Berliner Geowissenschaftliche Abhandlungen

BERLINER GEOWISSENSCHAFTLICHE ABHANDLUNGEN

Reihe A Band 176-I1

Jörg Hettler and Bernd Lehmann Environmental Impact of Large-Scale Mining in Papua New Guinea: Mining Residue Disposal by the Ok Tedi Copper-Gold Mine

UN P

. .. ..,.....* .. lWi'-'.& 41",

FU.TU.TFH Berlin 1995 BERLINER GEOWISSENSCHAFTLICHE ABHANDLUNGEN

Reihe A: Geologic und Paläontologie Reihe B: Geophvsik Reihe C: Kartographie Reihe D: Geoinforniatik Reihe E: I'alaohiologie

Reihe A: Geologic und Paläontologie

l-lciauiccbeii 'on cowicnchjli1 chini 1iit itu1ii de,' Ficicii und dci Tcchnkchcn Lm critJI Berlin so\\ ic \ on dci Tcchnkchen Richhochchu Ic Berlin

Schriftleitung: Di. Cli. Kuhncrt ( FLI ). Prof. Dr. U. Ripke (TFH ). Dr. F.Sc hrink (TU ). Prof. Di'. H. Kcupp (Ft

Fdr den Inhalt dci BcitrLic ind die \iitiicn alicia ciant )iIIich,

ISBN -89582-019-9 . ISSN 0172-S7S1

0 1995 by Selhst\erhnz Faehhereich Geo iciicIiiitcn. LU Berlin I Berliner geowiss. Abh. (A) 176-111 52pp1J 31 figs., 17tabs.,2 maps Berlin 1995

JörgHettler and Bernd Lehmann

Environmental Impact of Large-Scale Mining in Papua New Guinea: Mining Residue Disposal by the Ok Tedi Copper-Gold Mine

J10

UN P

Selbstverlag Fachbereich Geowissenschaften, FU Berlin Printing of this publication was funded by United Nations Environment Programme (UNEP)

Druck: G. Weinert, Offsetdruckerei Saalburgstr. 3 12099 Berlin Germany Environmental Impact of Large-Scale Mining in Papua New Guinea: Mining Residue Disposal by the Ok Tedi Copper-Gold Mine

Jorg Hettler and Bernd Lehmann **

Summary

The Ok Tecli copper-gold mine, located in the eastern part of the central mountain range of New Guinea, discharges daily approximately 80,000 tons of ore processing residues and a similar volume of waste rock and overburden into the headwaters of the Ok Tedi River. The Mount Fubilan orebody, which is the source of heavy metal-rich sediments deposited along the Ok Tedi and on the Fly River floodplain, contains a suite of base metals, of which copper is of primary environmental concern. The Ok Tedi flows into the Fly River 200 km downstream of the discharge point, where the mining wastes carried as suspended load are diluted. In the present study, the deposition of mine-derived sediments in the lower part of the Middle Fly River floodplain, and the hydrochemisiry and potential mobilization of trace metals, particularly copper, in the alluvial plain was investigated. To this end, a total of 156 sediment cores and sur - face sedinient samples and 117 water samples from the upper Ok Tedi and the Middle and Lower Fly River floodplain were taken. The suspended matter content in Middle Fly River water today is about 5-10 times higher than the natural background of about 60 nig'l. Near-surface sediments deposited along the river channel contain up to 1100 mg/kg Cu, the mean value is 530 (:La = 240-820) mg/kg. Of the floodplain water bodies, cut-off meanders receive the largest quantities of mine-derived sediment. Deposits of up to 70 cm in thickness of copper-rich material with 800-1000 mg/kg Cu, which have accumulated since the mine started discharging residues in 1984, were detected in oxbow lakes. Very high deposition rates (around 4 cmlyear) of mine-derived sediment were determined in locations close to the creeks and channels which link the Fly River with the outer floodplain. Due to the flat terrain, turbid Fly River water intrudes regularly upstream of the floodplain tributaries (measured intrusions up to 25 km). A thin layer of 1-5 cm of copper-rich material (400-900 mg/kg Cu) was usually found on the bottom of drowned (tributary) valley lakes. Copper in sediment deposited in the pre-mining period gave a median value of 44 (+a = 25-63) mg/kg. Riverine particulate matter also settles down on the floodplain at times of overbank flow, which leads to extensive copper contamination in low-lying swamp Sites close to the river. Natural deposition rates in the floodplain were determined by C 14 age dating to range between 0.1-1 mnilyear, with the exception of oxbow lakes, where natural sedimentation is much higher. Leaching of copper from material deposited on swampy, vegetated floodplain sites was detected in sediments which were also strongly depleted in calcite and sulfide, which are the components most easily mobilized of mine-derived material. The variable water table, oxidizing environment and acidic conditions generated by decomposing vegetation, typical features of low-lying floodplain swamps, facilitate the mobilization of copper from the solid into the dissolved phase. Water taken from swampy floodplain locations showed copper values of up to 50 pig/I Cu in the filtered sample (membrane filter 0.45 pim).

* Final report on a study funded by the United Nations Environment Programme (UNEP)/SPREP

** present address: Technisehe Universität Clausthal, Institut für Mineralogie und Mineralische Rohstoffe, Adolph- Roemer-Str. 2 A, D-3 8678 Clausthal-Zellerfeld, Germany Average copper content in mixed waters of the inner floodplain is around 9 (±cr = 5-14) g/l, the Fly River water has around 17 (±i = 13-19) jig/i Cu. Copper concentrations in unpolluted floodplain waters were measured at below 2 jig/i. Nearly all dissolved copper in the Fly River system is complexed by dissolved organic carbon compounds, however, a fraction of these complexes appears to be labile and reactive. Comparison of dissolved copper levels measured during the present study in the Middle Fly River floodplain with literature data on copper toxicity and international water quality guidelines shows that chronic toxicity of the metal to the aquatic community is to be expected. Significant negative ecological effects, particularly on the local fish population, may develop with a constderable time lag, since aquatic organisms at the base of the food web are the biota most sensitive to copper.

Foreword by Dr. David Mowbray, Professor of Environmental Science at the University of Papua New Guinea and Government Adviser to the PNG Department of Environment and Conservation

The Ok Tedi mine in PNG ranks amongst the world's largest copper mines, It is a major contributor to the PNG economy accounting for up to 45% of PNG's total annual export revenues. The PNG Government is a 30% shareholder in Ok Tedi Mining Ltd. (OTML). Mining for gold coimnenced in 1984, and for copper in 1986. Mining for copper is expected to continue for a further 10-15 years.

Initially it was intended that a tailings dam should be constructed to contain the mine's waste. However for economic reasons and given that the area is structurally unstable, no tailings dam was constructed, and so since 1986 from 100,000 to 150,000 tonnes of waste per day have been discharged into the Ok Tedi and Fly River system. In 1989 the PNG Government set compliance standards covering suspended particulate load, dissolved and particulate copper levels, biological parameters like fish catch, and others. It is argued by some that the environmental management of the mine is based on monitoring for compliance with standards that have been established 'to facilitate the realization of OTMLs mining schedule" rather than to protect the environment.

Ok Tedi Mining Ltd. has an extensive environmental monitoring programme. Assessment of compliance monitoring and interpretation are done routinely by the company (and the Government), but no detailed data arc available for public scrutiny (if at all) for at least a year after data are collected. This lack of public accountability for environmental performance is unacceptable.

Inevitably the Ok Tedi mine evokes different emotions and perceptions by different groups in PNG. For example, Mr Kipling Uiari, the former Deputy General Manager of OTML and now 13l-IP PNG General Manager, stated in 1993 at the 20th Waigani Seminar on "Environment and Development' here at the University of PNG:

"Importantly, the mine has brought positive changes to the quality of life of neighbouring communities in the previously underdeveloped Western Province. Investment in project infrastructure has amounted to about K300m and many of the basic services now established are available to all citizens in the area. This is particularly true of the road, power, health and water systems, transport and communication facilities. These services are meeting the basic needs of local villagers by increasing life expectancy and providing access to education, employment and business opportunities.

Ok Tedi Mining Ltd. has taken a number of iniatives to encourage as much participation as possible from people in the region and so ensure human development goes hand in hand with mineral development.

Sustainable development also requires good environmental stewardship. During the devlopment of the Ok Tedi project there have been criticisms of its environmental impacts, many of them based on inaccurate information. Ok Tedi Mining Ltd. is a responsible resources development company which recognises that all environmental effects must be carefully considered. It is the extent of these impacts and whether they are reversible or not, which must be considered in relation to the social and economic benefits to the communities involved and the nation as a whole."

Whereas at the same seminar Mr Alex Maun, a prominent landowner stated:

"With the mining operation our life has changed. The loss of our rainforest and degradation of the environment cannot be calculated in terms of short-term money. We are rural village subsistence farmers who depend on the environment for survival.

Before the Ok Tedi Mine operation our life was paradise, we enjoyed both the aquatic and terrestrial resources. We used the river for fishing, washing, drinking and transportation. We made gardens near the river banks which lasted 3 to 5 years.

Now we river people can no longer drink from the river nor can swim, bath or wash clothes in the river. We lack the protein in our diet that was formerly provided by the aquatic and terrestrial resources. Overflow ol' the Ok Tedi River has caused the wild life near the river banks and floodplains to disappear. Some game animals were drowned by sudden floods. Trees in the floodplains are dying completely forcing the wild life to migrate to other areas. Now gardens are no longer made near the river banks.

t would like to stress that Ok Tedi Mining is causing irreversible destruction along the Fly River an Ok Tedi River. Whatever it is destroying will never come back to nornial, e.g, customary land, sago swamps. etc. OTML. is bringillg unsustainable development. We affected river people think it is nonsense talking about sustainable development when the mess done by the Ok Tedi Miiiing is not cleaned up."

In 1994 Ok Tedi again hit the headlines: "Ok Tedi has to build tailings dam - Zeipi". so says Post Courier headline on 18 March 1994. The report goes on:

"Ok Tedi Mining Ltd (OTML) will be told to build a series of dams to dispose of mine wastes or "ship out" if it refuses. Mr Zeipi has always insisted a tailings dam be built because, he says, the river dumping has damaging effects on the ecosystems." (Mr. Zeipi is Minister for Environment and Conservation).

"K2b case looms on mine damage" again says Post Courier headline, this time on 4 May 1994. Then in Post ('ourier on 6 May a report states:

"The leader of a Papua New Guinean clan lodged a writ in the Melbourne Supreme Court yesterday against Australia's biggest company. BI-IP, seeking unspecified damages for allegedly poisoning the Ok Tedi River and destroying his people's subsistence way of life. It also alleges the PNG Goverment. a 3() percent shareholder in Ok Tedi Mining Ltd., had "failed, neglected and refused" to enforce environmental agreements and conve- nants."

These matters are still being dealt with in both PNG and Australian courts and have not been resolved. Furthermore present debate continues between Government ministers and leaders and Mr Zeipi on the type of compensation pay- ments for environmental damage and whether a tailings dam should and could be built.

The Floodplains

The Ok Tedi [nine adds about 58 million tonnes of sediment to the Ok Tedi River per year. It is estimated that 30% is deposited along the Ok Tedi, much of the rest reaches the Fly Delta. An unknown amount is deposited along the Fly River and in the floodplains in the middle Fly. Studies on the coastal and marine ecosystems in the Fly Delta, Papuan Gulf an Tones Straits are being done by both Ok Tedi and Australian scientists funded by Ok Tedi. Studies on the floodplains of the middle Fly are being done by Ok Tedi scientists, and are also the focus of the present 1JNEP study. 4

OTML, in its booklet 'Ok Tedi, the Environment and You" states that

"copper in the sediment in the Fly River is also being transported during floods onto the floodplain where it settles into the lakes, streams and the flooded forest, At this time there is very little scientific information to tell us what effect this copper on the floodplain might have on fish life, feed and breed in this area".

Consequently OTML has commissioned studies of the amounts in and effects of copper on the floodplain ecosystems. Unfortunately the PNG Government and independent researchers generally lack the financial and human resources to do independent monitoring and research of depth and detail. Hence most studies are done by or commissioned by OTML.

This study by Jorg Hettler and Bernd Lehmann is valuable since it is done by overseas scientists funded by UNEP through the universities of Berlin and Clausthal and the University of PNG. It is apparent from their work that copper is being deposited in the floodplain at higher than expected levels. Hopefully future independent scientists can assist in deciding if this copper contamination is likely to have biological effects, for examples if fish populations may be affected.

The present study is a valuable contribution to our present state of knowledge of the Ok Tedi/Fly River system. In fact more environmental research has been done on this tropical river system over the last 15 years than probably any other tropical river system in the world. Yet there is still much uncertainty!

Back in 1982 in his book "The Pot of Gold", Richard Jackson made a statement which (in slightly modified form) still is applicable 13 years later:

"In our present state of knowledge of the workings of natural systems, it is only in the rarest of occasions that honest environmental experts agree as a body and confidently predict a sequence of future events and outcomes. The Ok Tedi project is not one of those occasions. Decisions have been made in the project hoping that environmental risks will be worth it, that is, hoping that in the light of the facts available the magnitude of the environmental impact will be minimal. But the facts available are, still, too few. In the Ok Tedi case, the experts, if they are honest, will be keeping their fingers crossed and environmental monitoring systems under constant scrutiny and hoping that the environmental damages remain acceptable! 5

SUMMARY

Foreword

Introduction .6

Sampling and Analytical Methodology...... 6

2.1. Sampling...... 6 2.1.1. Sediments ...... 6 2.1.2. Water...... 7 2.2. Analytical Methods ...... 7 2.2.1. Sediments ...... 7 2.2.2. Water ...... 9

The Environment of the Ok TedifFIy River Region ...... 10

3.1. Geology of the Mount Fubilan Ore Deposit ...... 10 3.2. Geomorphology and Vegetation...... 11 3.3. Hydrology and Climate ...... 12 3 .4. Aquatic Biology ...... 15 3.5. Population...... 16

Input of the Ok Tedi Mine into the Ok Tedi/Fly River System...... 16

4.1. Description of the Mining Project ...... 16 4.2. Discharge of Tailings and Waste Rock ...... 17

Sedimentology of the Fly River and its Floodplain ...... 18

5.1. Natural Sedimentary Processes in the Floodplain ...... 18 5.2. Deposition of Mine-Derived Material...... 20

Potential Mobilization of Heavy Metals and Hydrochemistry ...... 27

6.1. Heavy Metals in Sediments of the Ok Tedi/Fly River System...... 27 6.2. Hydrochemistry of the Ok Tedi/Fly River System...... 32 6.2.1. General...... 32 6.2.2. Fly River Water...... 34 6.2.3. Floodplain Waters ...... 36 6.2.4. Mixed Waters ...... 37 6.2.5. The Behaviour of Copper ...... 42

Discussion of Biological Impacts ...... 45

S. Conclusions ...... 48

Acknowledgements ...... 49

References ...... 49

ANNEX (all analytical data) ...... 53 rsl

INTRODUCTION

The environmental problems of mimng activities have been receiving increasing attention worldwide in the last deca- de. The shift in the search for raw materials outside of the classical producer countries to the world's marginal regions like tropical rainforests has led to a number of environmental and social problems. In many cases, the extraction of natural resources has been the first industrial activity being developed in peripheral regions. Requirements of infrastructure, logistics and workforce may completely change the environmental and social patterns in regions which have previously been nearly untouched by man. One of the best documented examples of such a development is the Ok Tedi gold-copper mimng project in Papua New Guinea. Before the economic value of the Mount Fubilan orebody was dis- covered in the late 1960's, the mountain was a sacred place to the sparse population of the local Wopkaimin Papua tribe, who were inhabiting one of the world's most isolated regions. The environmental impact of the mining operation, located in a difficult physical environment, has been subject of controversy since the project's early planning stages. The fact that the Ok Tedi mine works without waste retention facilities has been in the focus of the debate in the last few years. In May 1994, local inhabitants living along the Ok Tedi River lodged a $A4,0000 million compensation claim against the mine's operator, BHP, over environmental damage caused by the discharge of tailings and waste rock into the Ok Tedi. (Mining Journal , May 6, 1994)

The rmmng company is engaged in an extensive environmental monitoring program and also has commissioned a number of studies on individual problems which have been executed by international research institutions and consul- tants. The present study, funded by the United Nations Environment Programme, is the first independent research project undertaken in the area. It investigates one of the most important environmental aspects of the mining pro ect. Based on discussions with the Environnient Department of Ok Tedi Mining Ltd., it focusses on the fate of niine-derived sediments deposited in the Fly River floodplain.

The project's coordinator was Dr. Bernd Lehmann, Professor of Applied Geology at Technische Universitat Clausthal. The responsible research officer was Jorg Healer, M. Sc. (Geology), of the Department of Environmental and Resource Geology at Freie Universität Berlin. The sedimentological part of the study was coordinated by Dr. Georg Irion, Professor at Forschungsinstitut Senckenberg, Wilhelmshaven. The University of Papua New Guinea offered comprehensive logistical support during the field and laboratory work. A draft version of this report has been presented by the study team in a series of discussions held in September 1994 in Port Moresby and Tabubil, Papua New Guinea, with Ok Tedi Mining Limited, the Departments of Mining and Petro- leum (DMP) and Environment and Conservation (DEC) of the Papua New Guinea Government, and the University of Papu.a New Guinea,

SAMPLING AND ANALYTICAL METHODOLOGY

2.1. Sampling

Sampling locations were determined with a handheld Global Positioning System (GPS) receiver and available topographic maps of scales 1:100,000 and 1:250,000. Water depth and bottom relief were measured with a portable electronic depth sounder with LCD screen.

2.1.1. Sediments

32 surface sediments from the river bank of the Ok Tedi at Tabubil and from a few locations in the Middle Fly region were sampled with a plastic spoon and collected in geochemical sampling paper bags. The large majority of sediment samples (124) taken along the Fly River was recovered with a gravity corer. Because of the difilcult access to swampy, vegetated floodplain sites, the large majority of sediment cores (and water samples) were taken from channels, lakes and small water bodies which were accessible by boat or dugout canoe. The ii gravity corer consists of a massive aluminium tube with a backslash (floating) valve mounted on the upper end. Fins are welded to the aluminium body to stabilize the corer during its free fall through the water column. Coring tubes of transparent acrylic glass (Plexiglas') of 1 in or 1.5 in length were fixed inside of the aluminium shaft. The corer was brought in an upright position above the water and then released for free fall. To increase penetration depth in the bottom sediment, a ring-shaped lead weight of 10 kg was fitted to the alumimium shaft. The corer is hoisted from the bottom of the sampled water body with a rope fixed to the backslash valve, which remains closed. The sediment cores recovered have a diameter of 36 mm and a length of between 20 and 70 cm depending on the nature of the bottom sediment. The cores, which showed no or minimal disturbance, where pushed out of the plexiglas tubes with a rubber stopper and were packed in clean polyethylene bags. On floodplain sites, the sampling tubes were dnven by hand into the sediment. After closing the open end with a rubber stopper, the tube was pulled out. The sediment cores recovered were usually in the same range of lengths as the cores from water bodies.

2.1,2. Water

A total of 117 water samples in the Ok Tedi/Fly River system was taken. Water temperature, pH, conductivity and oxygen content were measured in the field using portable electronic equipment. The reduced species N}14+ HS and N07 were also determined in the field using "Merck Aquaquant" reagent kits for rapid water analysis, based on a colorirnetric method. All water samples were gulp samples taken by hand 20-50 cm below the water surface, the volumes ranging between 250 and 1500 ml. The wide-mouth polyethylene bottles were soaked in dilute nitric acid for two days and washed with demineraiizcdldistilled water before use in the field. The containers were rinsed twice with water from the sampling Site before the sample was taken. Upon return to the field laboratory within a few hours after collection, alkalinity was determined by titration with 0.02 mol HCI to pH 4.5 (APHA Standard Method No. 403). The water was filtered through 0.45 jim cellulose nitrate membrane filters (Sartorius, Germany) with a portable "ANTL.IA" pressure filtration system (Schleicher & Schuell, Germany) which consists of a pneumatic pump, a 50 ml syringe cylinder and a filter holder of 50 mm diameter. Sometimes a pre-filter (Schleicher & Schuell Blue Ribbon ashless) had to be used for highly turbid water, Alter sullicient rinsing of the pump and filtration system, a 120 ml water sample was taken and was immediately acidified with 3-4 ml of concentrated nitric acid (Merck Suprapur) per litre of sample. Filters were retained for gravimetric determination of suspended solids. The suspended matter content in some samples was measured using a 1,5 1 "lmhoff' ftnnel-shaped sedimentation cylinder. Few water samples were split and a subsample was acidified unfiltered to determine the trace metal content associated with the particulate matter. Water samples selected for anion analysis were filtered but not acidified, DOC samples were stabilized with 2 mL/l of 50% H2 SO4. Great care was taken to avoid cross-contamination of collected material. All equipment was acid-washed and rinsed with distilled and deionised water between use for different samples in the field laboratory. Blanks were included and submitted to the same procedure as the samples.

2.2. Analytical Methods

2.2.1. Sediments

Sediment cores were cut in halves in the laboratory with a stainless stell knife to allow visual inspection and taking of photographs. Sub-sections of sediment cores were wet sieved with distilled water through nylon sieves of 20, 60. 100 and 200 jim mesh size. The resulting size fractions <20 jim. 20-60 jim, 60-100 jim, 100-200 jim and >200 jim were washed from the sieves into plastic containers and dried to constant weight at 80° in a drying oven. Grain size distribution was de- tennined after weighing. The fraction less than 20 jim in samples selected for clay mineral analysis was submitted to gravity separation in "Atterberg" sedimentation cylinders. Of the resulting three fractions (<2 jim, 2-6.3 jim, 6.3-20 jim), the finest was used to prepare smear slides which were submitted to X-ray diffraction analysis. 8

cc I- en C cc C90 — — to — -4-I — cc cc cc cc cc O .4-F O'i —In —cc

o iL en CD 664 r.UN 16 - cc a, 14

- a' cc cc cc -,-f cc en 16 I.. cc cc I. In 14 cc cc cc cc cc

In cc In ma elI — 0 to

43 to cc cc — cc cc In

UI 01 cc cc —I 4-'

U) •0 cc a, C cc C11 - e cc elI .0 In -4-I t .1 U

-el I--Il

ma 00 ta-ma — 4i-.4

14 10. 4*4 In o s too rl Wb

q- 'D to -t-I a, cc c-a c-I 16 n-I 1414 cc cc V a, In -I-I Ia- —. 14 88 8 8 Id en cc I en — >.

16 00CC 0C 8 — to Sn Efl Ia, 4 UI UI

2 8 00 01 cc to 4.1 a- to to

2 8 8 cc. 8 to to a, — .1_F I a, cc cc cc 0 cc cc cc cc ._ cc cc 1416 80 cc to cc c-a -I-I en b4 c-a elI 8 2 016 to 01 4d. ccH. N-

'-4 cc 41 U) en-I — cc cc II-' 4-44) 14 .;. .. . - 1616 cc- - - cc --4 cc N— cc 160 cc C.) E4 9

Samples selected for chemical analysis (grain size <20 jim was the standard fraction used) were shipped to commercial laboratories (ACME and XRAL, Canada), where the sediment was homogenized and pulverized in an agate mill, In the "total digestion procedure, 250 mg of sample is digested with 10 ml HCI04-HNO3 -HC1-HF at 200°C to fuming and diluted to 10 ml with diluted aqua regia. The vigorous digestion procedure leads to potential loss of As. Sb and Cr due to volatilization during HCJ04 fuming. The leach, however, is partial for magnetite, chromite, barite, oxides of Al, Zr and Mn. 35 elements were determined by inductively coupled plasma spectrometry (ICP-S), of which 24 are reported in the present study. Arsenic and antimony were additionally determined by hydride generation with aqua regia digestion and ICP-S analysis. Gold together with some other elements were analyzed in selected samples by neutron activation analysis (NAA) in a commercial laboratory (Bondar-Clegg. Canada). Insoluble carbon and sulfur were determined by Leco furnace, in which CO2 and SO2 gases are released during combustion and then detected. A 15 % HC1 leach prior to combustion was performed to remove soluble sulfate and carbonates. This procedure ensures that only organic carbon and sulfur bound as metal sulfide is detected.

Two reference sediment samples, NBS 2704 Buffalo River Sediment and EECJBCR RM 280 Lake Sediment, were submitted to the commercial laboratory for quality control. Results are shown in Table 1. Radiocarbon age dating was performed on five samples of peaty sediment from drowned valley lakes at the C 14 laboratory of Kiel University, Germany,

2,2.2. Water

Water samples were analyzed at a commercial laboratory (XRAL. Canada), at the laboratory of the Department of Environmental and Resource Geology at Freie Universitat l3erlin (FUB), and at the laboratory of the Environment Department of Ok Tedi Mining Limited at Tabubil (OTML). Conmiercial analysis for metals was by multi-element JCP-S, of which measurements for 13 elements are reported in this study. The trace metals copper, lead, zinc and cadmium, which were present in some samples in concentrations close to the detection limit of ICP-S, were additionally determined by graphite furnace atomic absorption spcctromctry (GFAA) at FLIB and OTML laboratories. Major anions (sulfate, chloride, nitrate) were analyzed by suppressed ion chromatography on non-acidified samples. Dissolved organic carbon (DOC) was determined with a 'Technicon' auto analyzer.

