comparison of methods for measuring acute toxicity to Hydra vulgaris. Chemosphere, 41, 1543-1548.

47

Karntanut, W., Pascoe, D. (2002). The toxicity of copper, cadmium and zinc to four different Hydra (Cnidaria: Hydrozoa). Chemosphere, 47, 1059-1064. Khangarot, B.S., Rathore, R.S., Singh, B.B. (2003). pH-Dependent toxicity of heavy metals to a freshwater sludgeworm Tubifex tubifex Muller. Bulletin of Environmental Contamination and Toxicology, 71, 283-289. Lee, K., Tay, K.L., Ewing, C.N., Levy, E.M. (1990). Toxicity and environmental impact assessment tests based on the activity of indigenous bacteria. Ocean Dumping Report 4, Environment Canada, Atlantic Region, Darmouth, Nova Scotia. 138 p. Lombardi, J.V., Machado-Neto, J.G., Brossi-Garcia, A.L., Kubo, .E. (2000). Acute toxicity of the fungicide copper oxychloride to the freshwater prawn Macrobrachium rosenbergii De Man. Bulletin of Environmental Contamination and Toxicology, 65, 383-390. Long, K.E., Van Genderen, E.J., Klaine, S.J. (2004). The effects of low hardness and pH on copper toxicity to Daphnia magna. Environmental Toxicology and Chemistry, 23, 72-75. Ma, M., Zhu, W.Z., Wang, Z.J., Witkamp, G.J. (2003). Accumulation, assimilation and growth inhibition of copper on freshwater alga (Scenedesmus subspicatus 86.81 SAG) in the presence of EDTA and fulvic acid. Aquatic Toxicology, 63, 221228. Markich, S.J., Brown, P.L., Jeffree, R.A., Lim, R.P. (2003). The effects of pH and dissolved organic carbon on the toxicity of cadmium and copper to a freshwater bivalve: Further support for the extended free ion activity model. Archives of Environmental Contamination and Toxicology, 45, 479-491. Markich, S.J., Batley, G.E., Stauber J.L., Rogers N.J., Apte, S.C., Hyne, R.V., Bowles, K.C., Wilde, K.L., Creighton, N. (2005). Hardness corrections for copper are inappropriate for protecting sensitive freshwater biota. Chemosphere, In press. Mastin, B.J., Rodgers, J.H. (2000). Toxicity and bioavailability of copper herbicides (clearigate, cutrine-plus, and copper sulfate) to freshwater animals. Archives of Environmental Contamination and Toxicology, 39, 455-451.

48

OECD (1992). Report of the OECD workshop on extrapolation of laboratory aquatic toxicity data to the real environment OECD Environment Monograph No 59, Organisation for Economic Cooperation and Development, Paris. Parametrix Inc., URS Greiner Woodward Clyde. (1999). Assessment of human health and ecological risks for proposed mine waste mitigation options at the Ok Tedi mine, Papua New Guinea. Detailed Level Risk Assessment, Final Report. Prepared for Ok Tedi Mining Ltd., 214 pages. Pechenik, J.A. (1996). Biology of the Invertebrates 3rd Edition McGraw –Hill, Boston, Massachusetts. Perez-Legaspi, I.A., Rico-Martinez, R. (2001). Acute toxicity tests on three species of the genus Lecane (Rotifera : Monogononta). Hydrobiologia, 446, 375381. Preston, B.L., Snell, T.W. (2001). Full life-cycle toxicity assessment using rotifer resting egg production: implications for ecological risk assessment. Environmental Pollution, 144, 399-406. Ranger Uranium Mines Pty Ltd. (1991). Report on the toxicity of dissolved copper and particulate copper in synthetic Fly River water. Ok Tedi Mining Ltd, Tabubil. Rathore, R.S., Khangarot, B.S. (2002). Effects of temperature on the sensitivity of sludge worm Tubifex tubifex Muller to selected heavy metals. Ecotoxicology and Environmental Safety, 53, 27-36. Rathore, R.S., Khangarot, B.S. (2003). Effects of water hardness and metal concentration on a freshwater Tubifex tubifex Muller. Water Air and Soil Pollution, 142, 341-356. Rogers, N.J., Apte, S.C., Knapik, A., Davies, C.M., Bowles, K.C., Kable, S.H. (2005). A rapid, radiochemical bacterial bioassay to evaluate copper toxicity in freshwaters. Archives of Environmental Contamination and Toxicology (submitted). Smith, R.E.W., Ahsanullah, M., Batley, G.E. (1990). Investigations of the impact of effluent from the Ok Tedi copper mine on the fisheries resource of the Fly River, Papua New Guinea. Environmental Monitoring and Assessment, 14, 315-331.