Quality control for water analysis was more difficult as compared to sediments since no standard reference material was available. Calcium and bicarbonate values, which were determined by two different methods, gave a correlation of 99% in the data set of all measured values. Values obtained for blanks and measurements of the same sample by different methods and laboratories are given in Table 2. The analyses of blanks (demineralized/distilled laboratory water) showed no or minimal contamination, with the exception of zinc. Reported values for this metal have been corrected by subtraction of zinc content in blank samples.

Statistical analysis of sediment and water values was performed using the U.S. Environmental Protection Agency Geostatistical Environmental Assessment software (GEO-EAS, USEPA 1988). Geochemical modelling and speciation determination was done with an updated version of the PHIREEQE software (Plunmier, Parkhurst & Thorstenson, 1980). 10

W1/28.10 Zn Cu Pb Cd -I

XRALICP-S 24 23 2 <1 FUBGFAA 25 22 3 <.1

W1/30.3. Fe Mn Zn Cu Pb Cd g/l

XRAL ICP-S 126 167 8 4 nd nd FUIB GFAA nd nd 10 5 <1 <.1 OTIvILGFAA 102 76 8 5 <1 0.05

W3/1.4. Fe Mn Zn Cu Pb Cd g/l

XRALICP-S 111 21 28 <2 nd nd FUB GFAA nd nd 20 15 <1 <.1 OTMLGFAA 87 24 19 13 <1 <04

W1/7.4. Fe Mn Zn Cu Pb Cd

XRALICP-S 319 13 17 53 nd nd FIJBGFAA nd nd 20 50 <1 <.1 OTMLGFAA 134 15 15 34 <1 0.1

W3/8.4. Fe Mn Zn Cu Pb Cd g/l

XRALICP-S 50 <5 <5 14 nd nd FUB GFAA nd nd <5 16 <1 <.1 0Th4L GFAA 48 4 1.3 16 <1 0.08

Blank (DDW lab water) Fe Mn Zn Cu Pb Cd pig/i

XR.ALICP-S <10 <5 11 <2 nd nd FUB GFAA nd nd 20 <1 <1 <.1 OTML GFAA 27 0.5 13 0.2 <1 <.04

Table 2. Quality control for water samples. The same sample was analyzed by three different laboratories

3. THE ENVIRONMENT OF THE OK TEDUFLY RIVER REGION

3.1. Geology of the Mount Fubilan Ore Deposit

Trace metal contamination of the Ok Tedi/Fly River system through mine discharges is closely related to the geochemical composition of the Ok Tedi ore. The orebody at Mount Fubilan consists largely of an intruded and altered monzonite porphyry stock of Lower Plei- stocene age, which hosts a mesotherrnal, stockwork and disseminated copper-gold mineralization (Rush and Seegers 1990). Intense weathering at Mi. Fubilan led to the formation of a copper-depleted, but gold-enriched cap (gossan) overlying the primary mineralization. The deposit is similar to other big copper porphyry mines in the Pacific Rim region. Pyrite and chalcopyrite are the dominant sulfide minerals in the protore below the supergene enrichment zone. Average metal concentrations in the orebody are 0.75% of copper and 0.67 grams per ton of gold (OTML 1993). At the contact between the intrusive and adjacent sediments, calc-silicate, sulfide and magnetite skarns have formed, which make up approximately 10 % by volume of the orebody (Jones and Maconochie 1990).

Porphyry base-metal deposits of hydrothermal origin like Ok Tedi typically contain a paragenetic sequence of the metals copper, molybdenum, silver, gold, lead, zinc, iron, manganese, selenium, arsenic, cadmium, tungsten and others, which are not evenly distributed within the orebody, but are located in distinct zones around the low-grade quartz core of the intrusion. Very high values for lead (1000-3000 ppm) have been found in overburden sampled in October 1991. The skarn mineral paragenesis, has an elevated and more variable content of trace metals as compared to the copper porphyry ore, and is more critical from an environmental point of view. However, arsenic and mercury, which are frequently associated with gold mineralisation, show no enrichment above background levels in the Ok Tedi deposit.

Ore minerals of the oxide zone: cupriferous goethite Cu.FeOOH, cuprite Cu20, native Cu, malachite Cu2CO3(OH)2, azurite Cu3 (CO3)2(OH)2, copper carbonates

Ore minerals of the sulfide zone: chalcocite Cu2S, digenite, covellite CuS (supergene formations) chalcopyrite CuFeS2, bornite Cu5FeS4 (very minor), pyritelmarcasite FeS2, molybdenite MoS2 (protore m ineralizctrion)

Sulfides of Sc and Ag are accessory minerals in the main orebody (Fubilan Monzonite Porphyry)

Pb, Zn, Cd, As, Ag and Se minerals are found mainly in skarn ores

Table 3. Ok Tedi ore composition

3.2. Geomorphology and Vegetation

The minesite at Mount Fubilan, which had an elevation of 2094 in above mean sea level (MSL) before mining started, is located in the upper catchment of the Ok Tedi (Map 1). The site receives rainfall of 10,000 mm per annum, which is amongst the world's highest. Dense tropical rainforest blankets the ridge and ravine topography. The area is seismically active and prone to landslides. North of the Ok Tedi catchment stretches the massive Hindenburg Range with maximum elevations of 3325 m. The range is a highly unstable, cliff-like structure build up of Tertiary Darai Limestone, which is responsible for the high natural calcium and bicarbonate content in Ok Tedi water.

The tributary streams in the area are very steep and form narrow, gorge-like valleys, with boulder-size alluvial debris. Upstream of Ningerum (60 m above MSL), the Ok Tedi is a braided river, up to 100 in wide. At Ningerum, 70 km to the south of the mine, the Ok Tedi River leaves the mountaineous region. Between Ningerum and Konkonda, the river passes through a transition phase and enters the Fly Platform, a vast alluvial plateau with very little relief and meandering streams. 130 km downstream of Ningerum, the Ok Tedi flows into the Upper Fly River at D'Albertis Junction at 20 m elevation above MSL. South of the junction, the river channel is flat and meandering, flanked by thick jungle, extensive backswamps, lakes and lagoons. The reach between DAlbertis Junction and the confluence with the Strick- land River south of Obo (Everill Junction) is called Middle Fly River, with a length of 410 river kilometers. The si- nuosily factor is about 2.4. The channel width varies between 200 and 300 m on average. Downstream of Kai and Agu Rivers/Lakes, the floodplain vegetation pattern is dominated by tall, dense reed (Phragmites, Saccharum) and smaller grasses. Jungle is restricted to a few isolated areas of higher elevation. Trees grow only in rarely inundated areas (with the exception of Sago palms and thin-stemmed Melaleuca) because of their susceptibility to prolonged flooding. The fact that trees are largely missing in the lower part of the Middle Fly points to frequent natural flooding. 12

Lakes and shallow water bodies and channels of the floodplain often have a dense vegetation of aquatic grasses, floating Azolla (Equisetum), Pistia (Araceae), water lilies, and the submerged Ceratophy/lurn. Below Everill Junction, the Lower Fly River is tidally influenced, reaching the delta 400 km downstream. Due to the increased flow rate and a lower gradient, the channel width and meander amplitude are greater than in the Middle Fly. The Fly River floodplain is made up of an imier part, the channel migration zone comprising the present river chaniiel, cut-off meanders and meander scroll complexes, and the outer part with drowned valley lakes and extensive backswamps (Fig. P1).

Fig. Pt. Backswamp close to Obo, several hundred metres away from the Fly River. Due to difficult access, only few samples could be taken from these locations.

The floodplain widens from 4 km width near Kiunga to over 14 km close to Obo. Discontinuous low natural levees, about 1-2 m above mean river level, flank the existing river channel. Drowned valley lakes are typically connected to the river by a narrow, sinuous channel with a length of 1-2 km, 3-10 m width and a depth of 2 to 5 m, becoming more shallow towards the lake (Fig. P2). The channel itself is being kept open by the scouring action of sediment particles in the water. The lakes have a depth of about 1.5 to 2.5 m, cut by slightly deeper channels, decreasing to less than 1 m towards the end away from the river. Their bottom relief clearly reflects their origin as tributaries to the Fly River. During the sea level rise which ended about 5,000 years ago, the river ends of the larger floodplain creeks were filled with sediment derived from the Fly River. Water can flow freely in both directions, depending on the water levels in lakes and the Fly River.

Due to the low relief, the edges of lakes or 'lagoons" are not well defined and tend to merge with neighbouring grass- lands at high water level (Fig. P3). Higher banks, on which the villages are usually located, are remnants of eroded plateaus of Pleistocene age. The water level in the lakes is highly variable. Following consecutive dry months, lakes may dry out completely which can result in massive fish kills.

3.3. Hydrology and Climate

The climate in the investigated area is wet tropical. The minesite receives well distributed rainfall on 325 days throughout the year. Average temperatures show a nocturnal minimum of 20°C and a mean daily maximum of 27°C. Mean annual relative humidity is about 85% in the Fly River lowland. Temperatures are constant throughout the region with an annual mean of 29°C. The Fly River has a total length of 1120 km. Its most important tributaries are the Ok Tedi and the Strickland River. The catckment comprises more than 76,000 km2. The mean annual runoff per unit of catebment area is about 2500 mm, which is higher than for any other river system in the world (Maunsell and Partners 1982).

!')Mt. Fubilan/OKTEDI MINE Tabubi1 'PORGERA

\ Ningerum/

i•7 • ____

Konkonda Kuambit Kiunga p - !' •Nukumba • (. — a-

L2ke Murray

l StCI Bosset tqb I Obo . Ogwa

I I Lewada

Daru

0 25 50km

142 144

Map 1. The Fly River Region. 14

As the river catchment is relatively small, streamfiow is characterized by short-term water level fluctuations with rapid changes between flood peaks and drought periods. Rainfall is highest in the upper Ok Tedi region, with recorded annual values up to 14,000 mm, and 7,800 mm at Tabubil, and decreases in the lowland. At Kiunga on the Upper Fly River the annual mean is 4,700 mm, at Lake Bosset in the Middle Fly area 2,670 mm, and at the coast at Dam 2,100 mm. Both rainfall and river flow show only limited sea.sonality. June to October usually are the dryest months in the Middle Fly region.

Fig. P2. The channel connecting Lake Bosset with the Fly River (right background). Deposition of riverine material is highest along the channel.

Fig. P3. The Fly River, inundated floodplain swamps and lakes, and jungle on the alluvial plateau in the background. 15

The long-term mean discharge rate of the Ok Tedi at the junction with the Upper Fly River (flow 1178 m 3/s at Kiunga) has been measured at 923 m3/s. The combined flow is 2161 m 3/s in the upper Middle Fly. There exists a water exchange between the floodplain and its water bodies and the Middle Fly River. The net supply of sediment-poor, but DOC-rich floodplain water to the Middle Fly, however, appears to be small. At Obo, the long-term mean discharge is 2244 m 3/s. The Strickland River discharges 3110 m3/s into the Fly River at Everill Junction (OTML 1994). Mean daily flow data for the last few years are summarized in Table 4.

Water level fluctuations in the Middle Fly are smaller than in the Upper Fly River, where water level changes of up to 15 m have been recorded at Kiunga. The discharge of the Fly River at its mouth of 6.000 m 3/s makes it comparable in size with the Niger and Zambesi in Africa or the Danube in Europe.

Flow rate (m3fs)

1988 1989 1990 1991/92 1992/93

Bukrumdaing * 25 27 24 27 Tabubil 175 110 * * * Ningerum 305 222 183 193 290 Konkonda 965 850 1122 456 805 Kiunga 1435 1197 1044 807 941 Kuambit 2400 2124 2094 1407 2061 Obo 2515 2613 2424 2057 1978 Strickland 3785 3869 3769 2870 3141 Ogwa 6300 5792 5461 * *

* no data available

Table 4. Mean daily flow data collected by OTML for the Ok Tedi, Fly and Strickland Rivers at different locations in the catchment.

3.4. Aquatic Biology

A study conducted prior to OTML's operation (Roberts 1978) came to the conclusion that the Fly River system supports the most diverse fish fauna in the Australasian region, with at least 105 freshwater species from 33 families. The Sepik River in Northern New Guinea, the river closest to the size of the Fly on the island, has a relatively low overall fish density with a total of 57 freshwater species (Allen & Coates 1990).

The fish population in the Fly is remarkable for the large size of some species (e.g. black bass and barramundi) and the abundance of endemic species like catfish. The most targetted species by commercial fisheries is the barramundi Lales ca/carfer, which roams the entire length of the Fly and resides in floodplain water bodies during most of its lifecycle. The barramundi migrates annually to the coastal areas for spawning (Eagle 1993).

Fish ecology is characterized by an overlap of species types resident in the two main habitats, the river channel and the floodplain with its waterbodies. Due to the high biological productivity of the Fly floodplain, the majority of the food for the Fly River fishes originates from off-river sources (Kare 1992). Overlap in diet and habitat requirements is an important mechanism for survival since prolonged periods of low water level may result in drying out of the floodplain and its shallow water bodies (Eagle 1993). Under these conditions, the fish take refuge to the stream channel and oxbow lakes, which may represent the only standing water on the floodplain. The fish populations decline substantially, however, the recovery usually is rapid with recolonisation of the newly flooded habitat by the surviving stocks (Smith & Bakowa 1994). The most important food items for fish are aquatic and terrestrial invertebrates (freshwater prawns, mayfly larvae and other insects, worms etc.), algae, plant and organic detritus. Predatory fish like barramundi and catfish feed on smaller fish like Nematalosa herrings. 16

3.5. Population

In the entire Ok Tedi/Fly River drainage area lived in 1980, according to a national census, about 73,500 people. They speak 28 different languages and form five different language families: The Ok and Awin people of the Ok Tedi and Upper Fly River, the Maririd of the Middle Fly and Lower Strickland (Lake Murray) region, the Suki-Gogodala people of the Lower Fly, and the Trans-Fly peoples of the southern coastal plain. The mountain people are hunter-horticultur- alists living on a subsistence economy. Fish becomes a more important part of the diet for the lowland riverine people. In the Middle Fly-Lake Murray region, habitable land and ground suitable for cultivation of plant foods is scarce. The inhabitants (about 4,500 in 1980) are mainly hunter-gatherers who collect wild sago, cooking-bananas and sugar cane, and use the fish resource of barramundi, black bass and catfish (Busse 1991). The sale of crocodile skins is an important source of cash. The largest populations in the Middle Fly area are at Lake Bosset, Lake Pangua and Lake Da- viambu. Small villages and homesteads are typically located along lakes and rivers. Virtually all travelling is by dugout canoe. During extremely dry periods, in which lakes may dry out completely, people abandon their villages and move to the main rivers which remain the only source of water. The development of the Ok Tedi Mine has resulted in massive cultural and socioeconomic changes in the affected region. A former government adviser emphasized the "psychological trauma that Ok Tedi development will bring to the simple life styles of the mountain people" (Pintz 1984). The mining company has established the "Lower Ok Tedi/Fly River Development Trust" which offers basic village infrastructure in the field of transport, health, education and business development. It supports an estimated 30,000 people in 101 villages living along the river system, who may be negatively affected by the discharge of mine residues and associated problems.

4. INPUT OF THE OK TEDI MINE INTO THE OK TEDILFLY RIVER SYSTEM

4.1. Description of the Mining Project

Mining of the Mount Fubilan orebody as an open pit started in 1984, sixteen years after the discovery of the deposit by Kennecott exploration geologists in 1968. The particular features of the ore deposit were responsible for the development of the mine in mainly two stages: from May 1984 to late 1988, the gold-enriched cap was mined and processed. This involved the use of sodium cyanide and other process chemicals for gold extraction. Following this stage, the phase of extracting copper ore commenced. The gold is now recovered in the sulfide flotation concentrate. Mining of the Mount Fubilan deposit will continue until 2008 and involves the handling of some 1,400 million tonnes of material, comprising 550 Mt of milled ore residues and 850 Mt of overburden (Salomons & Eagle 1990). Currently, the production is about 80,000 tons of ore and a similar volume of waste rock per day. The latter is material in which the metal content falls below the cut-off grade of 0.2% for copper or 0.8 g/t of gold. The currently produced ore has 0.84% Cu and 0.71 ppm Au (oral comm., OTML, 1994). The sulphide flotation process has a recovery rate of 85% which leaves about 1,300 ppm Cu in the residual material. The copper content of waste rock has an average of 1,100 ppm (OTML 1994). The ore is crushed and subsequently fed to a series of baIl mills where the material is ground to fines. A copper-rich concentrate, containing all valuable metals, is recovered by means of a conventional sulfide froth-flotation process. For maximum sulfide recovery and pyrite depression, the pH of the flotation mixture is adjusted with time to 10,5-I1 and organic process chemicals are added (England et al. 1991). The final concentrate is transported by a 160 km slurry pipeline to the river port of Kiunga on the Upper Fly River. At the Kiunga wharf, the slurry is filtered, dried and stored to await shipment by bulk carriers down the Fly River. From a storage facility in the Fly River Delta, the concentrate is sold under long-term contracts to smelters in Japan, Germany, South Korea, Finland and the Philippines. In 1994, OTML produced 206,329 tons of copper in concentrates, containing also 15.14 tons of gold, 30.20 tons of silver, and other metals. The mine is the world's fifth biggest copper producer. Ok Tedi's mineable reserves in early 1994 were estimated at 459 Mt with 0.72% Cu and 0.70 g/t Au (DMP 1994), sufficient for another 14 years of operation. 17

Following a recent restructuring of ownership, the shareholders of Ok Tedi Mining Limited are as follows: 52% of shares are held by Broken Hill Proprietary, the biggest Australian mining company, which is also the project's operator; the Government of Papua New Guinea has increased its stake from 20 to 30%, and 18% of shares are now being held by Inmet of Canada (formerly Metall Mining, a German-Canadian mining company) (Mining Journal, August 18, 1995).

4.2. Discharge of Tailings and Waste Rock

The tailings from the copper extraction in the mill, approximately 98% of the original feed to the processing plant, is piped without further treatment to the Ok Mani, a tributary of the Ok Tedi (Fig. P4). Waste rock is hauled to erodible dumps adjacent to Mt. Fubilan, from where the material is washed into the headwaters of the Ok Tedi.

Fig. P4. The narrow valley of the Ok Mani, which drains the Mt. Fubilan area, has been filled up by some 20 m, as can be seen at the drowned rainforest trees.

The thickened, alkaline tailings slurry consists of approximately 55% solids of which 78% have a grain size less than 100 urn. The material has about 10-20% of its original copper content, varying amounts of trace metals including zinc, lead and cadmium which occur naturally in the porphyry ore, and small quantities of organic flotation chemicals. Waste rock and overburden is coarser, but has a high percentage of soft material (siltstone and limestone) which breaks down easily, containing copper and significant quantities of other heavy metals. Although some of the waste rock remains temporarily in the dumps and adjacent creek valleys (depending on the rainfall activity), most of it is washed down rapidly into the Ok Tedi River. Since the beginning of mining at Mt. Fubilan, all waste rock and overburden (with the exception of a short period during the gold stage) has been disposed off in the river system. The construction of a tailings darn was halted in early 1984 when a land slide forced the abandonment of the construction site, which was located in a geotechnically highly unstable zone. The larger part of the mine-derived material entering the upper Ok Tedi has a particle size of less than 100 jim and can be transported as suspended load throughout the entire length of the Ok Tedi./Fly River system, given sufficiently high hydraulic transport capacity. The massive input of mine-derived sediments into the Ok Tedi exceeds the sediment transport capacity of the river system, which has led to severe aggradation of the river and a rise of the Ok Tedi channel bed by ten meters and more in the upper reaches. Riverbank food gardens and plantations have been flooded and the natural ecology of the river and subsistence fishing has been seriously disrupted. About 60 million tons of material are delivered to the Ok Tedi.fFIy River system per year, containing approximately 69,000 tons of copper. The daily discharge rates are 160,000 tons of rock material with 190 tons of copper. 18

The presence of iron and base metal sulfides in the waste rock and tailings indicates the possibility of acid mine drainage (AMD) development, which arises from the oxidation of sulfide minerals and subsequent sulfuric acid production. AMD generation in the Ok Tedi waste rock dumps was considered a potentially serious problem in the Ok Tedi Environmental Study by Maunsell and Partners (1982). Low pH in waste rock drainage may result in en]iariced solubility of trace metals like lead, silver, zinc, cadmium and copper (Ferguson and Erickson, 1988). However, acid formation may be buffered by alkalinity released from carbonate minerals in the mine waste, such as calcite (CaCO3). Given the relatively high calcium content in the waste material, development of AMD in the waste rock dumps appears unlikely. Despite the fact that sulfide oxidation is occurring in the mine discharges, no pH values below 7 have been recorded in water of the upper Ok Tedi, neither by OTML nor in measurements during the present study.

5. SEDIMENTOLOCY OF THE FLY RIVER AND ITS FLOODPLAIN

5.1. Natural Sedimentary Processes in the Floodplain

The Fly River has a highly variable flow regime where extensive flooding and overbank deposition on its wide floodplain alternates with extremely low water levels in drought periods. The sedimentary processes in the Fly River floodplain are dominated by the river itself and the suspended load it carries. Sediment delivery by floodplain creeks and channels to the system is negligable. The catchment in the outer floodplain and the adjacent piedmont plain comprises forested areas with some swamp savannah. The content of particulate matter in floodplain creek waters generally is very low (below 15 mg/I), consisting of organic material, mainly plant debris, and strongly weathered soil material.

At mean flow in the Fly River and low water table in the floodplain, intrusion of riverine suspended sediment occurs along pre-existing channels linking the river with the floodplain. Sedimentation is highest in oxbow lakes and tie channels, where medium to coarse silt from the river suspended load settles down, and much lower in drowned valley lakes and floodplain depressions where clay and fine silt may become deposited. Drainage direction throughout most of the Fly River floodplain is towards the Fly, although the inner floodplain, the active meander belt, tends to be higher in elevation than the outer floodplain due to sedimentation along and in the main river channel. No broad continuous levee is developed along the river and where there is a dam, it is cut by small drainage channels. During periods of high water level in the Fly River, which are associated with higher sediment loads because of increased transport energy, river bank overflow and inundation of the alluvial plain occurs. Overbank flooding during moderate floods is localized close to the Fly River channel in flanking swamps. The water returns to the river when the flood recedes, although most of the suspended load carried into the floodplain will be deposited there due to vegetative filtering in the grassland bordering the main river, and in sediment traps such as lakes.

Maunsell and Partners (1982) recorded a period of very high flow in the Fly River in July 1981, which was the highest since 1977. The water level was about 1 in above the natural levees. According to their observation, grassland and fo- rests adjacent to the river channel in the reach between Lake Bosset and Obo were flooded to a width of about 16 km on either side of the channel. Although this situation is highly anomalous, it is evident from aerial photographs that at high water level in the Fly River, turbid flood water may flow several kilometers across the floodplain.

The background deposition rate (i.e. before the mine started discharging material) in drowned valley lakes is very low. Analysis of the fraction less than 2 tm of sediments taken from small islands (0-1 in above mean lake level) in Lake Bosset and Bai Lagoon (Map 2) defmed a mineral composition dominated by kaolinite, aluminium chlorite, quartz and gibbsite. All four minerals are typical of highly weathered tropical sediments. The fact that sediment of probable Pleis- tocene age was found on the surface of these shallow islands indicates that no Fly River material has been deposited there in the last 120,000 years, when the sea level began to drop and Pleistocene sedimentation in the alluvial plain ended. 19 kILL

• 1 :i.

j1 . -- I . •', "•

:- I

\ I,

.. -, H 1 I

I N

( 1

Map 2. The lower part of the Middle Fly River (study area). 20

Core 3130.3. 3/11.4. 1/5.4. 1/9.4.

Locality L. Daviambu W L. Bosset NW Bai L. C Kai L. SW

Depth of peaty layer 13 cm 28 cm 27 cm 28 cm below sediment surface

C14 age 5265 2950 4030 4380 (years before present) ±215 ±70 ±110 ±125

Sedimentation rate 24 95 67 64 mm/1000 years

Water depth at site 1.10 2.40 1.70 2.60 (mean water level)

Distance from Fly (air kin)

TableS. Results of radiocarbon age dating of sediment cores from drowned valley lakes.

Table 5 shows the results of Cl 4 age dating which was performed to determine sedimentation rates in drowned valley lakes. The overall deposition rate seems to be well below 1 mm/year and may be as low as 0.1 mm/year. The sediment deposition at a particular site within a lake is very much dependent on bottom relief and shape of the lake. Core 3/30.3. was taken from the distal end of Lake Daviambu. The upper 13 cm consist entirely of peat-like organic debris. Accu- mulation of this material, derived from the catchment of the lake itself, is much slower than at those sites which receive suspended sediment from the Fly River. In the case of the other three cores, the sediment layer above the peat was made up of Fly River material, as determined from its characteristic clay mineral assemblage with dominantly (low charged) smectite. A thin layer of copper-rich material on the top of the cores was observed.

Due to the lack of dateable organic matter in sediment cores from tie channels and their banks, where sedimentation certainly was much higher, no data are available to establish deposition rates at these sites. The simple fact that the shallow lakes still exist after about 5,000 years of possible sedimentation from the Fly River points to very low deposition rates. Assuming that the maximum channel depth of 5.4 meters, measured in Lake Bosset channel close to the Fly River, reflects the original channel depth of the drowned tributary, then the deposition rate was about 1 mm/year in the last 5,000 years at sites close to the Fly.

At times when discharge from the Strickland River is greater than the flow of the Fly, a backwater effect develops, which results in decreasing river velocity in the Middle Fly and subsequent settling of suspended material on the river bed. The same occurs during low-flow periods. The deposited material may be resuspended at high flows. Pickup et al. (1979) estimated the natural suspended load of the Middle Fly at 7-10 million tons and bed load at 1-2 million tons per year.

5.2. Deposition of Mine-Derived Material

The suspended sediment concentrations in the Middle Fly River today are 5 - 10 times above the natural load of about 60 mg/l. Significant quantitities of mine-derived sediments are deposited and trapped in creeks, lakes and swamps adjacent to the Fly River, Such off-river sites play an important role in the food web and in the reproduction cycle of aquatic organisms like invertebrates and fish. Flow inversion in channels linking the floodplain with the Fly River is an important process because it is responsible for suspended sediment transport to off-river sites during mean flow conditions (Fig. P5-6). Reverse flow upstream the tributary channels was frequently observed during the three field trips undertaken. On April 8, 1993 slightly turbid water (sample Wl/8.4.) with elevated calcium, copper (9 .ig/l) and cadmium (0.12 g/l) levels was encountered in the 21

Agu River/Lake system about 25 km upstream of the junction with the Fly River, which had about mean flow. The Agu River/Lake in this reach runs roughly parallel to the Fly. As a result of a recent Fly River intrusion upstream the Agu River, turbid water was visible in the densely vegetated parts of Agu Lake, whereas in the Agu main channel, Fly water was slowly pushed out in southerly direction (downstream) due to heavy rainfall in the days before the site was visited. A similar observation was made in the Kai River/Lake system which was sampled on April 9, 1993, although in this case Fly River water was still flowing upstream. The upstream current observed close to the Kai River mouth was remarkably strong and made the use of the outboard motor necessary for passage.

Fig. P5. The channel linking Lake Daviambu (background) with the Fly River. Lake water of light orange-brown colour, rich in humic substances, drains into the Fly. The vegetated areas in the foreground are swamp.

Fig. P6. The same location, but at a higher water level in the Fly River. Turbid, white river water flows into the lake.

Water with high conductivity, neutral pH and elevated copper values (39 ig/l in unfiltered sample W4/9.4.) was encountered 19 channel km upstream of the Kai River confluence with the Fly River. Water movement at the sampling location was in northerly direction (upstream). One could expect that there exists a flow-through mechanism in which Fly River water enters the Kai River/Lake system (which is about parallel to the Fly) through an upstream connection.

C C ('1 6 C I I: E C U OE D C -LC) cD.j Cu, o C 60 60 Cci) 0 0 ('J O Ec C C F—b co (9 (c,

('J cli V V

rfl % cr

8 cJ 8 a.) Cj CD 8 FZ E 00 C-, C -alt-) Q Cu, _p C >

(5u) D C cci C 0 ED CR cJ cq 0)

V V

% ~ 23

However, satellite images and aerial photographs show no such connecting channel and give no indication of flow across the floodplain, away from the river. The same observation was made for the Agu River system, although maps show a northerly channel connection of the Agu with the Fly River. According to local villagers, this flow-through mechanism is only active at very high water level in the Fly.

The frequent intrusions of highly turbid Fly River water should result in widespread sedimentation of mine-derived material in the Kai and Agu Lakes. Sediment sample analysis revealed moderate copper pollution. Only the uppermost 3-5 cm of sediment cores showed strongly elevated trace metal contents with Cu around 300-600 ppm. In the Kai River samples, the contaminated sections consisted of deposited riverine suspended matter. In the Agu River system, the uppermost, copper-rich sediments were mainly made up of humic material in which the high copper content may be secondary, due to adsorption on flocs of organic matter. Deposition of mine-derived sediments generally is highest at locations close to the channels which connect floodplain water bodies with the Fly River. Comparatively coarse material (medium to coarse silt) is deposited there, whereas the remaining fractions of clay and fme silt of the riverine suspended load is carried into the floodplain water courses (Fig. S-S4). Trace metals are generally enriched in the fmest particle fractions.

In a sediment core from the Kai River channel bank (1/10.4.), taken approximately 3 km upstream of the Fly River confluence, the upper 38 cm showed strongly elevated copper values (400-700 ppm). This material must have been deposited within the last ten years, since the Ok Tedi mine is operating. The sedimentation rate is about 4 cm/year. Similar observations were made at the other drowned valley lakes investigated (Bosset, Kongun, Kemea, Bai, Da- viambu). Where the well defined channels extend into the lake, substantial deposition of 20-40 cm of copper-contaminated sediment was detected several kilometers away from the Fly River, whereas in the proper lake copper-rich sedimentation did not exceed 5 cm. Sediment core 4/11.4., which showed elevated trace metal levels in the upper 23 cm with about 200 ppm copper at the bottom and 800 ppm at the top section, was taken in Lake Bosset at the location where channel and lake water merge, about 2 km inside the lake (as depicted in topographic maps). Sedimentation of mine-derived material was highest in oxbow lakes with up to 70 cm measured at sites close to the river end of Lake Pangua and Lake Kibuz (Fig. P7). 29 oxbow lakes at different infilling stages exist along the Middle Fly River.

Fig. P7. Inflow of Fly River water, rich in mine-derived suspended matter, into the Pangua oxbow lake. Deposition of copper-rich material is highest at the river end of the lake.

Two sediment cores were taken from the Fly River channel bed (Fig. S5-S7), one at Lake Pangua (water depth 10.5 m) and the other downstream of Obo (depth 14.5 m), both in the deepest part in the channel cross section. The entire bed core from the Lake Pangua site (56 cm) consisted of fine grained sediment with 55-75% finer than 20 jim (clay, fme and medium silt fraction), and displayed a constantly high copper Content of about 850 ppm.

o 0 N EE N C A C

C 0 c Oo 0 - E -c) '0) cE 10 o 01 (OC co Nt I I-> LO o > c"J (I) (tIN- cc C

>'c) N

LL N

0 N N V F V o a c 0 0 0 0 C) 0 N -

N 00 % 01

o 0 N A

C) cE 0 o_ C) OLc 0-

-oc'J LQ 0.9 —Jc -ou (0 1) U)- ° 1) N > I) c'J o - (C (0 (Dç' NO) ND) Uc)

0 0 > N N LL V V

73 % % 25

The upper 30-35 cm of the core taken close to Obo were very similar in composition, however the bottom section (41- 43 cm) showed much coarser material with 75% fine sand (63-200 rim). The copper content dropped to 48 ppm in this section. From both cores, it is evident that suspended load has settled down on the channel floor. Only the section with mainly fine sand, from the sample taken downstream of Obo, shows the typical grain size of bed load. it is known from earlier sedimentological studies (Pickup et al. 1979) that fme material does temporarily deposit on the river bed upstream of Everill Junction. Due to the cohesive forces among fine grained particles, high current speeds are necessary to erode these deposits. The resuspension of mine-derived clay- and silt-size material from the channel floor today may be limited because much more sediment settles down, and will be flushed downstream at high flows only at hydraulically preferred sites like meander bends and immediately upstream of the Strickland junction.

In this connection, it is interesting to note that two islands within the Fly River channel, one immediately downstream of Lake Bosset and the other at the junction of Tamu Creek, have been observed during the field trips undertaken. Se- diment samples taken showed relatively coarse, copper-rich material (Fig. S8). No such islands are visible in aerial photographs from the 1960's and 1970's. The formation of islands in the river channel also indicates insufficient transport capacity of the Fly River to carry all mine-imposed waste material. OTML (1993) reports an increase in bed level at Kuambit, immediately downstream of the Ok TedifFly River confluence, of more than one meter above the 1984 pre- mine baseline.

Two effects probably are responsible for the strongly increased deposition rates as compared to earlier sedimentation: Firstly, because of the increase in the (mine-derived) suspended sediment load of the Fly River, which today also contains more silt than the natural particulate load. Secondly, as a result of the reduction in channel capacity due to bed aggradation, overbank flow frequency increases. In the lower part of the Middle Fly, close to the Strickland River junction, aggradation of the main river channel is higher due to the backwater effect from the Strickland. Sedimentation studies which were undertaken before the Ok Tedi mine was developed (Pickup et al. 1979) assumed that frequent landslides, particularly in the upper Ok Tedi catchment, have lead to a high natural suspended sediment load in the Fly River. Had this been the case, this should have resulted in much higher natural deposition rates in the Middle Fly floodplain. The comparison of sedimentation rates before and after the mining at Mt. Fubilan started, points to the fact that the background suspended sediment concentration in the Fly River was very low, probably in the range of 40-60 mg/i.

ss/gulp (mg/i) ssfDI (mg/I)

1988 9/88-8/90 10/91-9/92 1990191 1991/92

Bukrumdaing * 151(3) 44(5) 57 64 Tabubil 4709 7825(25) 6696(12) 8095 10626 Ningerum 1802 524(124) 3477(12) 5846 5882 Konkonda 698 1420(24) 949(12) 2072 2744 Kiunga 60(24) 26(12) 186 149 Kuambit 305 520(24) 3 16(12) 625 573 Bosset * 171(23) 118(12) * Obo 65 92(25) 103(12) 649 450 Strickland * 414(24) 239(12) 690 509 Ogwa 326 283(23) * 677 323

* no data available

Table 6. Average data on suspended sediment concentrations in gulp (surface) and depth-integrated (DI) samples. Medians shown and number of measurements in brackets. Data from OTML (1988-92).

Attempts were made to establish a suspended sediment balance using data collected by OTML. A proper balance would show sediment losses (or gains) in the system. A brief examination of data in Table 6 shows that there is a large scatter in the measured concentrations over several sampling periods. Of main interest to the present study is the deposition of suspended riverine material in the Middle Fly region. The measured suspended matter Content in gulp samples of near- 26 surface water in the Fly River reach between Kuambit and Obo shows a strong decrease in all sampling periods. This effect can not be seen clearly in the depth-integrated measurements in which the suspended matter content is determined in vertical sections through the water column.

Typically, the suspended sediment concentration of near-surface river water (gulp samples) is 70-90% of the depth- integrated suspended matter content (Higgins 1990). No such relationship can be established in the data of Table 6. Differences between suspended sediment content determined by gulp and depth-integrated sampling were up to 90% in measurements from the same site and date (data in OTML 1993). It is difficult to explain this discrepancy, which is particularly evident at Obo. There obviously exists a pronounced stratification within the flowing water body, with particulate-rich water travelling close to the bottom. Due to the shift of the main current in river bends, which leads to a more turbulent (spiral) flow, the bottom strata with a higher suspended solids content rise to the surface. This effect was observed in the depth-integrated measurements at Obo where suspended sediment concentrations are usually highest close to the outer bend. The stratification in suspended sediment concentration within the water body points to the settling down of suspended load on the channel floor. River bed sedimentation is largely dependent on flow velocity, which at Obo is significantly lower than at Kuambit/Nukumba. A detailed comparison of flow data gathered in February and April of 1992 by OTML gave a mean value of 1.12 rn/s for Kuambit and 0.66 rn/s flow velocity at Obo.

Ok Tedi Mining Ltd. has tried to establish an annual suspended sediment mass balance based on its regular measurements taken in the Ok Tedi and the Fly River (Table 7). The calculations illustrate the complexicity of the system and the difficulties in predicting sediment transport in the two rivers. The measured load values (calculated from a small number of measurements) in the 1991/92 data indicate a loss of sediment from Ningerum to Obo (46 to 35 Mt/a), whilst the model predicts a steady increase in the same reach (43 to 55 Mt/a). In the data set for the following year, the discrepancy between observed and predicted load is minor. Given the fact that sediment input by the Ok Tedi mine into the system is fairly constant over the years, the massive adjustment of the predictive model between the two years (at Obo 38 instead of 55 Mt/a) is hard to understand. The flow weighted load data for 1992/93 suggest heavy erosion of material in the Middle Fly between Kuambit and Obo (increase of suspended load by 4 Mt/a), which is hard to explain, too. This may be attributed to the limited number of data collected (9 depth-integrated samples).

Station Flow Weighted Flow Weighted Predicted Predicted Load (Mt/a) Cone. (mg/I) Load (Mt/a) Conc. (mg/I)

Ningerum 46 5750 43 5991 Konkonda 40 2001 43 1862 Kuambit 37 646 44 724 35 * 450* Obo 55 687 Strickland R. 84* 900* 85 - Ogwa 119* 323* 140* 807

Station Flow Weighted Load Predicted Load

Ningerum 40 - Konkonda 31 35 Nukumbal 34 38 Kuambit Obo 38 38

* insufficient records; estimates based on available data

Table 7. Suspended sediment mass balances for 1991/92 (above) and 1992/93 (below) calculated by OTML (1993, 1994). 27

6. POTENTIAL MOBILiZATION OF HEAVY METALS AND HYDROCHEMISTRY

6.1. Heavy Metals in Sediments of the Ok Tedi/Fly River System

Analytical results for sediments from the Fly River floodplain are shown in Tables 8 and 10 (Al and A3 in annex) and from the upper Ok Tedi in Table 9 (A2 in annex). The samples from the Fly River section are grouped into two populations, i.e. sediments which were deposited before the Ok Tedi mine started discharging residues into the Ok TediIFly River system (Table 8), and sediments controlled by mine-derived material (Table 10). This classification is easily practicable due to the distinct geochemcial signature in the mine-derived material, which shows for some elements a significant enrichment above the natural background. The frequency distribution plotted in log probability graphs for copper in 385 sediment samples is shown in Fig. Dl and for gold (27 samples) in Fig. D2. The distributions are roughly bimodal. The subpopulations have approximately lognormal distribution.

Parameter Mean SD Median Percentiles 25 75

Na mg/kg 4365 3146 3700 2000 5400 K mg/kg 10829 4448 12050 7200 14100 Mg mg/kg 5335 2320 5900 3400 7200 Ca mg/kg 8787 7493 6100 4875 8025 Almg/kg 85538 25348 91100 71700 102400 Fe mg/kg 37302 15185 33700 7200 46300 Mn mg/kg 289 225 197 136 334 Zn mg/kg 142 54 130 115 164 Cumg/kg 45 19 44 32 54 Pbmg/kg 18 9 IS II 23 Cd mg/kg 0.26 0.13 0.2 0.2 0.2 Auj.tg/kg 6.7 6.5 3.0 1.0 8.5 Ag mg/kg 0.24 0.26 0.2 0.1 0.3 As mg/kg 4.6 3.6 4.0 2.0 5.0 Crmg/kg 61 16 61 55 70 Mo mg/kg 1.7 1.4 1.0 1.0 2.0 Comg/kg 13 6 12 9 16 Nimg/kg 33 13 29 24 41 V mg/kg 165 49 171 143 196 Ti mg/kg 3765 1269 4100 3100 4400 Zrmglkg 67 25 67 49 83 Lamg/kg 24 7 25 20 28 Ba mg/kg 309 119 309 255 361 Sr mg/kg 137 57 127 101 162 Scmg/kg 17 5 18 14 20 C% 6.7 9.4 2.6 1.3 6.0 S mg/kg 1447 1445 800 300 2525

Table 8. Mean, standard deviation, median values and 25% - 75% confidence intervals for 27 elements in Middle Fly River background sediment samples (n = 128).

Gold gives the highest enrichment factor of 53, followed by molybdenum (factor 23), copper (factor 12), lead (factor 4.4), calcium (factor 4.1), silver (factor 3.5), sulfur (factor 2.6) and strontium (factor 2.5). Zinc shows an enrichment factor of 1.6. The elements A1>V>Co>Cr>Sc>Ni>Ti>Zr (in decreasing order) are present in mine-derived material in lower concentrations than in background sediments. It is evident that the elements associated with the copper-gold ore from Mount Fubilan are also found in the lowland depositional sites. Those metals which are typically enriched in soils during tropical weathering are found in higher concentrations in unpolluted lowland sediments as in material from the mine, deposited in the floodplain. The comparison of mine-derived sediments from the Fly River floodplain (Table 10) 28 with material from the upper Ok Tedi, which consists nearly exclusively of tailings and waste rock (Table 9), gives the following results: Calcium is found in upper Ok Tedi material 4 times and sulfur 3.8 times higher as compared to lowland mine-controlled sediments.

Parameter Mean SD Median Percentiles 25 75

Namg/kg 11921 4159 10950 8100 15000 K mg/kg 30350 7395 27950 24400 37200 Mg mg/kg 6733 1463 6350 5700 7700 Camg/kg 86117 33796 98450 54800 115900 Almg/kg 61462 11766 62750 54800 65800 Fe mg/kg 47512 29854 40900 29300 54200 Mn mg/kg 700 323 775 436 965 Zn mg/kg 541 450 378 182 779 Cumg/kg 1523 976 1158 805 1791 Pbrng/kg 463 744 123 63 488 Cd mg/kg 1.6 1.3 1.0 0.6 2.2 Auig/kg 426 381 266 107 609 Ag mg/kg 2.4 2.5 1.4 0.8 3.3 As mg/kg 14.9 115 II 5 19 Crmg/kg 30 12 26 21 33 Mo mg/kg 38 14 36 27 45 Comg/kg 13 10 11 5 16 Nimg/kg 17 9 14 10 23 V mg/kg 111 21 105 99 120 Ti mg/kg 1788 305 1750 1500 1900 Zrmg/kg 22 II 19 13 29 Lamg/kg 23 6 22 19 26 Ba mg/kg 410 202 368 275 599 Srmg/kg 576 86 548 510 620 Scmg/kg 7.3 2.4 7 5 9 C% 1.8 0.5 1.9 1.3 2.1 S mg/kg 15211 20643 7900 4425 10625

Table 9. Mean, standard deviation, median values and 25% - 75% confidence intervals for 27 elements in sediments from the upper Ok Tedi, mainly tailings and waste rock (n = 24).

Factors for other important elements which are found in the Mount Fubilan orebody are: copper (2.2), silver and cadmium (both 2), zinc and strontium (1.8), gold (1.7), arsenic and molybdenum (1.6) and lead (1.5). The elements K>Ba>Mg>La>V>Ti>Ni>Sc>Zr (in decreasing order) are present in tailings and waste rock in lower concentrations than in mine-derived sediments of the Middle Fly region. Evaluation of relationships between elements in the data obtained for the upper Ok Tedi sediments give the result that sulfur versus Fe, Mn, Zn, Cu, Pb, Cd, Ag and Mo is strongly positively correlated (r = 0.80-0.99, significance level p <0.05). This can be explained by the fact that most metals are dicharged in sulfidic form into the Ok Tedi River. In the Fly River floodplain sediments, the positive relationship between trace metals and sulfur is much weaker (in the range of r = 0.31-0.71 for the metals mentioned above) with the exception of iron, which shows no correlation with sulfur in lowland mine sediments. 29

Parameter Mean SD Median Percentiles 25 75

Namg/kg 8010 3218 8200 6025 9800 K mg/kg 28807 11610 30300 18850 36950 Mgmg/kg 7242 1831 7400 6200 8575 Ca mg/kg 0139 23842 24800 8200 50350 Almg/kg 0994 14915 81100 72325 91025 Fe mg/kg 0004 11012 39500 32200 46850 Mnmg/kg 528 254 510 310 710 Zn mg/kg 211 68 204 163 245 Cu mg/kg 530 289 529 278 732 Pb mg/kg 79 36 80 48 104 Cd mg/kg 0.5 0.3 0.5 0.2 0.7 Autg/kg 166 81 160 120 205 Ag mg/kg 0.7 0.5 0.7 0.3 1.0 As mg/kg 8.3 5.8 7.0 4.0 12 Crmg/kg 48 14 46 39 56 Mo mg/kg 24 13 23 13 32 Comg/kg 10 5 10 7 12 Nimg/kg 21 9 19 15 24 V mg/kg 145 35 139 123 163 Ti mg/kg 2753 801 2600 2200 3200 Zr mg/kg 44 16 41 32 52 Lamg/kg 26 7 27 22 30 Ba mg/kg 415 100 420 363 466 Sr mg/kg 330 139 313 226 438 Sc mg/kg 13 4 12 10 15 C% 1.6 1.5 0.9 0.6 2.0 S mg/kg 2464 2160 2100 675 3425

Table 10. Mean, standard deviation, median values and 25% - 75% confidence intervals for 27 elements in mine- controlled sediments from the Middle Fly River floodplain (n = 197).

In sediments from the upper Ok Tedi, the metals Fe, Mn, Zn, Cu, Pb, Cd, Ag and Mo are all highly intercorrelated (r = 0.65-0.95). As in the case of sulfur, the positive relationship is lost or becomes weaker in the Fly River floodplain data for the same metals. It is interesting to note that gold shows no relationship with any of the trace metals mentioned above in the upper Ok Tedi sediments In the mine-affected lowland sediments exists a positive correlation with lead (r = 0.68) and molybdenum (r 0.66).

There are mainly two processes operating in the river system which are responsible for the differences in the element concentrations in upper Ok Tedi and lowland sediments: Sediment admixture/erosion and mobilization of metals from the solid phase. The Ok Tedi River on its 200 km long way to the Fly River junction receives water and suspended sediments from several tributaries. Further dilution with uncontaminated riverine particulate matter occurs at the Ok Tedi/Upper Fly River confluence. Due to the fact that unpolluted tributaries have a much lower suspended load than the Ok Tedi, dilution effects, however, are small. Lateral erosion in the highly sinuous channel of the Middle Fly appears to be a more important process. Admixture of weathered floodplain sediment in the river course is responsible for increasing concentrations of elements like zirconium, scandium, titanium and chromium (which are highly persistent in chemical weathering) in the mine-derived sediments deposited downstream in the Middle Fly floodplain. 30

0 4 99.50 0 Fly River 99.00 • C 98.00 .._ 95.00 90.00 ) ppm 80.00 14 70.00 4LM- 60.00 50.00 40.00 30.00 20.00

10.00

5.00

2.00 1.00 0.50

10 100 1000 Cu (ppm)

Fig. Dl. Copper distribution in alluvial sediments of the Middle Fly River floodplain. The probability graph of the composite population of 385 data points (open squares) separates into two approximately lognormal subpopulations (closed squares) with a natural background population of about 50 ppm Cu, and a second population of mine-derived material at 610 ppm (geometric means).

98.00 - Fly River 95.00 Pop.2:

90.00 -j = 170 ppb -

- ± 1o= 120-2 O ppb 0 80.00 0 4

.0 - 4 0 70.00-a 0 0 0 4 Cn 60.00 0 4 0 0 00 50.00 4 0 0 40.00 4 (1) 0 4 ~ 30.00 S -: + 5 20.00 Pop.1; 0 E T 0 =7ppb 4 0 1 0. 00 0 + 10' 2-21 ppb • 0 5.00 - 0• T 2.00 0

1.00 1 T 1:11:11

1 10 100 Au (ppb) Fig. D2. Gold distribution in alluvial sediments of the Middle Fly River floodplain. The probability graph of the composite population of 27 data (open squares) separates into two approximately tognormal subpopulations (closed squares) with a natural background population of about 7 ppb Au, and a second population of mine-derived material at 170 ppb. 31

A sediment particle which is discharged into the upper Ok Tedi travels about 7 days until it reaches Everill Junction 610 km downstream of the discharge point. Durmg this time, mineral dissolution occurs. It is evident that particulate, mine-derived calcite is being dissolved in Ok Tedi and Fly River water. Particulate sulfide minerals, mainly pyrite and chalcopyrite, are also unstable in the oxygenated river waters and are oxidized to sulfates. The trace metals associated to these minerals either go in solution or become adsorbed to particulate matter, mainly to unsoluble iron oxyhydrates or organic matter. They are partitioned between the solid and dissolved phase according to their geochemical mobility in the aquatic environment of the Ok Tedi!Fly River system.

Sediment core 1/1 0.4. Ca Cu Pb S section ppm ppm ppm ppm 0-2 cm 35400 663 51 2300 2-4 cm 44300 668 54 1700 18-20cm 50000 664 98 nd 28-30 cm 56800 618 78 2700 30-32.4 cm 50200 574 94 nd 32.4-34.5 cm 35200 483 83 <100

Sediment core 5/27.3. Ca Cu Pb Au S section ppm ppm ppm ppb ppm 19.5-22 cm 48700 977 76 205 5300 30-32 cm 56000 929 79 nd 3000 46-48 cm 47800 891 71 nd 4600

Table 11. Typical vertical profiles of mine-controlled sediments (fraction <20 rim) in the Middle Fly River floodplain dTeposited at sites close to the river at high deposition rates.

Sediment core 4/7.4. Ca Cu Pb Au S section ppm ppm ppm ppb ppm 0-2.5 cm 7800 583 92 130 600 2.5-5 cm 6600 576 122 160 300 7-9 cm 4200 557 140 nd nd 19-21.5 cm 6400 37 16 <2 <100

Sediment core 1/7.4. Ca Cu Pb An S section ppm ppm ppm ppb ppm 7.5-10cm 8700 401 102 310 200 27-29 cm 14000 356 79 291 500 33-35 cm 7500 69 29 nd 200

Sediment core 2/27.3. Ca Cu Pb An S section ppm ppm ppm ppb ppm 3.5-5.5 cm 6100 430 110 240 <100 5.5-8 cm 6900 320 81 nd 100 30-32.5 cm 5900 40 19 nd 100

Sediment core 1/30.3. Ca Cu Pb S section ppm ppm ppm ppm 2-4 cm 7300 409 81 100 26-28cm 5800 35 14 400

Table 12. Vertical profiles of mine-controlled sediments (fraction <20 .tm) deposited at swampy sites on the Middle Fly River floodplain. Table 11 shows analytical data for two sediment cores sampled from locations close to the Fly River channel (Kai River channel, 1/10.4., and Fly River bank at Obo, 5/27.3.), whereas Table 12 shows data from swampy sites (locations in Table A3).