49

Smith, R.E.W. (1997). Review of laboratory based ecotoxicological testing of Ok Tedi Mining Limited Wastes. R&D Environmental Pty Ltd., prepared for Ok Tedi Mining Limited, 18 pages. Shao, Q., Harch, B., Apte, S.,Simpson, S., (2002). Temporal variability of dissolved copper concentrations in the Ok Tedi and Fly River, CSIRO Investigation Report CET/LH/IR 501R prepared for OK Tedi Mining Ltd., 56 pages. Soto, E., Oyarce, G., Inzunza, B., Bay-Schmith, E. (2003). Acute toxicity of organic and inorganic compounds on the freshwater cyclopoid copepod Eucyclops neumani neumani (Pesta, 1927). Bulletin of Environmental Contamination and Toxicology, 70, 1017-1021. Stauber, J.L., Critelli, C. (1993). Toxicity of waters from the Fly River system to freshwater and estuarine algae. CSIRO Investigation Report CET/IR 187 prepared for OK Tedi Mining Ltd., 20 pages. Stauber, J.L., Apte, S.C. (1996). Bioavailability of copper to algae in the Fly River system. CSIRO Investigation Report CET/LH/IR 464R prepared for OK Tedi Mining Ltd., 78 pages. Stauber, J.L., Davies, C.M. (2000). Use and limitation of microbial bioassays for assessing copper bioavailability in the aquatic environment. Environmental Reviews, 8, 255-301. Stephan, C.E., Mount, D.I., Hansen, D.J., Gentile, J.H., Chapman, G.A., Brungs, W.A. (1985). Guidelines for deriving national water quality criteria for the protection of Aquatic organisms and their uses. PB85-227049, National and Technical Service Springfield, Virginia, USA. Sunda WG. Ferguson RL (1983). Sensitivity of natural bacterial communities to additions of copper and to cupric ion activity: a bioassay of copper complexation in seawater. In: Wong CS, Boyle E, Bruland KW, Burton JD, Goldberg ED, (eds) Trace Metals in Seawater. Plenum Press, New York, p 871

USEPA (1986). Quality Criteria for Water. US Department for Commerce National Technical Information Service, US Environmental Protection Agency, Springfield, Virginia PB87-226759. Verdonck, F.A.M., Jaworska, J., Thas, O., Vanrolleghem, P.A. (2001).. Limitations of current risk characterisation methods in probabilistic environmental risk assessment. Analytica Chimica Acta, 446, 429-438.

50

Verdonck, F.A.M., Aldenburg, T., Jaworska, J., Vanrolleghem, P.A. (2003). Limitations of current risk characterisation methods in probabilistic environmental risk assessment. Environmental Toxicology and Chemistry, 22, 2209-2213. Wilde, K.L., Stauber, J.L., Markich, S.J., Franklin, N.M. Brown, P.L. (2005). The effects of pH on the uptake and toxicity of copper and zinc in a tropical freshwater alga (Chlorella sp.). Archives of Environmental Contamination and Toxicology (submitted).

51

Appendix 1

Algal Bioassays on the Fly River System (Stauber and Critelli, 1993) Test Species: Chlorella protothecoides (CSIRO Division of Fisheries Culture Collection) Dissolved copper Survey Date

Sample

pH

µg L-1

Algal Growth (% of Control)

Algal Bioassay

1993

Konkonda

8.64

12.6

105

No Inhibition

1993

Nukumba

8.35

9

129

No Inhibition

1993

Fly Bosset

7.4

10.4

95

No Inhibition

1993

Obo

7.21

8.1

97

No Inhibition

1993

Ogwa

8.21

3

92

No Inhibition

1993

Bosset Lagoon

8.39

1.3

109

No Inhibition

1993

Lake Daviumbu

9.34

1.5

119

No Inhibition

1993

Ok Mani

8.08

0.5

157

No Inhibition

1993

Kiunga

7.86

0.7

94

No Inhibition

1993

Strickland

8.73

0.7

109

No Inhibition

No samples were directly toxic to Chlorella protothecoides using a synthetic water control pH 7.9, matched to the same pH, hardness, nitrate and phosphate concentration as the samples.

52

Appendix 2 Complexation Capacities of Fly River (Stauber and Apte 1996) Test Species: Chlorella protothecoides, Chlorella sp 12 (PNG isolate) and bacterial isolate (I37) (EC15 endpoints).

Date

Sample

pH

Complexation Capacity (µg L-1)

Dissolved copper (µg L-1)

Electrochemical

Bacterial Bioassay

C. protothecoides Bioassay

Chlorella sp. 12 Bioassay

Jan-95

Konkonda

7.77

5.3

9.1

-

8

-

Jan-95

Nukumba

7.77

9.3

23

-

14

-

Jan-95

Lewada

7.99

1.8

27

-

4

-

Jan-95

Obo

8.15

17.9

20

22

30

23

Jan-95

Ogwa

8.20

8.0

19

27

17

17

Jan-95

Burei Junction

7.28

11.3

33

-

25

36

Jan-95

Kiunga

8.10

2.3

18

11

10

9

Water samples collected over a 12-month period. No samples directly toxic to Chlorella protothecoides or Chlorella sp. 12

53

Appendix 3 Speciation Surveys, September 1996 – June 1997 (from Apte et al 1997) Algal species: Chlorella sp 12 (Stauber and Apte, 1996), Bacteria Species: Isolate 37 (Davies et al. 1998) dCu (µgL-1)

ASVlabile Cu (µg L-1)

CuCC (µg L-1)

Algal Bioassay (%EDTA Control)

14.5

3.8

10.7

59

1