Calcium and sulfur values are much higher than in the cores shown in Table 12. Very little calcite dissolution and sulfide oxidation has taken place in the material from the river channel. The declining copper values towards the base of the sediment core 1/10.4. reflect the development stages of the mine. In the early mining phase of gold extraction only, tailings and waste rock consisted mainly of oxidized gossan material with a relatively low copper content and almost no sulfide minerals, which can also be seen in the base section of core 1/10.4. It was frequently observed that copper values increased in steps from the base to the top of a sediment core whereas gold showed the opposite behaviour (Table A3 in annex). Sediments of Table 12 display very low calcium and sulfur values, although high lead and gold contents indicate that it is largely mine-derived material. Gold is a very resistant element in tropical weathering, and lead also shows very little mobility in the Ok TediIFly River environment (see following section). Copper, together with calcium and sulfur, is clearly depleted in the sediments in Table 12.

Ca and S remain in the sediment body when the deposition rate is high and each layer of riverine suspended matter is rapidly covered by fresh sediment. Sulfide minerals are stable under conditions of oxygen deficiency, which are generated by the decay of riverine organic matter. The pore water in the fine-grained sediment will soon be saturated with calcium which prevents further calcite dissolution. Mobilization of trace metals will be minimal under these conditions. To the contrary, mine-derived material deposited sporadically on swampy floodplain sites is subject to intense leaching. The undulating water table permits penetration of atmospheric oxygen into the sediment body, which is facilitated by the roots of floodplain vegetation. Both factors result in sulfide oxidation. Rainwater leaches calcite from the deposited material, and the rotting of swamp vegetation generates acidic pore waters which may bring trace metals in solution.

6.2. Hydrochemistry of the Ok TediJFly River System

6.2.1. General

Earth alkaline and alkaline metals. These elements have a high solubility of their salts in common, which is independent of redox changes and partly independent of pH (Na, K). Because of the reaction: CaCO3 + H2CO3 Ca2 +2 HCO3 calcite dissolution requires the presence of carbon dioxide, which forms the carbonic acid consumed in the reaction. The carbon dioxide partial pressure in water decreases with increasing temperatures. Dissolved Ca, Mg and Sr compounds are stable in water in the presence of carbon dioxide. Calcite is abundant in the mine waste discharged into the Ok Tedi/Fly River. Its dissolution exerts a buffering effect on pH which will remain in the range of about 6.4 to 8.0, in spite of concomitant dissolution of sulfides from mine discharge. Calcium forms complexes with humic acids. Mg compounds are, in general, more soluble than their Ca counterparts. Na is the most mobile metal in the aquatic environment. K is usually present in lower concentrations than Na. The natural leaching of K from minerals does not result in significantly increased solubilization because K is usually readily incorporated into new mineral structures like clays.

Iron and manganese. Both metals show a similar hydrochemical behaviour. Manganese tends to be more mobile as compared to iron. Compounds of ferric iron (Fe3+) and Mn4+, which are the stable oxidation states in aerobic waters, are nearly insoluble. In reducing waters, the divalent reduced forms (ferrous/manganous) can persist in the absence of sulfide and carbonate anions. Iron and manganese may also occur in both oxidation states in inorganic or organic complexes. In oxygenated, alkaline waters both metals are present as colloidal suspensions of ferric resp. manganic hydroxide particles which pass through a 0.45 im membrane filter. For this reason, the "dissolved" concentrations reported for iron and manganese in neutral or alkaline waters in most cases do not correspond to electrolyte solutions.

Trace metals. Copper, zinc, molybdenum, cadmium and lead were analysed in waters. 33

Between pFI 6.5 and 8.5, lead speciation in the aquatic environment is dominated by carbonates and hydroxides (Hem and Durum, 1973), which are almost insoluble. Pb may be complexed with organic ligands, yielding soluble, colloidal and particulate compounds. It can easily be adsorbed to particulate organic and inorganic matter, which is the dominant mechanism controlling the distribution of lead in the aquatic environment above pH 6 (Farrah and Pickering 1977).

Copper generally shows a higher geochemical mobility than lead. The distribution of copper species in the aquatic environment depends on pH and on the presence of inorganic and organic ligands. Carbonates and hydroxy-anions are the favoured inorganic complexing ligands in oxidized freshwaters with pH above 7. In the presence of soluble organic matter, sorption of copper to particulates may be relatively ineffective because complexation with humic acids is the dominant process (Jackson and Skippen 1978).

Cadmium and zinc are characterized by a lower affinity to carbonate and hydroxy anions at near neutral pH values. The influence of pH and alkalinity on the solubility of Cd 2 and Zn2 is much lower as compared to lead and copper. Zinc also forms complexes with humic substances, a process which is favoured by increasing pH. Molybdenum occurs in oxidized waters as molybdate (M004 2 ) and bimolybdate (HM004) anions. The anions are readily adsorbed by iron and aluminium oxyhydroxides at pH values below 5. Above this value, molybdenum in natural waters is essentially dissolved. Aluminium in the aquatic environment is mainly present in the form of dissolved and colloidal aluminium hydroxide, Al(OH)3. Its minimum solubility is at pH 5.5-6. Above pH 6.5, soluble aluminium exists primarily as Al(OH)4. It is capable of forming complex ions with inorganic and organic substances. pH. The pH value in the Ok Tedi/Fly River system is not only controlled by the dissolved carbon dioxide concentration in waters and calcite dissolution, but also by the biochemical processes of photosynthesis and respiration. Photosynthe- sis of subaquatic plants is accompanied by the assimilation of ions such as NO3, NH 4+ and HP042 . Charge balance is maintained by the uptake or release of H+ or OW, which lead to alkalinity changes. Alkalinity increases in oxygenated waters as a result of photosynthetic nitrate assimilation according to the bulk reaction (Stumm and Morgan 1981):

106 CO2 + 16NO3 + HP042- + 122 H20 + 18H 1 V C106H2630110N16P1 (mean algae composition) + 13802 NH4+ is the dominant nitrogen species in reducing waters. NH4+ assimilation during photosynthesis causes a decrease mpH: 106 CO2 + 16 NH4 + HP04 2 + 108 H20 C106H26301 10N16P1 + 10702 + 14 H Respiration of plant material leads to reactions in opposite direction. When photosynthesis and respiration are in overall equilibrium, no change in pH will be observed, although there exists a pH variability over the day/night cycle. When the rate of production of organic matter (assimilation of NH4+) is larger than the rate of decomposition, alkalinity will decrease. Peat formation, which was frequently observed in the Fly River floodplain, leads to low pH values in the overlying water.

Reduced species. NH4+, HS and NO2- were determined in waters as indicators of reducing conditions. All three compounds are unstable in oxygenated water. Nitrite is metastable in both reducing (conversion to N2 and N',O) and oxidizing waters (nitrate formation). The stability of NII4+ and HS is pH-dependent. HS is converted to volatile H2S at pH values below 7. To the contrary, NH4 forms volatile NH3 at pH above 9 (at 300C). Within the pH range observed in the investigated waters with oxygen deficiency, only NH4+ was a reliable indicator of reducing conditions. Many redox reactions are slow and depend on biological mediation, hence the concentrations of species encountered in natural waters may be far from those predicted thermodynamically (Stumm and Morgan 1981).

Only the elements C, N, 0, S, Fe and Mn are important participants in redox processes in the aquatic environment. However, since the mobility of trace metals is influenced by the adsorption to Fe and Mn oxides and fixation in reduced species (e.g. formation of metal sulfides), redox conditions are of importance to predict trace metal behaviour.

Dissolved organic carbon (DOC). The term DOC is a sum parameter for a number of polymeric organic substances which contain a sufficient number of hydrophile functional groups (-COO, -NH2, R2NH, -RS-, ROH, RO-) to remain in solution despite their molecular size. Polypeptides, amino acids, certain lipids, polysaccharides, humic and fulvic acids and Gelbstoffe belong to this group. 1-lumic substances in general are a result of the transformation of biogenic material. DOC levels in interstitial waters of deposited organic debris may be much higher than in the overlying water. 34

Humic acid is extracted from humic matter in alkaline solution. Fulvic acid is the humic fraction that remains in acidified solution and is soluble over the entire pH range. Humin is the fraction which cannot be extracted by either acids or bases. Structurally, the three groups are similar; differences are in molecular weight and functional group content. The analytical determination of soluble chelates in natural waters is very difficult, particularly with the minute quantities of metal ions that are usually present. Humic substances have a strong tendency to become adsorbed on inorganic surfaces like hydrous oxides and clays through a mechanism involving ligand exchange of humic anionic groups with 1120 and OH- of mineral surfaces (Tipping 1981). These negatively charged coatings may themselves become active absorbers of trace metals. Colloidal iron and aluminium oxides are stabilized by humates. In highly dispersed form, these colloids pass through an 0.45 im membrane filter. The most important feature of humic and fulvic acids is there tendency to form complexes with metal ions. As polycar- boxylic acids, humates precipitate in the presence of dissolved Ca and Mg due to coagulation. Humic substances display colloid-chemical behaviour.

In the present study, three types of water were distinguished for their different chemistry and main constituents: Fly River water, unpolluted floodplain waters, and mixed waters.

6.2.2. Fly River Water

Water from the Middle Fly River (Table 13 and Table A4 in annex) is characterized by a moderately high content of earth alkaline and alkaline metals which is due to the active mineral dissolution of freshly eroded rock material which is carried in suspension from the Mount Fubilan minesite and the Ok Tedi catchment. The major anion is bicarbonate, followed by the minor anions sulfate, chloride and nitrate. The dominating dissolved electrolytes are calcium and bicarbonate, which account for approximately 90% of conductivity (calculated from mol equivalents), and which also control the moderately alkaline pH of 7.7. The content of dissolved organic carbon (DOC) is fairly high at about 6 mg/I (low number of samples, not reported in Tables). Oxygen saturation measurements of Fly River water gave a mean value of only 66% which is obviously influenced by the oxygen-consuming decay of dissolved and particulate riverine organic matter. Reduced species (NH4+, HS, NO2 -) were present at or below detection limit which indicates oxygenated conditions.

Within the temperature range observed in Fly River water, Pc02 does not change to an extent which would markedly affect calcite dissolution. The plot of pH versus calcium (Fig. 1) shows a significantly negative correlation (p <0.05). The plot of suspended solids versus calcium (Fig. 2) displays a significantly positive relationship. It can be concluded that at low pH, more calcite from the suspended particulate fraction is dissolved, which means that the dissolved calcium concentration is controlled by pH (assuming constant atmospheric PCO2). The fact that the highest suspended solid concentrations are associated with the lowest pH values may indicate that high flows in the Fly River are associated with low pH. Since high flows are a result of heavy rainfall, and because rainwater is saturated with atmospheric CO2, the increased input of carbonic acid may explain near neutral pH values in the river water at high flows. There exists a positive relationship between conductivity and suspended solids (p <0.05, Fig. 3). A high concentration of suspended matter is associated with high values for earth alkaline and alkaline metals (Ca, Mg, Sr, Na and K). No correlation was -found- forsuspended solids versus-Fe, Mn and Mo.-A negative correlation between suspended mat- - ter and dissolved copper and zinc was observed, however statistically not significant. Na, Ca, Mg and Sr are highly intercorrelated. Sample Wl/21.2., which has the highest Fe value of 339 .tg/l, also gave the highest values for dissolved Zn and Cu. In the acidified water sample, the colloidal iron oxide particles are dissolved and adsorbed trace metals are released. This explains the high Zn and Cu values associated with iron. This correlation was not observed in other samples. Dissolved trace metal levels, with the exception of copper (median value: 17 tg/I) are generally low due to alkaline pH and high bicarbonate content. Solubility of inorganic copper species in the p1-I range measured in the river water is very low. The fact that, in spite of the high concentration of suspended matter (which offers adsorption sites), dissolved copper values are relatively high, points to the presence of soluble organic copper complexes. 35 Fig. I Fig. 2 8.2 600

8.1 I 500 8

79 i 400 I

7.8 11 0) r 77 300 I CL E (f 7.6 200 7.5 1 I I 7.4 RM 1

7.3 I I

7.2 [!I I I 22 24 26 28 30 32 34 36 38 40 42 22 24 26 28 30 32 34 36 38 40 42 Calcium mg/I Ca'cium mg/I

Fig. 3 Fig. 4 600 IOU

160 500 140

400 120

47, 100 CD 0) E 300 (I) 80 cn 200 60

40 100 20

0 0 110 120 130 140 150 160 170 180 10 15 20 25 30 35 40 45 50 Conductivity ..iS/cm Conductivity .iS/cm

Scatter plots for selected parameters in Fly River water (Fig. 1-3) and floodplain waters (Fig. 4). 36

Analytical data from the pre-mining period (Maunsell 1982, Kyle 1988) are of poor quality and make comparison with present data difficult. Dissolved calcium concentrations are reported to have been in the range of 13 to 16 mg/i, which corresponds to about half of the present values. Fly River water has a high natural content of earth alkaline and alkaline metals because of the limestone formations mainly in the Ok Tedi catchment. Fly River water chemistry probably was not much different from present day conditions, although Maunsell (1982) and Kyle (1988) report slightly acidic pH values in the range of 5.5 to 63 which appear erroneous. Trace metal levels measured by Maunsell (1982) were at or below the detection limit of 1 tg/l with the exception of Fe, Mn and Zn. Pickup et al. (1979) report mean suspended solid concentrations in the range of 60 to 80 mgll for the Middle Fly and note that "the Fly and the lower Ok Tedi are very clean rivers by Papua New Guinea standards".

Parameter Mean SD Median Percentiles 25 75

Temperature °C 28.2 1.6 28.9 26.4 29.4 pH 7.7 0.2 7.7 7.7 7.8 Conductivity jiSlcm 138 15 136 128 146 Oxygen Saturation % 66 0.2 64 60 68 Suspended Solids mg/I 199 149 155 92 207 Naig/l 1564 299 1425 1350 1660 K g/1 622 87 615 536 667 CaLg/l 29380 6654 26000 24750 31300 Mg.tg/I 1237 170 1190 1075 1345 Srig/l 172 36 158 144 181 Al .tg/l below detection limit of 50 .tg/l Fe Jg/l 72 102 30 8 75 Mn 1g/l 18.4 13.2 12.5 8.5 25 Zn tg/1 9.2 6.2 8.0 4.0 10.5 Cd j.tg/l below detection limit of 0.1 rig/I Cujigll 19.6 11.8 17.0 13.0 19.3 Pb ig/l below detection limit of I ig/l Moig/1 7.9 5.0 7.0 3.0 12.0 HCO3 mg/i 77 4,0 76 73 78 SO4 mg/I 5.0 2.5 3.5 0.0 6.0

Table 13. Mean, standard deviation, median values and 25% - 75% confidence intervals for Fly River water samples (n= 11).

6.2.3. Floodplain Waters

The term floodplain waters is used for courses of water which drain the outer Fly River floodplain. Their catchment comprises forested areas and low-lying swamps of high biological productivity with dense subaquatic and swamp vegetation. The main features of floodplain waters are low pH and conductivity (Table 14 and Table A5 in annex) and their yellow-brown colour ("blackwater"). The content of suspended solids is very low as compared to the Fly River. The material retained on the membrane filter is orange brown, consisting of iron oxides (about 10-15% Fe) and particulate organic matter. The high content of iron, most probably in the form of oxyhydrate colloids, is particularly evident in the non-filtered water samples, but also in the filtered water in which the iron values are much higher than in the Fly River water. Sodium, an ubiquitous element, is present in floodplain waters in a similar concentration range as in the Fly River. Aluminium values were highest at low pH and conductivity (Fig. 4). The overall Al content was higher than in the Fly River, where the metal was below the detection limit (50 1ig/I) in the four samples measured. An influence from the mine discharges was detectable in some water samples (slightly elevated trace metal, calcium and sulfate levels). However, since the main features of floodplain drainage such as low pH and conductivity and high DOC levels were observed in these samples, they were included in this category. Metal levels for Cu, Pb, Cd and Mo in 37 floodplain waters unaffected by mining are at or below detection limit (e.g. <2 pig/i for Cu, Table 14). Zinc is the only heavy metal present in measurable concentrations. Since floodplain waters drain lowland areas of intensely weathered Pleistocene sediments, extremely low trace metal are to be expected. Low alkalinity is a result of active accumulation of organic material. Peaty sediments were repeatedly encountered during lake bottom coring. Oxygen deficiency in stagnant or slowly moving, warm waters does not allow complete respiration (oxidation) of organic matter. There is a positive correlation between temperature and oxygen saturation (Fig. 5), which is contradictory at first sight. However, the oxygen measured was a result of photosynthetic activity. which is highest at intense sunshine, which in turn is responsible for high water temperatures. It is obvious that the waters are not in redox equilibrium. The photosynthetic oxygen production masks the overall oxygen deficiency of the system, which is evident from the relatively high concentrations of reduced species like NH4 and HS. Of both compounds. NH4+ is the better indicator of reducing conditions because hydrogen sulfide is volatile at the low pH and high water temperatures measured. NH4+ shows a significantly positive correlation with dissolved organic carbon (DOC)(Fig. 6), which is also positively linked with water temperature. NH4 and DOC are both a result of decomposition of organic matter. The DOC produced in off-river sites is the main source of organic ligands which are responsible for trace metal complexation in polluted waters.

Parameter Mean SD Median Percentiles 25 75

Temperature°C 30.6 2.2 30.0 29.2 31.4 pH 6.0 0.5 6.0 5.7 6.4 Conductivity p.tS/crn 28.5 10 27 20 34 Suspended Solids mg/I 21 II 19 II 24 Napig/1 1233 409 1140 1020 1280 Kj.tgIl 188 107 193 89 244 Capig/l 4295 1992 4110 3060 5890 Mg pig/I 477 141 480 372 499 Sr pig/i 27 10 24 21 34 Al pig/I 67 48 64 25 80 Fe pig/i values highly variable Mn pig/I 15.4 15.1 7.5 2.5 27 Zn pig/I close to detection limit of 5 pig/l Cd pig/i below detection limit of 0.1 pig/I Cu pig/i close to detection limit of 2 pig/i Pb pig/I below detection limit of I pig/i Mo pig/I below detection limit of 5 pig/i HCO3 mg/I 13 6.5 13 9 17 NH4 mg/i 0.3 02 0.20 0.14 0.38 SO4 mg/i 0.37 0.26 0.27 0.18 045 DOC mg/I 9.4 3.0 8.0 7.0 11.0

Table 14. Mean, standard deviation, mean values and 25% - 75% confidence intervals for outer floodplain water samples(n = 15).

6.2.4. Mixed Waters

Mixed waters are generally intermediate in composition between Fly River and floodplain waters, although the inilu- ence of the Fly River is clearly dominant (Table 15 and Table A6 in annex). Due to the generally flat terrain, the location of the mixing zone of floodplain drainage and Fly River water depends mainly on rainfall in the upper catchment of the Fly River/Ok Tedi and in the Fly River lowland. When there is high rainfall in both catchment areas, the mixing front between Fly water rich in suspended solids and blackwater from the floodplain will be located close to the mouths of creeks and tie channels.

38 Fig. 5 Fig. 6 160. 1;I

140 16 p 0) E 120 0£ 14 C -o L p 100 C C L 12 0 p p C C 0 0) 10 I C (1) 60 0 0) > 0 x a)8 I o 40 > 0 I

.? 6 I 20 0

0 iI I I 41 1 T T F I F 27 28 29 30 31 32 33 34 35 35 37 0 0.1 0,2 0.3 0.4 0.5 0.6 0.7 Temperature °C NH4 mg/I

Fig. 7 Fig. 8 go g0 I

80 I 80

70 70

60 60 C) E50 0) 50 E :L I U . 40 £ 40 U p 1 c..) 30 1 30 p 20 20 IK 1 if I p II I I I p P 1 * 10 I I 10 1 Pp 1 1 1 P II I I I* JK IpI a a I I 0 r T r 1 0 F F I I F 5.5 6 6.5 7 7.5 8 8.5 9 9.5 5.5 6 6.5 7 7.5 8 8.5 9 9.5 pH pH

Scatter plots for selected parameters in floodplain waters (Fig. 5-6) and mked waters (Fig. 7-8). 39

At high flow conditions in the Fly River and previously little rainfall in the lowland, river water intrudes several kilometers upstream of the channels and lakes which drain into the main river under reverse conditions. The intrusion of Fly River water is associated with transport of mainly mine-derived suspended matter, which is deposited in the waters of the inner floodplain. Hence, it is not possible to distinguish between dissolved metals in mixed waters which are directly derived from an intrusion of Fly River water and those that may be mobilized secondarily from the sediments already deposited. Because of the different composition of Fly River waters and those draining the outer floodplain, the physical and chemical interactions occuring are of environmental interest (Fig. P8). Particular attention has to be paid to the behaviour of trace metals, of which copper is the most relevant.

Parameter Mean SD Median Percentiles 25 75

Temperature °C 30.2 2.0 30.1 28.9 31.7 pH 7.3 0.8 7.1 6.6 7.7 Conductivity j.tS/cm 130 90 119 80 151 Oxygen Saturation % 77 42 77 41 107 Suspended Solids mg/I 36 72 15 7 28 Na pig/I 1587 541 1480 1185 1755 K pig/I 539 657 385 272 603 Ca pig/I 21390 8372 21900 13425 28050 Mg pig/i 1226 725 1140 881 1345 Srpigll 150 iii 138 82 176 Al pig/i 102 139 25 25 101 Fe pig/i 357 590 145 32 457 Mn pig/I 82 308 2.5 2.5 24.5 Znpig/l 15.4 15.5 12 6 19 Cd pig/I close to detection limit of 0.1 pig/i Cu pig/I 11.8 11.3 9 5 14 Pb pig/l close to detection limit of 1 pig/I Mo pig/i 7.8 9.9 5.0 2.5 9.0 HCO3 mg/I 53 28 47 39 62 NH4 mg/I 0.21 0.16 0.15 0.10 0.25 SO4 mg/l 2.5 2.0 2.1 0.7 3.4 DOC mg/I 9.0 2.5 9.0 7.3 11.0

Table 15. Mean, standard deviation, median values and 25% - 75% confidence intervals for mixed water samples (n=65).

As mentioned above, the pH of floodplain waters is much lower than in the Fly River. Decreasing alkalinity may lead to increased heavy metal mobility due to desorption from particulate matter and formation of free dissolved metal species. The plot of pH versus Ca shows a weakly positive correlation (Fig. 7). Mg, Sr, Na, K and flCO3 give similar plots. Calcite in particulate form carried into the floodplain waters is rapidly dissolved and exerts a buffering effect on the local waters. Opposite to the main trend in the p1-1/Ca plot, there are samples showing a high calcium content at comparatively low pH. These data are from small ponds on the swampy floodplain and slightly acidic floodplain seepage, i.e. extremely iron-rich water trickling from exposed channel or river banks into the river during low flow conditions. No clear influence of pH on dissolved iron and manganese concentrations was observed, although the highest Fe values tend to be associated with low pH. None of the trace metals showed a significant correlation with pH. The plots of pH versus zinc (Fig. 8) and cadmium display a slightly negative trend. No correlation between pH and Cu (Fig. 9) and pH and Mo was detected. DOC and pH are weakly negatively correlated (Fig. 10) which is due to the acidic nature of humic and fulvic substances. In the absence of calcium bicarbonate buffering, the organic acids control water pH.

Fig. 9 Fig. 10 60 14

50 0) 12 E C 0 40 .0 10 0 Q) C-)

30 0 B C. a 0 0) 0 0 20 - 6 1) > 0 10 (1

0 2. 5.5 6 6.5 7 7.5 8 8.5 9 9.5 5.5 6 6.5 7 7.5 8 8.5 pH pH

Fg. 11 Fig. 12

50 Id

45 8

40 7

S 35 0)

30 0) E C a, -0 20 >' (1) 3 0 15

2 10

0 0 0 10 20 30 40 50 60 70 80 90 0 10 20 30 40 50 60 70 80 90 Calcium mg/I Calcium mg/I

Scatter plots for selected parameters in mixed waters (Fig. 9-12). 41

Reducing conditions, indicated by an elevated NH4+ content, also seem to have little influence on dissolved trace metal concentrations. Only zinc gave a moderately positive correlation with N114+. Ca shows a weakly positive correlation with Mo (Fig. ii). A positive relationship of the alkaline and earth alkaline metals with sulfate (calcium versus sulfate, Fig. 12) was observed. Elevated concentrations of Ca, Mg, Na, K, and Sr are clear indicators of the influence of mine discharges. High sulfate levels are a product of active dissolution of mine- derived sulfide minerals. Although the metals iron, zinc, copper and molybdenum are discharged into the river system primarily in the form of sulfides and undergo oxidation to sulfates, no positive correlation between sulfate and any of the metals was observed. Iron and zinc even displayed a weakly negative relationship with sulfate. This illustrates the complexity of solution chemistry in the investigated waters, and the fact that aqueous metal transport is not controlled by sulfate comptexation. There was also no significant correlation of dissolved organic carbon (DOC) and trace metals found. Only the plot of zinc versus DOC shows a weakly positive correlation.

Fig. P8. Mixing of acidic, humic-rich floodplain drainage (blackwater, DOC 12 mg/l, 18 jig/I Cu) and sediment-laden white water of the Fly River (DOC 5 mg/l, 9 jig/i Cu) in the Lake Bosset channel.

Intercorrelations between elements in mixed waters are very high in the group Ca, Mg, Sr, Na, K and HCO3, with the exception of Na versus K. Zn and Cd display a positive correlation, too. A significantly positive correlation was found for Al and Cu in unfiltered water samples (Fig. 13). Aluminium hydroxide may complex or adsorb dissolved copper. At/Zn displayed a similar correlation. Fe versus Cu in unfiltered water shows no clear trend. Fe is weakly positively correlated with DOC and Mn. The heavy metal concentrations in waters of the Fly River inner floodplain, the zone of mixed water, are of particular interest because of the prominent role which off-river waters play in the ecology and biological productivity of the entire river system. Dissolved trace metal levels in mixed waters are controlled by a number of abiotic and biotic factors. The most important inorganic factor is the moderately high bicarbonate content of Fly River water and the high earth alkaline metal concentrations, dominated by calcium in dissolved and particulate form, which are responsible for the neutral to alkaline pH values in mixed waters. The most prominent biotic factor are the dissolved organic carbon substances which play a very important role as complexing agents. Both factors interact in a complicated manner which cannot be predicted from thermodynamic equilibrium calculations.

Despite the fact that lead in mine wastes is clearly enriched above background values, the dissolved metal contents were in the great majority of samples below detection limit (< I jigll). Elevated concentrations were found only in unfiltered samples and in a few samples which also showed a high content of presumably colloidal iron and manganese, which offer adsorption sites for lead. 42

Speciation modelling of inorganic lead with PHREEQE suggests that at near neutral pH values and oxic conditions in waters, more than 70% of soluble lead exists in the form of the PbC03 0 complex which easily precipitates out. Bet- ween 0-30% of lead may be present in the free ionic form Pb2+ under the environmental conditions in the floodplain.

120

100

80 IN

i... 60 a) U-

0 100 200 300 400 500 600 700 600 900 1000 Aluminium

Scatter plot for Al versus Cu in mixed waters (Fig. 13).

Cadmium and zinc are among the geochemically most mobile elements. Their mobility is less pH dependent as compared to lead. Cd was found in concentrations above detection limit (> 0.1 ig/l) only in waters with a pH below 7. Be- cause of the low Cd values in mining residues, the metal is not considered as being of environmental concern. The same holds true for zinc, which showed a similar concentration range in all waters investigated. Molybdenum, because of its anionic form in water, showed a different behaviour than the other trace metals. Highest concentrations were usually found in alkaline waters, or linked with high calcium values. Mo as well as Zn is considered an essential element to biota. The concentrations observed are much below the toxic threshold. Copper is an essential element to plants and animals, too. Because of its environmental importance, the fate of copper in the investigated waters will be discussed in some detail in the following chapter.

6.2.5. The Behaviour of Copper

The statistical data analysis showed that copper concentrations in water do not appear to be significantly controlled by inorganic factors. Alkalinity and pH, adsorption to oxide surfaces (with the exception of aluminium) and reducing conditions (e.g. formation of soluble copper amine complexes) do not seem to have a significant effect on dissolved copper levels. The abundance of organic complexing agents in the investigated Middle Fly floodplain waters may be responsible for the observed behaviour of dissolved copper. No correlation was observed between DOC and copper levels in the floodplain waters. It is evident that dissolved organic ligands, even at comparatively low concentrations in waters of the Fly River system, are always present in excess of trace metals which may become complexed (although some complexing capacity may be occupied by major cations like Ca and Mg). In the Lower Fly River, downstream of the Strickland junction, measured dissolved copper values are still moderately high at 6 j.tg/l (Table A6, annex) despite the massive 43 admixture of uncontaminated suspended matter from the Strickland River. It appears that the lowering of dissolved copper levels in the Lower Fly is mainly due to dilution effects and that adsorption to riverine particulate matter is of secondary importance. Davis and Leckie (1978) tested the influence of dissolved organic substances on copper adsorption. Certain organic ligands which form coatings on suspended mineral particles enhance the extent of trace metal adsorption, while others show opposite behaviour. Some organic acids inhibit copper adsorption by forming soluble, stable complexes which keep the metal in solution. When the complexing capacity of dissolved organic carbon substances is larger than the rate of adsorption to inorganic surfaces, copper will remain in solution.

Sholkovitz and Copland (1981) also investigated the solubility and adsorption properties of a number of trace metals in the presence of humic acids. Contrary to similar studies which were performed with synthetic laboratory waters, Shol- kovitz and Copland used natural water from a small stream in Scotland which drains peaty soils. The properties of this water (DOC 7 mg/I, pH 6.5, low conductivity) are similar to those of the Fly River floodplain. The authors detected a coagulating effect on humic acids and trace metals, particularly iron and copper, upon adding of only 0.5 mmol/l of Ca, which is equivalent to 20 mg/I Ca. The mean Ca concentration in mixed waters of the Fly River inner floodplain is about 25 mg/l. Coagulation of humic substances with adsorbed trace metals may be an important process when calcium-rich Fly River water mixes with floodplain blackwaters. This is consistent with the detection of elevated copper levels on the top of black, peat-like sediments sampled from lake bottoms far from direct Fly River influence. The experiments carried out by Sholkovitz and Copland (1981) also gave the result that there was no precipitation of trace metals and humic acids when pH (starting point pH 6.5) was changed in the range of 9.5 to 3, below which humic acids, Fe, Mn, Cu, Ni and Cd began to precipitate. This behaviour is contrary to that predicted by inorganic solubility considerations. Complexation of the trace metals by dissolved organic matter is the most reasonable explanation. Or- ganic copper complexes are known to be particularly stable because of their favourable electron configuration (Stumm and Morgan 1981).

Fe and Mn in floodplain waters, despite their chemical similarity, showed no significant correlation in any data set. Of the elements investigated by Shotkovitz and Copland (1981), Fe>Cu>Ni>Cd (in decreasing order) showed the strongest affinity for the dissolved humic substances, Mn and Co the least. The differing tendency of Fe and Mn to form complexes with dissolved organic carbon may explain the missing correlation in data from the Middle Fly River region. In a copper adsorption experiment (spiking to yield 20 j.tg/l Cu) with natural waters containing different concentrations of suspended solids and dissolved organic carbon, Sholkovitz and Copland (1981) obtained results which suggest that solubilization of Cu by dissolved organic ligands, forming ultrafine colloids (<0.01 im), is a more important process than the adsorption onto riverine particulate matter. Even in water containing 100 mg/I mostly inorganic suspended matter and 3 mg/I DOC, most of the copper was kept in solution as the pH increased from 4 to 9.

CSIRO of Australia (1989) performed mixing experiments commissioned by Ok Tedi Mining Ltd. with water from the Fly River and a floodplain tributary. Mine derived sediment with high copper values was added to different admixtures. Total dissolved copper concentrations increased with sediment admixture, the pH remained fairly constant. Dissolved copper species in the resulting solutions were analyzed by Anodic Stripping Voltametry (ASV). The method is used to determine labile trace metal species supposed to be present in the ionic, most bioavailable form. Interestingly, between 30-50% of dissolved copper in the fmal test solutions was measured as labile or "ionic' copper. Assuming that the copper was present as dissolved organic species, the results point to great reactivity of humic copper complexes in the Fly River system. This is consistent with more recent laboratory work undertaken by CSIRO on behalf of OTML (OTML 1994). Electrochemical measurements of bioavailable copper in the Ok Tedi/Fly River system gave the result that the fraction of potentially bioavailable copper is up to 50% of the total dissolved copper concentration.

In the presumably unpolluted waters of the outer floodplain (it is difficult to establish the maximum intrusion range of Fly River water) copper was at or below the detection limit of 2 jig/I. In the Fly River, copper values were about tenfold above this "background", which may actually be much lower than 2 jig/I, i.e. 0.2 jig/l. Mixing of both waters should result in dilution. This is the case in most mixed water samples, however in some waters copper concentrations were much higher. Sample W1/7.4., which had a dissolved copper concentration of 52 jig/I in the filtered and 106 j.tg/l in the unfiltered sample, was taken from a depression in flat swampy terrain about 150 m behind the low dam paralleling the Fly River channel.

Table 2. Particulate and dissolved copper and fish catch for Kuambit/Nukuxnba, Obo and Ogwa.

YEAR KUAMBIT/NUICUMTBA OBO OGWA

pCu dCu fish catch pCu dCu fish catch pCu dCu fish c ug/gg/l kg ugfg ug1 kg u/g ugjl kg

1990 1236 11 15 1202 17 35 524 3 95 1991 905 6 28 879 9 66 401 2 118 1992 715 3 38 693 5 94 321 11 35 1993 875 4 29 850 6 70 388 1 120 1994 581 3 49 564 4 119 270 1 147 1995 581 3 49 564 4 119 270 1 147 1996 562 3 49 544 4 119 253 1 147 to 2008

Source: pCu -Supplementary Investigations, VoL II, Appendix B. dCu -Supplementary Investigations, Vol. L Figure 3. Fish Catch -Suppiementaxy rnvestigations, Vol. Ut, Appendix H, Figures lOa, ha, & 12a and Table 4A.

Table IS Dissolved Copper Predictions (.LgfL) Provided to the State in June 1992.

Year Nukurnba Obo Owa

1991 11(10.5) 15(14.2) 6(5.4) 1992 13(13.1) 17(15.0) 6(6.6) 1993 14 17 7 1994 ii 15 6 1995 11 15 6 1996 12 18 7 ErrorEstixnaxe +-78% +8O%

()annual (lendar vir) mcans of observed data

Table 5.8 Revised Predictions for dCu in the Fly River and Mean Data for 1993

Year Nukumba Obo - Ogwa - - - 1993 16 (4-28)(13.S) 17(3 -311 (17.6) 6.3 (1- 12) (7.4) 1994 18(4-321 18 (4-32) 7.5 (I - 14) 1995 1413-251 14(3-251 6.7(1-12) 1996 14(3-25) 14(3-25) 6.6(1-•12) 1997 13(3 -25) 13(3 -251 6.1 (1- ii) 1998 14(3-25) 143-251 6.611-12) UUpper and lower error ewrwics 0 Annual (lcndar year) rncans of observed data

Table 16. Predictions made by OTML in 1989 for key environmental parameters in the Fly River system, and revised predictions for dissolved copper in 1992 and 1993 (OTML 1990, 1993, 1994). 45

Water sampled from small pools and shallow water courses in the periodically flooded grassland sites usually gave high copper values. The water samples taken generally had low pH andlor alkalinity. It appears that active leaching of copper from deposited mbie-derived material is responsible for the high dissolved values observed. In the floodplain swamps, redox conditions change easily depending on the undulating water table which may result in trace metal mobilization. Because of the dense vegetation and the low rate of water exchange, soluble organic chelates are abundant and may facilitate copper mobilization. The buffering effect of calcium disappears as the easily soluble element is leached from the sediments. Because of the favourable conditions for trace metal leaching, even a thin layer of deposited mine-derived material will be an important source of copper. Due to difficult access, only a few samples from the swamp sites were taken (water was sampled within a maximum distance of one kilometer from the main river channel). It seems that the copper levels in floodplain waters investigated in the present study are biased towards low values (in samples from large lakes and tie channels) and are not representative for the entire floodplain.

Where floodplain swamp waters with leached copper drain into lakes and the main river, they will increase copper concentrations. This may explain why OTML (1991, 1993, 1994) reports higher dissolved copper levels at Obo as compared to the upstream site at Nukumba, below the Ok Ted itlJpper Fly River confluence. A different type of water in which high trace metal concentrations could be expected are the samples called 'floodplain seepage". These moderately acidic waters percolate through the floodplain sediments and drain spring-like into the channels and rivers. The waters are highly reducing and have a very high Fc2+ and M112+ Content which is immediately oxidized to orange coloured gels and floes upon exposure to atmospheric oxygen. Although the floodplain seepage samples showed high Cd and Pb levels, this was not the case for copper, which probably remains in sulfidic binding in the sediment body.

Table 16 shows predictions based on computer modelling by OTML (1990) of various environmental parameters in the Fly River system in response to the impact of mine discharges. These additional environmental monitoring conditions were established to ensure that the Acceptable Particluate Level of 940 mg/i at Nukumba does not result in actual envi - ronmental damage to the Fly River System beyond the level actually speqfled in the predictions (OTML 1990).

The predictions resulting from the Supplementary Environmental Investigations (undertaken in 1986-89 by OTML) have been established for testing whether or not the State's conditions for environmental management of the Fly River System and off-shore are met. Due to a number of reasons, there has been a non-compliance with monitoring conditions beginning in 1990. Particularly the dissolved copper levels were much higher than expected. Measured values at the APL sites Kuambit/Nukumba, Obo and Ogwa have been in a steady increase between 1990 and 1993 (OTML 1994) contrary to the trend predicted by the models. Revised prediction values for the key environmental parameters were developed by OTML in July 1992 and again in December 1993.

7. DiSCUSSION OF BIOLOGICAL IMPACTS

The discharge of ore processing residues and mining wastes has significantly changed the aquatic environment of the Ok Tedi/Fly River system. From the discussion in previous sections it can be concluded that there are three factors of particular environmental concern: the increase in particulate copper, in dissolved copper, and in suspended solids concentrations. The negative impact of each effect on the fluvial system is spatially different.

The massive increase in the suspended load of the Ok Tedi is mainly responsible for the decline in aquatic life upstream of the confluence with the Fly River, where fish populations and species diversity have been dramatically reduced (Smith 1991). The persistently high concentration of suspended sediments in the water interferes with gill respiration of fish and aquatic organisms, modifies the movement and migration of fish and prevents successful development of eggs and larvae. Heavy metals in particulate and dissolved form, and toxic process chemicals like xanthates probably play a minor role in the adverse impact on the aquatic life of the Ok Tedi River. In the Fly River itself, a combined detrimental effect of suspended sediment, particulate and dissolved copper on the aquatic ecology is to be expected. Monitoring undertaken by OTML (Smith 1991, OTML 1993, 1994) shows that fish populations in the Middle Fly River at Kuambit, immediately downstream of D'Albertis Junction, are continuously 46 declining, in the range of 30-50% (M. Eagle, pers. comm.). Statistical analyses gave the result that particulate copper (pCu) was the best negative correlate of fish catches, but dissolved copper and suspended solids also have a negative effect. OTML (1994) speculates that the relationship between pCu and fish catch may either be attributed to an avoi- dance of pCu by fish or by a correlation of pCu with a toxic fraction of dissolved copper.

Both total fish catches and the diversity of the aquatic fauna are affected. No clear negative impact of mine discharges on the abundance of fish in the Middle Fly River downstream of DAlbertis Junction can be established from the data collected so far because of their high variability. Fish biomass monitoring in the last three years has been strongly influenced by low river levels concentrating fish into the river channel as the floodplain dried, hence leading to high catches at the river sites. Clearly, the exact determination of fish populations (including migratory species) in a dynamic fluvial ecosystem is a difficult undertaking since fish catches are influenced by a number of factors which cannot be attributed to the mine waste, like periods of droughts and commercial and subsistence fishing pressure. In addition, OTML (1994) admits that its database on fish biomass is limited with respect to baseline data and comes to the conclusion that "changes between before and after mine start-up conditions cannot be quantified for any Site (in the Fly River system)".

Considering the homogeneity of the Middle Fly River system and the fact that the key water parameters influenced by the mine's operation (dissolved and particulate copper and suspended solids) show little changes in the Middle Fly Ri- ver reach between Kuambit/Nukumba and Obo, negative effects on the fish populations in the lower part of the Middle Fly region are to be expected. OTML (1993) report elevated copper levels in body tissue of the catfish Anus berneyi caught in Lake Pangua, a deep oxbow lake. Anus is a bottom feeding omnivore. Aquatic invertebrates form the most important part of its diet, but detritus/mud is also an important food item (Kare 1992). It is not clear whether the copper detected in the fish was taken up via the bottom dwelling invertebrates, or directly from the copper-rich sediment, the metal being released in the intestinal tract of the catfish. In both cases, it is evident that copper in particulate form on the lake bottom is bioavailable.

In the off-river floodplain sites of the Middle Fly region, elevated concentrations of dissolved copper may negatively affect aquatic life. The floodplain plays an important role in primary productivity to support fish stocks in the river system, and for the recruitment of fish. OTML (1991) states that the floodplain supports a greater stock of fish than do oxbow and drowned valley lakes. The Fly River channel is to a large extent a fish migration route and a refuge during dry periods. Subsistence and commercial fisheries target mainly the off-river water bodies because of their high productivity. Although the vegetated floodplain sites may not be inundated over the whole year, fish invade rapidly into the alluvial plain at higher water levels because of the abundance of food sources.

Animals at lower trophic levels, such as aquatic invertebrates, and juvenile stages of fish are known to have a high sensitivity to dissolved copper. According to Moore and Ramamoorthy (1984), sensitivity to copper is inversely related to the age or size of an animal. A decrease in the abundance of macroinvertebrates, which are at the base of the food web, may lead to a decrease in the overall fish population. Copper is a strong toxicant to aquatic organisms but does not biomagnify in the food chain. End consumers like carnivorous fish show a much lower copper body burden than benthic invertebrates like burrowing mayfly larvae which feed directly from the contaminated substrate (Karbe 1988). The average copper concentration in mixed waters of the inner floodplain is around 10 g/l; the Fly River water has about 17 jig/i of copper. These are concentrations which may be harmful to biota. The applicable United States Envi- ronmental Protection Agency (USEPA) "Water Quality Criteria for the Protection of Aquatic Organisms and Uses' (1986) recommends a maximum value of 6.5 ig/l (hardness 50 mg/I as CaCO3, unfiltered water) for the "chronic" 4- day average concentration, and 9.2 jig/I for the "acute" 1-hour average concentration. The Canadian Water Quality Guidelines (1987) are even more stringent (Table 17).

Dissolved copper levels detected in mixed floodplain waters of the present study are in the concentration range which is reported in the literature to be harmful to sensitive freshwater aquatic organisms. Clements et al. (1989) found a significant decrease in total aquatic insect abundance after 4 days of exposure to 6 .tg/l of copper in an outdoor experimental stream. The most sensitive species, a chfronomid larva (midges, Diptera), was eliminated at 13 jig/I copper after 10 days. Moore and Winner (1989) also found out that benthic chironomid and mayfly larvae show a high sensitivity to dissolved copper. 47 Table 3-1. Summary - Guidelines for Frhter Aquatic Life Pammeter Guideline Comments

Inorganic paramezers -- - - Aluminum 1 0.005 meL pH<6.3: 1Ca)<4.0 mgL: DOC(2.0 mgL 0.1 meL pH2-_6.5: [Ca]4.0 meL* DOC2.0 mgL

Anumony 1D

Arsenic 0.05 mgL - I

Beryllium ID

Cadmium 0.2 p.gL' Hardness 0.60 mgL (CaC0) 0.8 p.i.gL' Hardness 60-120 mgL (CC03 ) 1.3 ieL' Hardness 120-180 meL' (CaCO) 1.8 i.gL - ' Hardness >180 mgL' (CaCO3)

Chlorine Itotal residual chlorrne) 2.0 agL Measured by aniperomethc or eguivajent method

Chmmium 0.02 mgL - 1 To protect fish 2.0 gL i To protect aquatic life. including zooplankton and phytoplankton

Copper 2 i.gL' Hardness 0-60 mgL' (CaCO3) 2 p.gL' Hardness 60-120 mgL' IC2CO3 ) 3 sgL I Hardness 120-180 mgL ' (CaCO3 ) 4 t.Lg -L Hardness >180 mL (CaCO3)

Canide 5.0 z.g-L 1 Free cyanide as CN

Dissolved oxygen 6.0 mgL WarTn-water biota - early life stages 5.0 mg-L I - other life stages

9.5 meL Cold-ater bioca - early life stages

6.5 mgL ' - other life stages

Iron 0.3 mgL'

Lead 1 ig.L_L Hardness 0-60 mgL' (CaCO3) 2 Hardness 60-120 mgL (CaCO3 ) 4 sg.L' Hardness 120-180 mrL (C2CO3) 7 sgL Hardness >180 mgL (CaCO3)

Mercuri 0.1

Nickel 25 gL Hardness 0.60 rngL 1 (CaCO3 ) 65 J.g- L' Hardness 60.120 mgLt (CaCO) 110 Hardness 120-180 mgL (CaCO,)

150 g.LI Hardness >180 mgL. (CaCO3 )

Nitrogen Ammonia (total) 2.2 mgL pH 6.5: temperature lOT (see Table 3-12) 1.37 mgL pH 8.0: temperature lOT Nimme 0.06 mgL 1 Nitrate Concentrations that stimulate prolific weed growth should be avoided Nitrosaxnines ID

pH 6.5-9.0

SelenIum 1 .tgL i

Silver 0.1 sgL 1

Thallium ID

Zinc3 0.03 mgL'

Phvswa! parameters Temperature Thermal addiciotis should not alter thermal stratification or turnover dates. exceed maximum weekly avera2c temperatures. and exceed maximum short- term temperatures I see Section 3.2.3.1.1)

Total susoended solids increase of 10.0 Background suspended solids 00.0 mgL -'

increase of 10 above Backeround suspended solids >100.0 mgL backeround I Concentrations of heavy metals reported as total metal in an unflitered sample. ID - insufficient data to recommend a guideline. Tentative guideline. Table 17. Canadian Guidelines for the protection of freshwater aquatic life (CCREM 1987) 48

Burrowing mayfly larvae of the genus Plethogenesia (together with Macrobrachium freshwater prawns) were the dominant constituents of the macroinvertebrate fauna in terms of biomass in the Fly River (OTML 1987). Both are important food items for the local fishes. Toxicity of particulate copper in mine wastes on mayflys has been proven by bioassays undertaken by OTML in 1988 (OTML 1989).

Williams Ct al. (1991) report a 50% mortality after three days of exposure of the extremely sensitive tropical freshwater shrimp Caridina sp. to 4 ig/1 copper. Most studies on copper toxicity consider the free ionic species as the most bioavailable form, and copper complexed with naturally derived dissolved organic carbon (DOC) as much less toxic or even non-toxic (e.g. Flemming and Trevors 1989; Meador, 1991). Recent scientific work (Winner and Owen, 1991), however, has provided further evidence that dissolved organic carbon also may enhance copper toxicity. The authors used the alga Chlamydomonas reinhardtii to test the bioavailability and toxcity of organically complexed copper in water from a freshwater pond.

Toxic effects of dissolved copper on alga deflagellation and population growth were observed at values above 12.2 ig/1 in natural waters with DOC concentrations varying from 5 to 14 mgIl. They found a positive correlation between DOC and copper on alga toxicity and concluded that labile organic copper complexes may be responsible for this effect, demonstrating that a simple relationship between DOC concentration and copper toxicity cannot be established. OTML (1994) has commissioned toxicity tests with the freshwater green alga Chiorella protothecoides. No reduction in algal growth rate occurred at dissolved copper concentrations of 12 ig/l, which was the highest concentration used in the tests. Since copper values measured both during the present study and by OTML in waters of the Fly River and its floodplain were much higher than 12 g/l, it appears justified to repeat the tests with copper concentrations of up to 50 g/l in order to obtain more information on the species' sensitivity.

8. CONCLUSiONS

The following conclusions can be drawn from the informations gathered in the present study:

Of the trace metals contained in the Ok Tedi mine waste, copper is of main environmental concern because of its strong enrichment above background (about twentyfold) and its relatively high geochemical mobility.

The discharge of mining residues into the Ok Tedi!FIy River system leads to a substantial deposition of copper-rich material in the Middle Fly River floodplain, although most of the mining wastes carried as suspended load finally reach the delta. Areas of standing water and vegetated parts of the floodplain play an important role in the recruitment of fish and primary productivity. Hence, this part of the fluvial ecosystem is particularly sensitive to mine-induced changes in the local environment.

The pollution by copper-rich material is not only a problem of quantity, but also of spatial distribution and the potential mobilization of the trace metal. Deposition of mine-derived sediments in the Fly River floodplain is highest in oxbow lakes. Because of the naturally high organic carbon content in sediments, which is responsible for reducing conditions, only copper in the uppermost layer (few centimeters) of bottom sediments may become dissolved. When copper-rich sediment is permanently covered by water, mobilization and bioavailability of the metal, except to bottom- dwelling invertebrates, is low. However, when mine-derived sediments settle on extensive areas of the vegetated floodplain, the conditions are quite different. Chemical and biological factors facilitate the mobilization of copper. Even a thin layer of copper-rich material may have a negative ecological impact.

The way in which copper affects the aquatic community is not clear. Dissolved organic carbon substances, which are present in high concentrations in the Fly River system, complex the metal and keep it in solution. Little is known about the bioavailablity and toxicity of these organic copper species. Recent research has shown that these compounds may exert toxic effects comparable to the free ionic copper species. 49

The most sensitive biota to dissolved Cu are juvenile stages of fish and aquatic invertebrates which form the base of the trophic web. Populations of larger fish species will only show a response after their reproduction cycle has been com- pleted. Measurable negative effects may develop with a considerable time lag.

The copper pollution in the Fly River floodplain is of persistent nature. Even after the cease of deposition of copper- rich suspended load, the environmentally detrimental effects will continue to exist. There is no mechanism to 'de- toxifi' the alluvial plain. Erosion processes will only remove deposits in the main river channel and on the banks immediately adjacent to it, and, to a minor degree, along the channels connecting drowned valley lakes with the main river. The mine-derived material deposited in oxbow and drowned valley lakes, and in the floodplain swamps, will remain and may be covered by less contaminated sediments carried by the Fly River in the post-mining period. How- ever, when sedimentation rates in the floodplain return to pre-mining values, it may take centuries until copper-rich sediment deposits are sealed by an unpolluted sediment cover.

The possibilities to mitigate the detrimental effects of the Ok Tedi mine wastes on the Ok Tedi/Fly River system are limited when the construction of tailings and waste retention facilities is deemed uneconomical by the mining company. In order to reduce the amount of metal discharged into the environment, the recovery of copper in the mill, currently at around 85%, could be increased. This may be possible through installation of additional flotation cells, re-flotation of tailings, increase of the residence time of ore in the cells, improved control of grain size in the flotation process, sulfi- dization of non-sulfide ores, and other optimisation strategies. Thus, a recovery of 90% may be achievable. The cut-off grade for waste rock could be lowered which would also result in less copper reaching the fluvial environment. Measu- res of this kind are most probably not viable from an economic point of view; however, as an environmental protective action, they should be investigated thouroughly by the mining company.

ACKNOWLEDGEMENTS

The authors of the study wish to thank the United Nations Environment Programme for fmancial support of the research project. The study team is particularly indebted to Mr. Gerhart Schneider, Programme Officer with the Water and Lithosphere Unit at UNEP headquarters in Nairobi, who offered prompt assistance in solving all emerging problems. David Mowbray of the Department of Environmental Science of the University of Papua New Guinea (UPNG) provided invaluable support througout the various stages of the research project. The authors of the present study wish to thank him and Dr. Ian Burrows, Head of the Biology Department of UPNG, who provided working space and laboratory facilities during the field work in Papua New Guinea. Very special thanks go to Mr. Peter Gelau of Obo Station on the Middle Fly River, who helped us in accomodation and transport problems during the various sampling campaigns, and Mr. Ben Kapi Mekeo, our reliable field guide and boat pilot. We also wish to thank the people of Aiambak, Bosset and Manda villages for the hospitality they have extended to us. We also may express our gratitude to Mr. Joseph Gabut, then with the Department of Foreign Affairs of the Papua New Guinea Government, for continued support of our research project. We wish to thank Ok Tedi Mining Ltd. for analyzing a batch of water samples and further logistical assistance.

REFERENCES

Allen, G.R. and Coates, D. (1990). An ichthyological survey of the Sepik River, Papua New Guinea. Rec. West. Aust. Mus. Suppl. 34, 3 1-116.

American Public Health Association (1985). Standard Methods for the Examination of Water and Wastewater. 16th edition, Washington, D.C. 50

Applied Geology Associates Limited (1989). Review oftheflnal draft report by Ok Tedi Mining Limited - Environ- ,nental Study November 1988, Unpublished report to the Department of Minerals and Energy, Port Moresby, Papua New Guinea.

Busse, M. (1991). Environment and human ecology in the Lake Murray-Middle Fly Area. in: D. Lawrence & T. Cans- field-Smith (eds.) Sustainable Development for Traditional Inhabitants of the Tori-es Strait Region, pp. 261-282. Great Barrier Reef Marine Park Authority, Townsville.

Canadian Council of Resource and Environment Ministers, CCREM (1987). Canadian Water Quality Guidelines. Environment Canada, Ottawa (Ontario).

Clements, H.W., Farris, J.L., Cherry, D.S. and Cairns, J. (1989). The influence of water quality on macroinvertebrate coninlunity responses to copper in outdoor experimental streams. Aquat. Toxicol, 14, 249-262.

Davis, J.A. & Leckie, J.O. (1978). Effects of adsorbed complexing ligands on trace metal uptake by hydrous oxides. Envjrøn. Sd, Technol. 12, 1309-13 15.

Department of Mining and Petroleum / DMP (1994). Mines, Projects and Prospects. Quarterly Report April to June 1994. DMP, Port Moresby, Papua New Guinea.

Eagle, A.M. & Higgins, R.J. (1991). Environmental Investigations of the effect of the Ok Tedi copper mine in the Fly River system. in: Rogerson (Ed.) Proceedings of the PNG Geology, Exploration and Mining Conference 1991. p. 232- 238. The Australasian Institute of Mining and Metallurgy. Melbourne.

Eagle, A.M. (1993). Copper Mining and the Environment in Papua New Guinea: The Ok Tedi Case Study. Ok Tedi Mining Ltd., Tabubil.

England, J.K., Kilgour, I. & Kanau, J.L. (1991). Processing copper-gold ore at Ok Tedi. in: Rogerson (Ed.) Procee- dings of the PNG Geology. Exploration and Mining Conference 1991, p. 183-190. The Australasian Institute of Mining and Metallurgy. Melbourne.

Englund, E. & Sparks, A. (1988). Geostatistical Environmental Assessment Software. Environmental M.onitoring Systems Laborator'v, United States Environmental Protection Agency, Las Vegas.

Farrah, H. & Pickering, W.F. (1977). Influence of clay-solute interactions on aqueous heavy metal ion levels. Water -fir .Soii Pollut. S. 189-197,

Ferguson, K.D. & Erickson, P.M. (1988). Pre-Mine Prediction of Acid Mine Drainage, in: Environmental Manage- ment of Solid Waste (W. Salomons and U. FUrstner. Eds.). pp. 24-43. Springer-Verlag, Berlin.

Flemming, C.A. & Trevors, J.T. (1989). Copper toxicity and chemistry in the environment: A review. JVater Air Soil Pa/fur. 44. 143-158.

Hem. J.D. & Durum, W.H. (1973). Solubility and ocurrence of lead in surface waters. I .4m. Water Works Assoc. 65, 562-568. Higgins, R.J. (1990). Off-river storages as sources and sinks for environmental contaminants. Regulated Rivers: Re- search & %fanagerneiu 5. 401-4 12.

Jackson, K.S. & Skippen, G.B. (1978). Geochemical dispersion of heavy metals via organic complexing: a laboratory study of copper. Lead, zinc and nickel behaviour at a simulated sediment-water boundary. I. Geochetn. Explor. 10, 117- 138.

Jones, T.R.P. & Maconochie, A.P. (1990). Twenty five million tonnes of ore and ten metres of rain. Proceedings of .iline Geologists' Conference 1990. p. 159-165. The Australasian Institute of Mining and Metallurgy, Melbourne. 5'

Karbe, L. (1988). Interactions with biota. in: Metals and metalloids in the hydrosphere: Impact through mining and industry, and prevention technology, pp. 165-186. UNESCO, Paris.

Kare, B.D. (1992). Diets of seven freshwater fish species in the Fly River system, Papua New Guinea. Unpublished report to Ok Tedi Mining Ltd.

Kyle, J. (1988). Water quality and sediment chemistry of Lakes Basset, Pangua and Daviambu. pp. 9-18 in: J.C. Per- netta (ed.) Potential impacts of mining on the Fly River. United Nations Environment Programme, Nairobi.

Loftier, E. (1977). Geotnorphology of Papua New Guinea. CSIRO and ANU Press, Melbourne, 196 p.

Maunsell & Partners (1982). Ok Tedi Environmental Study. Main report - Environmental Impact Statement, and sup- porting studies in seven volumes. Unpublished report to Ok Tedi Mining Limited, Papua New Guinea.

Meador, J.P. (1991). The interaction of pH, dissolved organic carbon, and total copper in the determination of ionic copper and toxicity. Aquat. Toxicol, 19, 13-32.

Moore, J.W. & Ramamoorthy, S. (1984). Heavy metals in natural waters. Springer Verlag, New York.

Moore, M.V. & Winner, R.W. (1989). Relative sensitivity of Ceriodaphna dubia laboratory tests and pond communities of zooplankton and benthos to chronic copper stress. Aquat. Toxicol. 15,311-330.

Ok Tedi Mining Limited (1987). Studies of the benthic fauna of lowland (potamon) localities of the Ok Tedi and Fly River, with reference to mining activities. ENV 87-11.

Ok Tedi Mining Limited (1989). Supplementary Environmental Investigations, Volumes I-I!!, unpublished.

Ok Tedi Mining Limited (1990). APL Compliance and Additional Monitoring Program, unpublished.

Ok Tedi Mining Limited (1991). APL Compliance and Additional Monitoring: Annual Report 1990. Report to the Independent State of Papua New Guinea.

Ok Tedi Mining Limited (1993). APL Compliance and Additional Environmental Monitoring Program. 1992 Annual Report. An annual report to the Independent State of Papua New Guinea as a part of OTML's environmental compliance monitoring requirements.

Ok Tedi Mining Limited (1994). APL Compliance and Additional Environmental Monitoring Program: 1992193 An- nual Report. An annual report to the Independent State of Papua New Guinea as a part of OTML's environmental compliance monitoring requirements.

Parkhurst, D.L., Thorstenson, D.C. & Plummer, L.N. (1980). PHREEQE - a computer program for geochemical calculations. US. Geol. Surv. Water Resour. Invest. 80-96, 210 p., Washington, D.C.

Pickup, C., Higgins, R.J. and Warner, R.F. (1979). Impact of waste rock disposal from the proposed Ok Tedi mine on sedimentation processes in the Fly River and its tributaries. Unpublished report to the Department of Minerals and Energy and Office of Environment and Conservation, Port Moresby, Papua New Guinea.

Pintz W.S. (1984). Ok Tedi - Evolution of a Third World mining project. Mining Journal Books, London, 206 p.

Remade, J., Houba, C. & Ninane, J. (1982). Cadmium fate in bacterial nlicrocosms. Water Air Soil Pollut. 18, 455- 465. 52

Roberts. T.R. (1978). An ichthyological survey of the Fly River in Papua New Guinea with descriptions of new species. ,'inif/r 'antrib. Zoal. 281, 1-72.

Rush. I.M. & Seegers, H.J. (1990). Ok Tedi Copper-Gold Deposits. in: F.E. Hughes (ed) (ieologv q[the Miueral leposirs of Australia and /'apua New Guinea. pp.1747- 1754. The Australasian institute of Mining and Metallurgy, Melbourne.

Salontons, W. & FagIe, A.M. (1990). Hydrology. sedimentologv and the fate of copper in mine-related discharges in the liv River system, Papua New Guinea. Sd. Thral Environ., 97/98, 3 15-334.

SILoikovitz. F.R. & Coplanid, D. (1981). The coagulation, solubility and adsorption properties of Fe, Mn, Cu, Ni. Cd. Co and luLmic acids in a river water. Geoc/iirn. ('osmoelzim. 4 cIa 45, 18 1-1 89.

Smith, R.F.V. (1991). Biological investigations into the impact of the Ok Tedi copper mine. in: D. Lawrence & T. CansIield-Sniitli (eds.) Susrainablc Development /r 7,'aditional lnhthitants f the Tvrres Strait Region, pp. 261-282. (ieat Barrier Reef Marine Park Authority, Fownsville,

Smith, R.F.W. & Bakowa, K.A. (1994). lJtilization of tloodplam water bodies by the tishes of the Fly River, Papua New (iuinea .t fat lnternat. I ercin. I. iowal. 24, 187- 196.

Stiiinm. WI. & Morgami, I.J. (1981). ('hL'mi.ctrl. Wiley and Sons. New York.

I'ipping. F. 1981). Ihe adsoiption of aquatic humic substances by iron oxides. Geoihun. ( nroc/iim.Acta 45, 191- I 00 .

United Slates EiiironmentaI Protection Ageiicv (1986). Qua/in ('ritt'ria for Water. Office of Water. Regulations and Standards. \\ashington , D.C.

\Vjiiiamus, N.J., Kool, K.M. and Simpson, R.D. (1991). Copper toxicity to tishes and an extremely sensitive shrimp in reLation to a potential Australian tropical mining waste seep. !,uertr J Environmental Studies 38, 165-I80.

\inner. R.W. & thcn. H.A. (1991). ToxicitN of copper to ('h/ann domonos reinhardili (Chiorophvceae) and ('erzo- 'Jima lnbia Crusiaeeae) in relation to changes in water chernistry of a !'reshwater pond. .'tquat. Toxieol. 21, 157- 1 79. ANNEX

Tables Al to A6 -

-I-,

- - - -

0 - - 0 0 0 0 Ca 0 CQ 0 c-a o Na 0 0 C%JraNa Na c-a c-.j c-a

In en .-I In en en 0-arnn en 0 000 c- en Le, en In

1-. 00 en CO en 0 0 0 N-I Ifl CO 0 0 '.0 0 In In CO 0 0 c-I c-I In In N-

0 00 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 CO 0 CO 10 = 0 0 0

F-. CO C-a In en o Cfl 0 01 00 01 0.0 0 0000.0 c-I CO In CO C-a

O N- SO .-I c.I e -j c-I C N- 0000c-I N- 50 001 F- 0- 0500 05

0. 0 00 N- 0 In a, It1 CO In In CO In CO In In CO '.0 0 fl N- I 0, 01 01 CO

o 01 '.0 50 en '.000 OLCI N- COWN- c-I .-1 c-I O , 0 0 05 Cl Cl CO c-a a, so o - -I -4 0c-100 0 01c-I 03 01 -

InCO c-lI eO 03 .0 .0 .0 c-I ..-Ic--I 01 03 0 - .I

N- CC-I 00 0s C-fl In 05 N- a, a, 50 0 = 00.0 Cl CO 0 0 N- 0 N- In 1.0 In SO In 05 CCI LII N-SOLO N- N- In N- N- N- CC 0 0 CC- sc-, C-, -

~ -I- -I- c-i '.0 010 05 05 03 00 0 01 Cl CI N- 0, 0 50 00 In 1.0 en a, 00 Cl

000 0.000 000 0000 0 C 0 10 0 00 .1 0000

•0 . 000 0000 O 000 0000 000 00000 CO CI

0 50 50 '.0 0 01 .0 N- CO SO In 0c- 05 01 '.0 0. 000 50 00 N-

10 0

CO 0505 N-C-a05CO C

ON-mO - 0, In C 0 00 N- 01 50 In 010 .0' 01 N- .1 05 Cc-COO 0 0 0 0. 010 '.0 50 O LOInOLO IflOOl (-I N- a, .10 LO 401N-0 0) CC> 41 CO 50 0 CC> '.0 1.0 50 0 C!. COIn0 InCOInLfl ttlCOO 0500100

000000 0000 Cl Cl 0 0 0. 00 000 00000 0

enrICO 010005 In 00 Cl .0 50 In COol LII SO CO .0 0 N- CO CII CC>

IC 10N-I N-0050 c-SOLO .-i.-I0050 In o - CII CII CO - 05 00 000 .0 01 000000 1 c-I - - -

000000 CIlO 0 0 C. = 10 000 000 000 = 0 0

CI N- N- SO 10 N- 10 50 CC 1.0 0 C-I CI 05 50 N- CO CII N- 0 1.00 CCC 00 0000 000 00 = 00 0 00 = 000 0010 0000 000 000 00000 0 000000 000 0000 015000 InN-la, 00050 0N- InC CO0 .0 .0NrmCOCON- - .0 - - 1 - - c-I - - -I • g -gg 001005 '.0 Cc- c-I In In CII S 01 '.0 l 00 In 00' N- 05 0. 50 0 0 0, - c-I CO c-I 001 0' 01 00000 In N- 0 0 c-I Id 0. o .0 001005 um Cl. - U

In T Eq LII 153

C In

.0 r3 C- Fl Es 5

CrT c. 00. -oo'.9 -0OCO0 , '00CS3 2o1.o 00.0 300 ft u • a,- • o H --iCr' .-i''j '' CO CC cc 0 000 = 0CC 0-n 00 oo CII CO -I C'. NJ In 0 In 10 CO /'. 01 50 - CO CO _ s CO In In - In CCI IC> In SO SO - -

0 0 0 0 = o e en ma . .

-I-,

Li 03 CM - CM en CM r.5 0 .en c-I - ,-c en

- 0% CC 0% tO 0 CM 0% In tO 0% c-i 104 tO 0% 0 CM CM 03 CM Oc en en en en en c-a c-i en en - en .-I c-I .-I en - CM

en en en C%i 04 CM en en CM en — CM .- c-I en ..-i en ,.M CM

00 tO 0% CM 0% eI 0% CM 04 c-S 0% en 0% 0% 03 0% 0% 03 00 0% 00 . In CO 0 000 0 CM tO 0% In CM en c--i C-i 0%

0 0% 0 0% 0% CO 0 00 0 00 00 00 0 00 0% 0 0 0 0 0 0 0 0 0 00 0 0 0 0 0 0 0 -I In CM .t en C- 0% tO 0% CM In CM In CM 10 0 Ic-

tO In 0% tO 0 CM In 0% c-I 0% en 0 - 00 0% 0% tO c-i En en CM en en en en CM eM en en en CM c-i en .-i .- en en c-I

o In CM en C'l CM .- CM en c-I en c- CM en c-i en CM C-i C-i '1

O '0 0% 0% MO nfl 0% 10 lOIn tO In C%i en c-a en C-- CM tO

- •0 -4- - -4- Ci 0% 0% 0% tO 0% '0 0 04 Ml C-- MO C In CM 0 tO

0 Ca c-I ca'0 MO en C-a CO t-I c-I en (Cl 0% cc -+ 0% In '10 '1 '10 0 '1 00 0 00 00 0 ti 0 00

C I IV - . .

0% 0 tO 0% CM In en en C-. CM CM '

.0 01.0 0%

C en CM '0 en en en 0% '0 0% In en 0% 0% en en 0% 0% c-i b-I

0 'CC 0 a% oIneneno r rM 0 C-- 10 Ca - to c-i

gg g g gg g gg g gg U, In. 0 0% 1-0 Into en ro C-a en CM i-I 0 en CM CM c-I C-, ,-1 Ill

c-i c-I

In N-0%In1-OInMOCClM%InLC% 0% C 0% '0 00 '0 en C% in U

00 0% 00 0 0 0 CC 0 00 0% 0 0 CC 0% CCC

to tO' 1.0 tO 4-- 0% tO Ml 0 CM 01 Ml - en c-I Lc- en .0

g g g g g gg g g g.g

010% CM COO In C-i 0,0 01 C-1 01 c-i ene- in EL Li. 8 8 8 c) C) P> 01 8 8 8

I CM 0 '0 I In I CM 'Cl I C-fl CM CM i i C-c.

MS - C 10 c-I C%C C'S F=i -8 8 -

5)0%

0 c-• _0% 59 . c- - -

c-i en enen I-i 5--4 8 en -i i- j

• - - - I - - - 1.-I 0 014 c-I c-O 01 en enenenen c-Ic-c C-) Cc- - en - • - • - - en - - - en

CO .

.o o .o o 0 0 0 0 0 0 0 0 0 CM CM 0 en

o r- I00 in e., to CO' CM fli - C -

SO 0S CO N- 0 CM t- to C-I CM -I CM CO CD C.E a C, C', N- o a S. C' IC' C', In 0'. 0 N- N- - CM tO N- C IC' to to

CO Ca - - CM - CM CM <'4 CM CM CM . - CD CM CM CM CM CM CM CM CD CM CM CM

CO N-- - CD .- CM CM CM CD CM CM CM CM C< C". CM CM CM CM CM CM CM CM CM

CM to ,,'l tO 0 CM CO CM 0'. CM tin N- 0'. CM N-- 54 CM to CM tin to C tO CM 0, N- In In .'-i 0..'i CO - C- C- C- 0 0 N- to CM CD CO -I to CO to CD .D - CM C4 CD N- CO

0 0 0 0 0 C C 0 0 00 0 0 0 0 0 0 0 0 0 0 0 0 0 -4 CM to CM 0 CM N- Itt 0 CM CD to SO CO tin CM ' 'S' ' In C', to

o to C-CM '.0 C', 40 - ..-IOO CO CO to a a to tori IC, CO CO CM CD CM C-- C)

o CM CM CD CM CM CM CM .-4 Cl C-I to CM.-) CMCM CML'. CM CM CM CM"0 CM CM

0 CM ..D to .CO to tin 'C' CO In 'C' CM In CO CO CO IC CO )fl SO IC N- CO In Li'.

I 'CO C-li tin to

CMCMCM CM CMCM CM ,-I -4 CM - 0, i C'. .D CM 50 CM

' cl . . . .' .

'C' to to 'C' 0' ' 0' CM CM CM N- tin CM 0'. to 0CC 0' '4' '0CM CM to to

,fl a to CM In CO to CM 'I 00 a'. CO to in CM In .-4 0 CM CM CM In en a'. 0. 5' CM - I CMCMCMCM CMCM CD .'D CM "4 CM CM -I 'C' CM CM

0 CM .-4 C'4I CM LflCM C-, in tO CO tin CM CM In In 'C' to to 'C*tO In C-- 0 •

to 0, CM In C-- U'. 0 C', CM N- In '0 CM tO N- CO In CO N- 0t4 tO CM

CD CM to CD to 0

gggg '0 a CO Li'. 0' 0 Cl C 100CM 1 in a, a to I CD 'C' CO 50 CM CD CO to tO -to 0 CD 0 N- - IC, C .I i CD CD 0 CM to 0' N- N- 0 0 0. CM CM CM CM C') CM CM to to

C-.

000 CC 0 0000 00 00 CI 0 00000 0 0 000 00 0 0000 00 00 00 00000 CM 40 C-,

000 0 00 0 0000 00 00 00 000 0 0 00 000 = 00 0 0000 00 00 = 0 00000 00 to - N- 0 ('6 N- 0 N- 0 0 C', CD In oto CM to to CO 0'. N- CM tin 0 CO CM .10 CM CV'. IC, CM N- In CO tin CD '.0 CC N- N- CD to '0 CO C-- N-

000 0 00 0 0000 00 00 00 0 = = 00 00 o a 0 C 00 0 0000 0 = 0-0 00000 00

CD CC CM CO to in CM CM t-C CM tin C', CO In to 0' C-iC'.) CM N- 0' 0 Cl CM CM Ito a t-C ,-I rI ,1 Cli CM .1CM CMCM CM ,4,-C CM i CM

000 0 00 CM 0000 00 00 00 00-000 00 000 CM 00 0 0000 00 00 00 00000 0-CC S. CMN-In N- to C, to to t', COC'6 ON- CM0 CMCM CMIn0WIn N- to '00' nfl CCCMCM Cl tin CO C), CM tin C%I 50 CM it, a, 0'. 0' CM en In -0' Ca 0. 4.'

I C-Dan ('4(0 CM I I I in CM OdD

CM C) 3.tin CMLI'. it'.Ci

CO LI .01 0 CM 0. to 14, CO CO It'. CO

t,C •.'-< • CM N- N- .tOgN- 0.1CC. CMtCCCM I•"C C-IN- - C') r en • 1-7 - - ..-..-. Z! 2 HI LI, CM CM ('30 CM 0 CM Cl CM CD CM CM en CM to -4' In tin Li, In 2

en too a, c-4 o a, La en 0 00 a' .0 c-a en 0 a' C

La t200 2 2 2

0 0 c-4 - in a a, — as en 0 - C- 00 • a, tfl 0

0 0 c-S 0 en en en 0 0 — — en c-I

a, en en ,0 0 0, 0 OLa 0 en en to a, en .-4 a, 0 0 a, a-- aS 00

00 00 0 0 0 0 0 0 = 00 0 0 = 0 00 0 0 0.0 0 00 00 00 Ce 0 0 0 0 00 en en so in ho sO 0 en to en - -a en so a— 0 I- hO C

en en a, a, is-, inc., a, en in a en in a, a, 0 to C r-

en 005 c-I 0 en en en 'C - en ho a, en en to to 0 -

r- it, La 0' en a en o a, at c-i C Ifl n Ifl in en a, in in

n2, en C-. iS-a c-a c-a en o c-a c-a c-a en eq eq ens-c. in -a- .-a s-c.

-0 -5- V V cC V V V en

-a, '0'0 a

a' C- tO CO en 0 in r- a, en hO 05 0 a— to ho a a Cs en eq

a, 2

0 0o0 — c- 1.0 .- .

g g g g g 2222 cc, .-i en 0, fl a— -S in I-i CC I c-s en a— is, en is, to ho a- C-- en it, oO a, it, -4 c-a C C-a en en en en en a— a— en en a— en en in .-, - 0 000 0 0 00 0 0 0 0 0 00 0 0 0 0 00 0 0 0 0 0 0 0000 O 00 0 00 P. 0, 0 c-arc. en a, a, a 4-4 cC

2 2522 2 22 2 22 22 22 2 2 2 2 en in - a 0 fl 00 CC- it, CCI CO C-I 0 en tO en CO

CC- = 0 0 0 C) 00 0 00 00 00 0 0 0 0 2 2 2 a,, 2 22 2 22 22 22 2

0000 0 00 0 00 0 0 0 0 0 C) 0 0 0

C-c. ho to en 05 it, in eq c-a a- en cc, en 05 en a- in a- SIC c-i en c-i .-. en en 00 0 0 0 0 0 0 0 00 0 00 0 00 00 0 0 0 0 0 00 0 00 0 = 00 0 CI CC 0 0 en C-c. c-i on c-. c-a c-s c-i en 0 0. 0 a— in in in in

C- 0

en C-I C-I to tc. 0 I C- I

I en ft -

41 cececen 92 22 2 9 2 2 882 8. 5 . •

w to - CM • 0 0%

o 0% 0 0% '0 CM 0%'0 - CM - to to to ento In 0) rj CM a) .-.4 t4 CM - CM CM C') CM 4 C4 •-1 CM C')

0, In In CM C 0% In - In 0 CM 0% 0% ICC In Eg I-' % CM CM CM CM - CM 0% CM CM CM CM CM CM CM

- CM IC eM ICC 4CC In '0 CM N- 0% 40 0% CM '0 Lfl 0% CM CM C') CM CM - CM C') CM CM CM -4 CM CM tfl ('4

ICC C AC' AC' In co 0% N- 0 CM 0% 0, CM CM *0 0% a 100000 N- An 00 C- 0' CM 0% 0% iCC ICC 0 14% - ICC 10 FM -

0% 0% 00CC 0 0 0 0 0% 0 0 -CM CMCMCM CM - 00

- - - - - (_4 - •-. .-4 •-4 CM CM C') CM CM C')

-4 Ifl C'. In 0% In CM 0 C-- CM tO 10 q1 ('4 ffl CM 0%

o Cr410 C') C-. CM 0,0% 0 CM In 0% CM CM eM CM 40*0 - CM C-') •-I CM

V CM V V V V

C P-I Ifl t)e CM CS) In C) ICC 4010 14% ICC ICC CCC C-- '0 10 CCC 0% 40 rI CM CM CM CM

Zr - V 0

CM 00 en ('4 CM 0% CM CM en 0% ('4 CM 0* C')

-4- .0% A 0% 0% • . .

0% CM 0% 0% 0% In 0% CM CM CM CM (-4 CM C) VVV 0 0 V VV V C V 0% 0% 000 00

In CM In Cfl CM 0 0% 0% 0% CM W 0 0 ICC ICC CM CM N- CM CM 10 104% In 0% CM ' 4% C') In C-p

0%' C'S 0% CC CM 0% 0 CM C-fl N- CM 00 '4' ri 0 In ICC 0% CM 0 C') CM CM 0% CM e-I CM 10 10 CM 0% 00 - CM - •-4 - - _I CM •-4 -I .-+ - _I - CM

C') '00% CM CM CM C') 0Cnr4In r 0% 10 CM C'. CM tO 0% .) ..M 10 CIn 0% 0 40CM._4 0% CM C') . -1 CM C') CM CM CM CM CM e-I CM CM

0 0 888 8 o 88 8 8 2 8 8 282 28

0% a-I en ci - C-fl CM CM Cfl CM CM CM ('4 CS) CM CM 0%

An CS) 0% CM CM ICC - CM CM ('C 0 40 In In 00CM In CM to ('410 .'4 0 CM 4% 0%' CM C'. 0' 400 -I .0%

0 0 0 0 0 0 0 0 0 0 0 00 0 00 000 0 5 5 0= 0 0% 0 0 0 000 00 40 1040 C'. CM N- 1-- 0% CCC tO tO ('4 CM eM 0' In C-C CM

to to 10 CM 0 N- 0% 10 C-- tO I CM - 40 In

000 0 0 0 0 0% 0 0 0 0 0 000 00 000 0 0 0 CC) 0 0 CS 0% 0 00 0 00 0400% rI CM ("I In CM CM en ICC CS) CM 'CC 0% C-nO 0% 0% 10 LCI 0% ICC 10 tO Cr4

000 0 0 0 0 0% CC 0 0% 0 0 000 00 000 0 0 0 00 0 0 0 0 0 000 0 0% 0% .M CM 0' tO C') On 0 0% 00 0% CM tO 04 CM CM 10 .0 ft 0% 43 AC' • CM C.) C) CS) AC' 10 40 CM 40 C.)

40 CC CM CM In S CCC AC' CM 0% In 0% C'S 4•1 0%' I 0% 0 C') C CM - 40 0 CM CM CM CC • CM CM CM 0% C_C - 8 _ _ 00 *1) 41

C--

CM SM CM PM 4-.) 4-') • • -

0% 40 .0' - - • 0 0 - - - - C') CM 0% 050 ICC ICC ICC C') C'S ICC ICC 54% ('4 ('4 CM., '.5') CM CC 0 0 0 00 0 0 0

in n to a

C) ¼OC to t' C 0 03 E- 0 0 Cl Cl 0 tfl 0 0

in Ifl in 00' = 0 in 4 Cl Cl 0r t- r- 03 00' La 0003 CC P- o in r 4 0 C — 0 = 01 Oar- or-. C,

7.11 I zi,

Elm

I; 00 0 C- 000 00 00 0 - .3-I (fl

4000CM wo 0 .00

—I flJ - -I >1 .0 In 4000 0 00 08 00 CM 04 (44 00 0(0808 '-1 W4-' (0 000 CM 0' 08 40 .4 00 CM CO 0 40 00 CM On CM CM CM (-4 CM CM CM CM Ci Fz.i.0 4J4 -1 0. - 0,0404 CO 000r 000 4000 CM CM CM CM CM ('4 CM (-40 ('4CM w z q-i -.-I 0 o cjI C-- '0(43 '0 F-1D WIJ .- W4-) 0 S U, 000 0 000 0 = 00 4(8 0(0 g C-CM40 400 r 1 1 4 (fl - . - - 00W

40 0 - 05 .-40 00 CM 40 0040 U, 00% 0 (44 CC (000 0 0 '004 CM CM r1 U, (d 0 't$'ci 00

-4 '0 eM C- C-M 40 CM *-t en 0% 0 CM r, 4i _ CMe,30 0 0 (4 o. CM 4 CM C-i SI-) w £0w

00 0 00 0 080 00 (a 05 00 00 3-400 o (-4 E - 4i.rl WItS 0400 o (-4 CM (-4 eM ('4CM (83 (83 CM CM U,

9j r- CM 40 CM cfl'4-i WU, 0 0 40 C- 40 4040 C-- n10 to (0 CM 40

'0400 (a 40040 00 - en CO 00 ('C 00 Cc V SI V V V '1 V SI SI jn9 to w 46_I 0(430 0000 -.- -m - (--00 i-I WU) U) '0403- 0443 4-' I,0 -0 4000 00 0 r-I 40 00' ('3 CM 3M OP ZIPi 0000 -40 fi q-vo 0

.00 00000 C- 0040 C- 40 -_ CM CM 000500 UJv 0. CM CM CM CM CM CM (83 (84 CM CM CM 046 U) : 00 04 .40 40 4-4 i4J

4-I CM 0000 .1-4-I-I r, 0 (453-

wwU, 84-) -i-I -.-I -I-I I1 0 in >1 000 0 000 00 00 000 0 000 00 0 00000 U) Wr-I 0540 CM - CM -00 CC%08 00 (a 444 en 05 0 OC.-*0 40 CM 40 CM Wd N .6 nOn CM ('400 CMfl 40 00 U)-.-' i:: 000 0 000 000 00 00000 00 ('C 4000

(43 0% 40 (0 (44 CM C-MO In CM In In n C-- 00CM 04 o -c-i

000 U)-.-, gg 0 050 000850 0 CMOS 00 00% 004 4646 (a (444000 40 0 44 4440 40 (44 00040 -1-31 a) 0 r-I ID t3

0000 0 000 00 00 (00 0 0 00 0 00 00 0 00 00 IC CM (OrlO 00' 000 0 0.4 40 444 40 144 0040 4040 - In In 444 ('4 4(5 40 00 40 '0 U) 8 000 000 00 00 08000 l-1 000 0000 00 0000 (4 (45 fl Cn 0) CM 0000 00 080 9-1.0 0000 CM 04 08 040 - CM 0004 - CM 1-40.1 (1.4 1l .1 0.4.4 CM CM .4 CM 1 04-) 0 - 00 0 000 00 = 0 In 400 46 000 00 00 0 00 00 C) 00 CM C- ..-4 004040 0040 CM - 400CM -.-I0 .6 00. 4) —I (0 P 8 >t4-4 1-) 0 8 00 8 8 r4rI -4 8 8 0040 In 3) 4 CM 00 co8 14: 01 -31 >t t0 W91r4 -4 a) :: 5 8 C el 43)0 CM -4 ii 0 CM CM 14:):: • ft -- Cc Cc , 0.4In 00 (a woo .-1 i -.1 0 00 0 ID 00 -8 1 (0 -00 CM 46,i OH --- _- -- '000 0) : 0 g ggg

9 0.'Ej .00to 0 0 0 0 0 0 04 0 0 0 c--I 0 0 0 0 04 0 0 c-f 04 0

o 04 r'f 50 40 0% 03 . to to 03 50 tO 50 0 50* 10 as C- 40 50 03 'Z 04

En am-tb I- U Sn to to to to to 40 4040 to 5003 In to 50 Sn

0 0, 0 0 afl c--i 0, 0 In En In 0 C I 04 to 0 0 - to 0005 10 en to to oi r- .0 .0 tO 05 - 0, 03 En en c- 0 - en C-, In -a• CM 1-IC C'4 tO tO to am 40

10 0 t- tO 03 00 c-a - to tfl- 0. .0 In CM 0 05 0 03 05 I-I .5

0 00 0 00 00 0 0 000 0 0 0 0 0 00 0 0

- - .-i .-i - c-i c--i •c-••

- C-a r- - a, 03030 In C C- - en C-, en Ce-Ic C-S 00 r to

C- 04 o CM 005 c-I 500 0 c-I En La 0 0 .-i 0, 0 *0 -0 50 CM 50 0 ** La 50 CM to am -evo -e Or, c-s -e c-s 04-a en oc OeM

S-I - 0 03 0 0 .1 50 40 03 to 030 0 10 .04 40 to 0' 4-3 03 03 O 05 I0 c-I C-4 . ('40CM 05 c-s CM - C-S - 8M CM CM CM CM CM - -

03 .0 In 10 000 0 50 to to en to 04 C-- CM 05

- to - CM 0050 ' .05000 In C-S c-I 0' 04 CM to 04 C- 03 in CS Os In - - cC 04 C-C C-- 0 03 i-I tfl .04 c--4 - i-I C c-i .4C4 c--I c-i

ItS to CM to 10 to 030CM 0 0..-i 00 0 10 0 10 50 0' - to to

o CM CM to C 1.-f .01 to (MO 00 c-i 0 CM CM CM c-1 0 0 .04 00

0 03

.0 50 0. 03 05CM 03 CM 50 I-I en en 40 CM

c-I c- 50 Ofl to tO en 50 - C- In 04 00500 3 10 CM 05 CO to In 0 0 40 tO 0 00 05 tO 50 .0 05 03050 03 0 tO en c--IC 10 c-I 04 101 1-n - to to 40 0 I-- c-n tO 03 0% 50 to 0 en en In c-s en In - . .04 c-I i-i

00 e -0.01 0, 0 0 to C- 0 c-I In c-I - 05 OEM en to - tO a, &n In n a' C e a' In c-n In In 40 I--C, c-I C-taO 0, 30 c-i 10 - 10 1i0 C-- 00 c-I C- .-I en In - c--I 03 - Mm C- en 40 05 03 to - 0CM_I

a' 00 OI ItS c-I CM 1 In 5 030300 In 00CM 0 0 C-a to MS to en *0 CC to C-a tO 0 tO c--I 40 In In 01 c-I In In 05 C em 0003 - d c-i .04 .04 03 010' 0 CC 05 00 C- to 10 -0 en 03

0 00 g•g.g g g gg g ggg

0. - as -e In I c-I 03 to 04. in ('a to -0. 50 .0 en 0 In (Mol .i 0 000 00000 00 0 0 00000 0 0 CC 000 00 0 00 0 00 000 0 0 0 0 0 Ifl I-s to as La a, 1--a' 05 to C- 05 c-tOO to MS Ml (MC-- - en to ItS 535 ItS 10 tO 010 IntO In 50 SOLo cc 00 0 00000 000 0 00000 0000 0,00 00000 0000 00 00 = 000 C 30 Cl 0, tO 00 In to to c-i 03 c-I c- to to tO 00 11-5 0305 L0 to a'

0 = 00 00000 0000 000 = 0 0000 0 000 00000 0000 00000 00 0 0 0' 03 03 00 ItS 1-4 0 C- tOto toto 005 10 00 I-S C-- 0 40 0 aS CM 00 535 to 05 11-5 03 .0 ItS C- 40 10.0 0 Mm to to -a- C-- 40 10 10

0 000 00000 0000 0, 0000 0 0 00 0 000 00 00 0000 00000 0000

(530CM CM .4 c-i OeM (MO I-S CII 004-aCM en en en em in

O 000 00000 0 00 00000 0000 40 - 40 IM CM 00 0 4-5 .-4 50CM 10 05 Ml tO In 0' MS .0 01 CM 04 10 03 0.--. tO 40 001 04 0 00 03 000CM (0 c--I to tO 00

q EL C 00 - 00 004-10 0.-4 CM OOCMO 0

71 g

-c-i: tjr (00 C-, — .g - r- 0 0' N- Sn SO Ci N- SO N— CI a, Sn 1.0 0 Its 00 E- 0 H 0

Id c-a In Oh WI SO Lfl SO 11 0 10 (4> 1.0 '.0 Sn Sn c-a Cl c-i (Cl C_I (J)r1 Q4 4i 040 s-I

N— Ofl -i 40>0 0 Oh 0 0'.tC (Cl c_S .-4 Oh WI-i

0 CI CI 0 0 0 0 CI 0 00 0 0 0 Sn 00 — o o 0 CI 0 0 0 0 = 0 0 0 CO In 0 4-1 0 4-4 U]

0's (OH

c-iflnqn C_I CII OH — (0 C_I 0 CI CI 0 CI c-i CI

a) CS c-i CII C_I .-1 c-ic-I — SO i fllc-_Ifl 0 — 04 (OW - t Oh 5th 0 i.W c-I rIO *4 .4 c-I (€O.Q

-1- -4- -4- 0 U] 0 000 00 0 000 S0 1.i (0

0' - -s-I to

.0 CC. a)4J .10 c-I 0 0 c-C

Sn c-h ICC_I a, n a, 0 CI CO C-- 0c-I (0 0> I-)

(0>0(0 -CSn N-Oh-ISO Cl0hc-C

CO >0(0(0' 000 CIa 4(0 Oh Oh a, 0 (00- C_I -I In 1(0 -I C_I C(0Cfla,C%3 In 00 N-- >c'I

gg g g 5-I CI 0> 0 c--I In c-I 0€-I 0(0 — Oh C_ISO N— Oh C ci) CC C-I 4-> (Ci a, 000 C_I 0000 a, r-a

O C> 00 000 00 0 CS 00 (.0 '.0 (Ci Oo N-- 4-i (OW (00 0 CI 0> (00 0-0 0 40 c-I 0 0 4-I> s-I-H

0 0 04 00 00' C> = CI 0000 000 (0 0 CC 00 000 00 0000 CI 1.0 €4> 00 N- so €n (Ci N— 1.00 1.0 CCI Sn (4> (0 SOc-I Sn

0<100 000 00 0000 0 (Ci 0 -H 00>00 000 00 0000 0 CC> 0 CI Oh r1U]

Ci 0 Ci Ci 0 C_I C-I C_I a, 01000 CI C_I (j

0000 000 00 0000 c-I 0' 0 0000 000 00 0000 Cl Sn 50 -H -H CC 0(0050 00>0> 4J( >iS-I C

(-4 0

C-I (I, E Fl-I C> ("I-li -s-I CC Oft ft CC < ft (a, 1jcoc. 0> Coa, C>o > 1.c_.-f I Ic- 0(0 e" C II) r14J 400

I <1 E1 0 0 = 0 •0 '0 '0'C 'C '0'0 'C '010 '0 .- I '-4 '0 '0 '0'O 'C 0 - _

t— en eq 00, - •0 'C 'C '0IH '0 'C 0 0 0 0 0 0 0 0 00 0 0 0 0 0 0 0

t eq 0000 0 — — eq eq c 4-.) •-4 eq eq eq eq

eq fl4fl&fl 0, I0 eq eq In CM 0, 0 '0 0, 10 en 0, 0 0, I — eq eq eq eq -e en eq en en - en In 40 'e en to en 0,

.-1

lea - eq 010 '0 *00010 en C, en eq -e C— In .-% In '0

0 0000 0 000 0 0 0 0 0 0 0 00000 0 0 000 0 0000 00 0 0 0 0 000,0,

eq eq

0' *0 CM CM 0,

'C '0 40 10 C- 0, 10 10 40 0 10 40

eq - - eq 04 eq eq eq %f f

10 C% en

0 0000 0 0 00 0 r-1 '-1 V 00 0 0, 0 St V =

4- -4- 4- & '0 '0'0 'C '0'0 '0'0 'C '0'0 0 0 0 0.0 0 .0 .0.0 .0 eq .0 eq CM .0 .0 .0 .0 .0 .0 .0

CM CM CM (*4 0% It CM CM 40 0% '0' 'a' (0 0% 0' — — %t 0000 0 000 = 00 '1 00 0 V ' V 0 V

2 eq —

2 2 22 CM 04 eq — eq 04 eq 0, *000eq0,eq.-'I .-, eq CM

0 eq c -

0% f-

g 2 2222 22 2 22 2 22222 CM eq 0% *0 0 CM 0, CM CM 04 eq 10 C- 0, 0 0 In 0 en (0 0, '0' '0*0 en fl 0, IC fl*0 .0, C-) 000 0, 0,00010 0 2 2222 2 2222 22 2 22 2 22222

0 0000 0 0 0 00 00 0 00 0 00000 2 z2 nnCMeqne%aInneqC*4In In eq —

0 0000 0 000 0, 0 = 0 00 0 0 0, 000 2 2222 2 2222 22 2 22 2 22222 2 2.222 .2 2222 22 .2 22 .2 282.22

0 0000 0 0000 00 0 00 0 0 0, 0 0 0 = 00,0 0 0000 00 0 00 0 0 000 0 O .0 0,. eq eq 0404 — eq 04 .0 0

M 0 0, 0, p 10 CM 0% CM

ti U) - 0, 0 0 II) C-

- - 0, - 0, 0 •- eq S.. eq

- C.) C'.) In CM eq eq eq MM n 0 In '0* • . eq - . - . . - • '0 10 '0 *Or C—r-r r CM Z en en to N— 0 tO *

o LaO -. — 0' 4,0 tO On 00 0 00 00 tO •0 .4 — — — en — — — en

00 2 2 0'3 en — en en en en r — en en en nenOn en 04

00 2 2 2 00 40 .04..O..4 0 00 04 tO neOn 40 00

- 004 O 0040 0000 an 0 00 tO 00 tO 00 N— N— In to 0

N-- 03 00 C'. 41 C', 00 to In tO Lao N-00 0 0 C' 00 C'. CO

O 0 0' 00 00 On 0 04 0 0 04 04 04 04 0 00 0 0 —4 22222 222 22 22 2 2 2 2228 2 8

en en en to 0 C- .-I 0n0 C'. 0 On t-'4 040,0 — .- — .-4 — — — .-. en .-I •-4

04 .0 N— to en N— 0 en 00 40 Itt 10 40 IC) '0 0' 40 4,0 -C IC)

C! 0 '0 0, 0, fl fl 4004 CI) N— 40 4- Itt OOr-O, -1 to

04 0 0 0 .-4 In -0 4,0 t,0 to .-4 '0 tO 10 C-4 N— 0004000 C- O

00 01 000e4 00 03 O'.04 en — to 4'. 00 CI, C It) C'. 40 to

00 Ct tO 0 000' 00 0 0 10 000 C'.

04 ft en 0 04 0 04 04 04 04 04 04 04 04 04 04 04 04 04 04

en en

22 22 2 2 2 2 5

04 #2 222 2 2 2 2

CO to Ot 40 rI N— t COO en .-. 40 CO 10 to 0 10 0 04 CI) ¼0 00 to SO N— 4 N— 40 40 4,0 1010 04 04 222 282 0 S S 00' .0 COOl It N— '0 en tO 0 01 10 .0 0 to 10 On en o 00 N- - 14,.. 'CO ' 0 10 N— 40 en.-.enen en nO C'. Cl 0 en

O 0 040 00 00 0 0 0 0) = 04 0 0 40 40 40 0 04 0 04 0 0 00 04 04 0 0 0 04 0 0 0 '-a 52 2 2 5 2222 2 .00 CO N— 00 0000 00 N— 0000 an C-- '0 00 C--. N— N— 01 N--

22828 282 82 22 8 8 8 8882 2 2 222 5 C) 0 CO SC) '0' Cl — 40 CI) en 40 00 00 C-) 0 00000 000 00 00 0 0 0 0000 0 0

22222 222 22 22 2 2 2 2222 2 2

Cl 00 Cl 4-' CI) 04 Ca .-I .-( n Cl enennen COt'.

000 000 0 040 0 0 0 00040 0 0 000 000 0 00 0 0 00-00 0 04 N— 0100 tO 0 tO to en 4 00 to -0- N— C-i C -40404 en C. 40 00 C-- 00 04. 00 C, 4)1 01 .-4 0104 Ct C'. 0 10 It 0 CI) 0400 N—

00 04 _ 04 C) C) C) tO

2 I Al C-4 C ST I - In 0004 04 0) 04

• en 4I - .0 _ Ca r La 0-C 44 0-O,44 C) • • - . .

0' o 04 o ' a sp tO - 'C I I , _ • t-. .E 2'- .a 41 00 0 0 = 0 00 fl 0 0 0 0 00

. —

0000 00 00 00 04 10 00 _ o ' o c0 -

00 04 04 00 00 00 00

I...

-. '0 0..4 10 04 00 05 00 00 0000 00 05 00 0 00 0 00 0 10 0 SO Efl '0 40 00 — 00 In Ic- 00 c-- 0000 SO 00 SO SO In — * — — 00 00 00 00 00 c-a — 00000 0000 00 00 00

0 05 0 00 0 sO 00 Ic- En — Ic- — 00 10 1000 CC 0040 00 p.4 00 eq c-a r U 00 00 00 00 00 c-Ic-I C%l 00 CSI 00 00 00 00 0 (4100 00

I-I — (4- 0 — nO asq CC CC — (41 04 1(4 C41 00 04 0 00 00 05 1(4 04 fIn 114 Sc- 0 0 10 -! '0000 400100 4(4 40 05

00 000 00 000 00 0 0 0 0000 00 0 0 CD 00 00 0 000 0000 0 0 00 0 0000 CD 0 0 00 00 - 00 00 CC 0 — Ic- Ic- p.- 10 C'4 04 00 04 04 05 I- CC 0000 F-. 00 ('lIsle, 0000000 p., 00 00 -4 en eq eq esa C4* ('1 0 00 00 00

-4 04 00 4(4 04 .'-SO' 00 CC c-I 00 eq in o 00 00 41Cfl 04 00 1(4 p.41 00 c-I — — *4 .10 — 0 00 .1 (41 c-I p.4100 0400 (41 41sl (44 (41

o c-I I(450 .40000 CC '0 10 Ic- 000504 00 0 00000

o 0 '0 40 04 00r4 (41 00 10 00 00' 00 eq tO p.- 00 00 1(4 00 00 (41 (Is

00 00 ('1 '0 00 40 000 00 0441 001(410 -. c--I 00 0 0 0' 05 O 41 (IsO '' Ic- '0 10 10 00 In ic- 0 'Si Ic- 4(4 00 CC SO In 1(4

• a ~ a q. 000. (Is 0041 0 0 00 1(4 (4- c-I 000 00 00 0 (41 0Cc-I0

4-4000 .-40050 0 05 1(4 40 P-c-lOrIs 0 00 10 00 .-4 0' ('1

-4- -4- -4- 0000 000 0 0 00 0000 0 0000 0 0 0 0 00 0 0 0 0 0 0 0 0 0 0 0 0 0 0 00 00

00000 0000 1(4 00 '(4 0441 0 1(4 0.14' 0 (41 1000 in .-,ICD0 0 .-4 -1 .100 00 00 00 0 Sf 0 0 5/ 5/ 5' 04 00.0-0 0

CO (41 0 100000 ' 5000 0010(4- 04 10 p.4 ('44-4 05 0 .0 05000 000.0 05 50 c-4 50(41 00 0000 1000 00 0050 c-4 00

(4-000 000500 50 (41 nc-i 0In .10 (41 -40 ICe.! (4- 00 50 000 0 0 04 0000 014 ' £0 in 10 00 o 00 1(41(4 Ps 00 P.- 00 00 141 50 en .10 00000 p.1.10 -4 C'- 1(4 50 (1

05000 00 0 OP. 0500(10 Op.- 00 00000 00 In (4U04 '0 00 110 00 40 00 00 00 10 00 10 0 10 00 0 115 '0 e'l0

10 00 00 (41 (4- '0000 04 0 1(5 CC (41 1(4 (Is 00 0 — in (41 05 (4- (4- 00 01000 000050 04 10 (41 1(4 '0000 10 1(5 00 0' 441 0000 00 (4- 0 SOel 000010 fl' c-f (nO (41 00 1(4 I/i 454 00 (41 10 .10 (Is 00

0 000 0000 CD 0 00 0 000 0000 0 00 00 000 00 00 0 00 0 0 0 0000 0 00 0 = 0-0

as 000.51 10 0 ..t 00 (41 'CSJ 0 P-Ic-CO 'O 414' 10 451 ((4 1(4 0000 gg ggg g gg

0 00014100 CD 05 P—P-P-CC 50(4-P-CC 05 00 (511(4 040500 05 05 (4-

00 10 '0 c-f P—SO 10 500 5000 00 04000000 O '0• p.- .-4 1010100 — e-i a .-i

00 g 0 0000 0 0 0000 0000 00 0 0.000

00 Cc-f '00' 00 I-I c-fl .-4e4 10 00 (4. 05 00, ('5 00 ("I 00 00 05-00 c-f 00 00050 50 410 C4 50 50 00 51 tOO 50 0015100(4-

0000 000-0 0 0 00 0000 00 0 0000 0000 0000 0 0 000 0000 00 0 0000 00 (4- 00 En '0 (44 00 (10 r-. C'-. 0t CI.110 0r-. .-.'I CflLCIO 10' 0050005 (4-000 00 40 ('4 1(4 00(4-0—P.S 0500 00 0 04000- a .-i a .-. .-1 — — 04 — ('4 ('4 g 05 '14' .10 0 05 000 0 0 000 — 50 0 Cs 054-4 0 00000 00 I14 0. 00(4-P- 10405050 -4' 00 00 00 00 00 10(4- En 00 934 0 0

• 0 0 00 111 O 10 I 4(4 0' • 0 0— 0 esa es1 ' 10

00 00 (0 10 04 (.4 (.3 (.4 om 49 -s0.-.!0 .4000 0. 0.0 • cw H' - •40••3 • 1.4 o •- -4 I W

C-, o • :3 :::

o - Cl 0 0 5CC 001000 CD cc cc SC .'I t 0 01 = en 0 en 0

t- a, en - - -0 to - In - -e to 000 0 Ifl 0 C1 In Lfl - en en

cc - In en en In - - -. en cc .c -eccen n en cc to -

0 0 00 0 0 0 = 0 00 0 0 0 0 0 0 0 0 00 0 000 0 -I

C'S 0 C to cc 000, CCC' c-I 10 en — — to c-- C- cc 00 cc In - en 0 0 0 Cl tO C'S Ifl en In CU is-I o CI 0 cc cC en 0 .'-I c-I 0

- en en 0' 0 to 0 CI t- 0 Cl In

tO '0 0 0 en en 0 0 C- CU t- c-I t- C- ic- to c-c a-i en '0 CI C-U CC C In o

a- o en en en to in in C, In a, um C 0' '0 -00, cc -'0' 0 0 en c-a in in 0

c-i en -I - SO s-I 0 0 nfl IC) en in 364 as 0 0' C-- en - 00 Cl C'l In 40 an a, en C) SO to In en In 0 to -'0. '0 La 0 Cl s-i cc en in SO C en rs n'S' 0

•1 S - •0 4* - cc en so .- c-a no s-s en s-a en en In 0 '0. -I r- en to s-i in 0, a 'U'

0 C'I LaCU -0.'- C'0e 2 C OC. -0 i-leO 0 i-'I 0 ,1 .4 00 c-i CC .-n - 00.-l000 00 0 0 0 en

•g ' - - - - ME

in in Cl cc CC C-I Is, a cc C, en 0 01 -aS Z- 0cc cc en cc -U- isS Cl cc to cc

po .0

cc In 00 CD In en 0 C, en c-a C, ('S cc s--i s-n en C, so cc en c-I 0.-i In cc ..U' In so a,' s-n to is, cc cc c- 0 to en c- en 0 0 to en CO en a, 0 s-I - en I-- - C in en cc a' o cc .-d Cl as so cc In Cl ai cc in in - c-I 0

O0C 0 c-q n-I 0 en C'S C'S en c-S c-I 00 en s--n s-s en c-a c-a no -n-s-son ens--Sen s-i

cc ens-lOs-na' 0 CD 0C-In 0ts-00tSS cc c-i cc to SC is- cc en a,en.,SoinSflLflcc

gg g g g 8ggg

-4 I-n, 155 s--icC' C "U' s-I in 0s en cc to cc -a- cc en en en c'a 0 155 CD Cl 155 cc - LC) 00 en a' Cl s-s CI Cc-i .0 C) CI C en 0.0' an tO .-i n55 en Ifl en s-s en o .-s '-a cc s-- is-I so cc C- C- 0 is'i nz, Ia-i Cl

0 = 000 0000 00 00 0 000000 000 = 0 000000 00000 0 C 000 000 00 0000 0N- en Cl CI 50 50 tO tonC- is, to Cl 0 0, 0 .-I CC a, in a, a a, N- N so to is, so Os CO

000 00000 0 0 000 g 000 00 0000 00 0 0 0 = 0 00 = 00 0000 0 0 0 000 0 0000 00 en en 01 C, en a, N- en 0.001 O CC- cc s--I 030 in en In-C -aS 0 C-i to cc in in c-a CC so c-S C'S a, en en N- ,-IC cc en c-a en CD n-I CC 0 in en c-d -I C'S c-S en en i- cc 0 .-I C'SCUC'SCU0fl OC'SC' C-,

ggg

. N. - en en - en to Cs en s-s en

0 I CC CC en O I C) I 0 cc en I _I_i •I C)

0 - 0 0 i '

Zle CO 0 OC C' o -.-s Ia stiis en s- occ- eio Zo en ft en _ 01 s-0O C-I 0.0 - 00

= C - - 00 e 0 00 CD CD CC 000 0

N- 0 0 0 C'S Is, U-i in 01 In ifl 0 to en en en c- en Co :

'-I CD C_I In C_I 01 0 en In i .-I .-d .-.4 0 In In C.Jr.IflIn.-1 0 IC) In C-Th CD CD

C_I 0 — o nj — .-.d C_I CD 0 CD CD CD In .-i Ci 0 CD ..DCSI0.CIn

In CD 0000 0 00 IC) 00)_I In C-- 44) 00 C_I Co CO-e-Cr-- CC_I0n

0 0 0 = 0 0 0 00 = 00000 0 0 0 00 = 0 = 0000 0

00 0 10 in — 0 = In — Ifi C_I CO 00 0 ' 00)_I .IIIn 00 0

00(51 In Co C- 0 0 C In ' o cm 0 C-. to — C- rI 05 0 10 0 IC) IC) CD C_ICC_I

000)_I In r- In C- 0) I-I 0 00 C IC) to In 0 0 o C_I 0 to 0 10 — '4) 0 — CD — 0 C'a In Co — In C_I In C_Il-I 0 CM '0' aD '0 C1' In

0004_C C, CD to In IC) .0 C- I 0, CD - C- 05 0 05 CD 0 i4)'sO Ln C-. In 0(54 '000 Cg0 In In IC) 14')

0 'CO 10 00 10 00 0 - C_I CO — C_IC_I t'4 I_I 0 0) 14) 'COIn 0 C_CO

..D1000l-)I0C-IsOIn(4) CD 104_COO 00 000) D.14_11D0

-4-

- - . - - .0.0.0 0.0 'CD a - - 14 14 14 14 14 14 14 14 - CD 0 0 14 14 0 14 14 0 14 14 14 14 14 14 14 14 14 14

14') (54 0 0— 1)1)0 to 10 0— C_I CD 44) In MD C_I CD 00 000O00C_I — — 0.-IC 0 00 0 CD 0 CI CC 0 00 0 0 0 00 00 00 0 0 0 0 5'

-a

10 ..-I 4'-) - am 00_I 0, 10 '0 4_I C) OS In CO Co CC ..4C4 50 0 10

0 CD 0 C'-. 0) 0— 0004_I 1)11)1 CD 0000 14)0 0 10 .-.4 C-- 10 00050)

0—CD C- Cl C-- 0 00.-I '0 0 0— 1)1 10 00 0 0 0 CO CO CO 'O 0 0 0)0 C") 05 ., C_to C'- 0) to .0 11 0 .-I Mi '0 .41• 0' (51 '0' on ,-. 0 0 .-.CoCflCnI CD 14 00 000)05 tO 14) 00 CC) C-. IC) '0 to 4'Cfl COOS In SO tO 0 0— CD (4')

S 0) I_I 10 (4C 0 C-- ICC 0— 0 tO tO 'CC) tO 00 00000 01_I fl_I C-Co CII .-I 1)1 '0 0 ,-4'0' SO 0 0— CC IC) 0 0— 03 CO to IC) to 0 O 10 000 CC IC) C— CC C1OCDCSMDC1 In C_I 14) 0 CD 0 14) 0 00000 0 Co C_I In CD C_IC_I 0 C)

g ggg 555S 5 5555 55 ggS55S 5 CI') 0— CC) C-. CO CO C- 050' 0— am CO 10 00 0 0' 0 C--MD 10 tO 00051)1 000

22222222 2 5555 52 22 522255522

.0 5 55

CD 0 = 0 000 000 CD 00000 CD 00 = 0 0 0000 00 00 000 = 00000 55 0' 0005000000 010100 CDt_I C_IC_I 000 000000 00 00 00000000 0 0000 00000000 = CC) 00 0 OCSOC 000 0 00 00 IC) C_I C> •••••I — am 50 CC) (15 CD 1%J 40 1.0 LCD SC) '0 4_I ('.1 0 '0' C_a am '.O C- )%J 0' C_I 0 C_I — 0 — 00000 1411 00-CO 0' 0 0)_I C> 000000000 00000 0CD C 0000 00 00 00 000 = CD 0 CD CD CO CC 0 00 CD 0 000000 0 CC> S C_IC)> 0) 04 0 '.0 0501 CI ,_) CD C_IC_I 0500) 00 CC_OS 0 Cm LCD 0 10 fl. Ci CD (CC 05 — - Li

T - 4) CD CD (4

C_ICC, CD I 50 '-I- -

C_I -

g, 12 9 ' - C— 1014

CD0 oC'' 0 00 0n1 o50 0' ,'0 .0 0J5,b000 0

0 1.0 05 0005 0505 05 - toto 0 0 '0 (4 .410 0 '0 0' 0 * CO 10 5010 0— 10 0' d' 100

C! - 0 .-. .40 .40 9 . 40 40 0 40 40 en 0 40 40 40 40 NI N! . - 8 40 40 40 CM CM 40 40 -

U) 88

- 0) 50 0 04 40 - SO CM 40 50 50 £44 - CM C-. 40 00 - 4 CM r- en 40 40

S4 0 en '.0 ! 0 tO 0) 50 04 4-) 40 0 tO 40 40 50 40 0 40400 40 - N- 4(4 0)40 en 4-4 CM en en en CM 40 - ' tfl 50 - 40040 00 0.'C! C.) en 40

0 000000 00 0 0 40 0 0 0000000 0000 0 0 40 40 0 0 00 0 0 000 0 0 0 0000 0 0 0 40 CC 0 - CM C-IC'!,.-) q- 4 £44 C'! .1CM 4-MC0N! C! 40 00000CM CI <44 _( CM CM CM

CM 4040..! C.) 40 40 0 It) SO 40 40 SO 0 CM 40 404-40 C- 40 0 en a, ,-4 C-.

o 0 040 40 £41 C.) CM 4-4 4-4 50 0 0 0 40 SO C! '0' SC) 40 -4 .-t 00CM tM 0 C tO 10 50 0 - 40 40 4-, 4-4 rI CM - -I 004--a 40 50 in 40 0040..' In

4-) 40 .4 40 0400) 40 0% 0. 40 50 4-- 0 40 40 0) N- in 04 40 o t 00CM CM 0 ('4 (0 '0 tO C CM 40 0' 'a.' .0' 50 04-4 CM 40 . 0'0 0' 00000

en 4000000 00 0(04-) 0 IS) 40 .1 N- 50 0('4(44 40 '0'0 0% .'4

~ 40 ft 0 '40 '00 '0 '00 '00 '00 '00 '0 '400 .40 - '0000 40 40 40 40 40 40 40 40 400 40 40 40 40 40 40 40 40 40 40 0 40 40

40 5fl'0'04000 0 .0' 'C' 4-'! 050 50 40000 004(40

50 0 CMOS 50 0 ('40 0 00 04 4 400050.-lOin 4-- 000450 .0 0) .-4 I!) ,4' 040 C--0 OEM tO 50 4(4 004.0' 40 0 0 0404000

C ('4 000504040 40 CM 5(4 'a. 40 4-- *-4 '0.400 0 00 50 0 .-4 4004 50 00)0 (MC.) C! - N- N- .-4 40 I 40404040000 '0').) t 040 0 40 00404000 40 40 ('4- CM CM 0) 50 50 0) - CO 40 40 50 0 00 C--

40 0C.IO)40 04 500 404-44-'! - 0. = 4000)flSO C.! 400000 40 C-IN-SO 0 Cat.) 5050 04(4 rI C.) 04 00 en0 50 04 40040400 CM 00000 4-! ,-4 .-'-a I..4CM CM Ca 40 40 00000 00000

a' '0' .-4 50 40 40 40 '0'N- N-tOO Ca - CM00 04 0)00 0 '0 ,.-$ 50 tO O 0 40050.-tOO 054-4 0040 0. 04 - '0 0400040 4(4 '.0 4040CM 0 en 40 404040 50 50 C.4 -'0CM 00 40 0 .-! C-. 04 (0 0) 0) 4-- 0) 00 0. 50 50 8 888888 88 88.8 .8 8 8.888888 88888 40 93 8 0. 40 4000 ('0 404-) 4-4 4-4 4-4 40 40 4.) 5(40 40 '0'-'C4 40 'C' 00404040 0400 0 0 00000400 00000 0 000000 00 000 0 0 000 0 000 00 0 00 0 000000 00 0) 04 a) ('4 rI 004004000 000fl

0 000000 0 CC 00 0 40 0000000 000 40 0 0 000000 40 00 0 0 0000000 000 0 40 I..! 0. 5040 50 40 .-4 .5OC.) 0', CM rio 40 40 40(444004 C.!SC)0 4-! 0) 40 0) 4050 40 (44 40 5(4 0) 400 0 40 0.-! 4-IC.) 50 50 0. 0 CM N- N- C) 0 40 C.) 4005CC- 40'! 4040400 50 0. 0) 4040404040

0 000000 00 CM 0 0 0 00000 0 000400 0 000000 00 0400 C 0 00000 0 00.40 40 0

0 0000040 4)4 0 000 0 0 CC 0 40 40000 00000 C) 000000 00 0040 0 0 400 0 0000 00000 40 40 '0' 10040 40 4040 404040 0 0 50 04 50 'C' IL) 0 (44 0) 40 C-. 0 C.) 4-4 0)400 0 040.0' CM 40 4000) .1 ('0050 000400

.8 .8.8.8.8.8.8 .8.8 88.8 .8 .8 .88.8.8.888 888.8.8

40 94. 40 4040 cl 40 15) ' 5CI"2.-(

• 40 I Sn - I C CM 40 4Lfl• 0 N- 0 40 . C! mm _ 0 '8 40 8 8

' . 8 8,. ;mL OO €. - 4747 U H 05000 _ - - - - • C! - ('4 14 - - - - - . . - - ('0 4-! _d'0 -C' - - - - It

------vo 8 8 8 8

o .rc, as en en en en 0% 0 eq eq 1010 eq eq 0 tO .c in in to 0%

La -. e- eq - 00 0-i in - i--i w in C-a in fl in c 0 •--i Co 0% 0 tO en eq e'. en 0 0 0 en In en eq en eq 0 0% 0 - - C-I i I i

ad 00 00 en c--i en t- - en en Co -1 tO 00 r. en r C-i tO tfl e en Co to Co Co - In ifl 0 U-i - c-% tO 10• - - o - e c-a en - C-c en

- 10 toO, C-- en Co in O in - as 'a in t- to a, C en r- tn en en en en eq i-i c-a en c-a c-a eq etc en C. - .- t eq eq c-i eq 14 en eq c-si 0

0 1.4 '0 10 C-, c-- in 00 00 - C-i c-I 0-i eq r- it, 00. c-a r eq in in C-a in in eq en -a C-a C'4 in en en eq en -. -a -e eq In it-i it-i in - in in

00 e 0% 00 000 0000 000 0 0 0 000 0 0 0 e 00 CC 0 0 0 0 0 0 0 0 0 0 0 0 0 00 0 0 -.0 .-4 .-f i-i en c-i c-a en CO Co - eq en a, itt Co eq 00 50 OS 00 0 en C-a C-I c- eq eq - en eq eq r4 en en eq en eq 0 0 eq a-, en

e .-f to in en en in -a to in - i-i .-I InC Co Co in to eq en en - eq e, to en C-f C so eq -a e, to eq eq eq - so es, it, t in -0-

- 0 Co to in in to Co C-- ifl Co In c-i Co 10 0 a, .-I -a- c-I en c-f PM

o e 1010 000 SO 000 to in In C-- r4 c-f en - 10 0% en In

eq 0% O c-a .-t to U eq eq 0% t0 en r, I- C-) -a c-i Co Co en In 10

to Co sti Co 0 C-- -4 Co Co it, eq c- en in Co

Co Co 0. to an -a a, to o to a-, C- C C-C o C-- iC to c-a eq e-a C-I 000 -

8 8 8 888 8

1 - Co Co en to e In en Ioeq '0 50 0% eq Co C-a 50 eq c--a C V V 0 00 V C V 0% 0%

eq eq 0 00 0% en e eq eq in c--i c-I en en C en 1040 Co 0 a a - en a' to Co Os C- Os eq Co c-i C Co Co C- Co SO -

50 eq tO en C- c-I U In it, - 100010 c-f .-f en Co CO a--I -

en Co eq o i Co CO C' c-a -a en C It, Co C - e'a 00 in 50 to c-a en Cs. in C- 50 C-- 0 CC eti Co en In Co 1 50 0 10 0% C

I8 8 8 - 888 eq '0 r en en -a- 411 C-a 0 .-4 i-I eq en so it, eq .-I q-, en c--a c-f

288 888 828 8828 888 8 8 8 828 8 '00% 0% en eq tO 500. C- en CD C- en q-I In ifl in e'a w c-i Co en o en a, .0 etc c-I en en .-4 .-i eq 05 50 10 eq eq no fl Sn &fl 0 in a, c-I in in a-, en en Co en a-' en -I Co c-i Co en in Co Co a- 0 00 000 = 0 000 = 000 0 0% ci 0 = 0 0 000 000 0 0000 0 0 0 0 0 000 0 Co Co 0 en Co to a ni 0s en In c-I c-i Co C-- 0% c--* Ci .-4 en eq c:: 8g 8

888 888 888 8888 888 8 8 8 888 8 en tO 0 en in a, in a- .-i eq so en en a, 50 it, 0% 5000 Co '0 0 Co 00 Co Co c-I to .-a en 0% a-- Co tO it) Co 10 In en Co 0% 0 eq c-I in en en en eq Co 5010

888 888 288 2888 888 8 8 8 888 8 888 5

888 888 888 8888 888 8 8 8 888 8 en 4-1 Co Co eq 00 .0 .-i in to c-a 000 1- to to CO C 50 In c-I 0 - en in to - eq en Co Co -I Co Co Co i-I in en en i-a - -

o o 0 000 0 00 000 C 000 0 0 0 0% 0 0 0 000 00 000 00% 000 0 0 0 00 0 0 1- it, CO CO U en C 05 00 i-I In en C- 0 100, C-n.-. 0 50 Co La. 0, i-I eq

o Co c-a Co en el-f .01 it) Co Co Co Co Co in ta-s it, in 0 0 U I C 0 C 0 0 C-- eq c- 0 I It, eq em it, en Co

Co 0

$4 C3 4.4

1411

Zn eq c--f eq eq c-il en en C-. -e -e Co In in UI c--i e-1 eq 0 en In -

0 4

0 ent-a, ene-,0cccb a'0000 0000

0] en - en en en 0 0' enene-. .fl, a- a' tn en en 0a'enen en-..' C-I 0

a) 0]

s0 0 en o- cO C- Ca C-- SO 00 0, en 0 00 IL, IL) 'a tO 0 tO en en en C en s-I C-, en C-a C-len carIe. en a. en .-. en en e.a C.. 0 c. en e%4 en en

0000 00000 = 00 000 00000 = 000 00 -

en 00 en- 00 C. C. - - en 11% .a en en cc c-I en - so C-, 0000 00, en en en en a) C-fl Ca en 0 .- tO .-( en C-a .-. en 00 en C] 00 50 00

-. C- 1000 en Lc'l to en so cc a, r.s en to in .ec a. 'a a, en en en c-a CM Cc, C%I c en - en ccl .-c CC .-i 0 - en enenenen -. - en C. en

o CII ci en 00 0 en a, a, 0 en Cn 0 c--I C- '0 eS 5040 ' CO SO SO O CC

000 0 ilfi In 10 cII

10 In 0 to tO — C-cl -e .C) In C-I in -a- 0 en 0 s-I 00 ' Ifl en c. ill In en ho tO 'cl -0' en a. en a. t'I C-a a, C-, a c-i -a- en a men toe-.

Ccl CM .o. Cd 0 ILl en c-i C-a en 00 00 0010 00 III to 0 C--, so t- to 00 en 0000 en 00 '1 %' 0 C-cl en 0 0 0 s-i 000 00 00 = 00

- '-

CII IC) C'I 'a a. a. LCfl to a en 0 5000 sOa. 'a a, e- 000 0 0 'l 00 0 'V 00 0 00 00 0 'V = 0 0 00 00 l 0

- a. 2 0. - s-c en en en en en

00 0 00 O so en so a. .-i to r- C-a en en a, C- tO 50 'a 'a 5050 In a' SO 'a 'a 5000

00 Ct ccl cc 04 en s-I C-I i-I .-I i-I C-C en c, C-a - en c-i C--a en en Cc, eg .-4 C., C-

-S. a' c-a a, n a. en 0 00 en Laenen enenennen a' C-cl en en a. so a. .L-Iencnenen a,en.-. -C-a' 50 enn ena'Cn enen

ggg g gg gggg gg g

0) .o en O''a ICfl a- a. 'a tCSen-0'-.eentt, 00 'aentd enna'eneclen C-I en en en IC, 000 = 000000 0 000 000 00 00 000 00 0000 000000 00 000 000000 000 00 00 01 .-I 0 C 0 tfl 0 CO en 'a to r-. ILfi so -a c-a 00 CC 0, L0 a--. C- 'a a' en en en 'a en en a' Ul tO a) - 'a a- en ILl 0 s-I a' a' 00 C- a. I0 en

0000 000000 0 0000 000000 000 00 0 00 0 000 0 0 000 00 e 000 00 0 00

0 en C-I SC, en C en to to.nn en -C' 'a Ill 'a C-, c-I 40 'a

gg gg g 0300 'a 'a 03 en I- CO 10 a. 50 tO e 00 tO 500 Cl 0 IC) en en fl en en 0 C-- en 00 en 000050 so C-- en C-- so c-- en C- CO t Co en 00 00 50 CO 0%

gg .g:gg g

C-a .- en c-a c- -C- en ,-, en en en nC-len 00000 CC- 'a Cl

0000 000000 0 000 000000 0003 00 00000 00 = 00 0000 000000 0 000 0 00 0 00 0 0 0 en tO en a. 0 50 03 en en en r--. cc 0 in a. C C-- o In C-I tO .0 50 en 0 c-I en 0 CI 0 en tan 0 -.soenenr- a, en en 0. en en en - en c-i - en en en 010 en a-a t> CC LO Ca -

00 enL1 & ' .oTc!,en Cc, 0 I I en 03 • 01 00 0 - 0 0 5 C 01 C--I CC - to

I- en iI .ii

'a 10 • 00 'a C) en Ce) Ce) 0- ..-I -.-4 0- -e-1 - I200CL

ft 00 'r --- •. ------

_) en e-ce 'f 41

- - 'C-_cl' .Cenen c-Ienci In ILl 000000 - - - cz_ !z . ez z_ 03 -a- a' -C- a. -- a' -a a' .-, C- en en '-I en en en en en o .-4

C 00 cc 4-JO

- •0 '0.0 'C '0 .0 .0 00 C C C PS

.01 C, 01 W.Q

C 0101

0Q -I-I U - cc u-fl on a, -t cc en F- C en lot- I- p-3 04 Cl] CM — 04 Cl] en Cl] Cl] I Cl] 0

-C-I

00 000 00 0 00 en p-I C = 0 00 0 00 0 00 in 0 0 04 E-] On I-I p-I 0 CM en p-I p-I 04

04 0e00 00 t In a, a, in a cc cc -c- a' u-fl In Ln 00 .0.0', 0

I] 0 — Cl] 0.1CM en CM Cl 04 p-I "4 0

c-I In 0 WV

"SW

01 -ri 0 00 1b -C - - - C-I cc InC. a- u-C to p-. cc Ifl Cl.

-,- N OlIn 0

In 01 Xk -4-4- 10 00 0. 00 C C C C C C C 0 0 c-i

•% 00

010 to en C--. C CO cc in 0 CII C- 0% to 0 -i-I

014J -'-I

-c-I

0In C—cc 00 C tOin CII p-I -I C] p-I p-I .I0 0 CM p-I 0 PiS JW CII It) 0 p-I Cl] 04 .1 ('4 C%I 10 In In 0

COO o - 00 00 = 00 0 00 cc 00 000 0 0 00 0 ,-4 0 01.1 0 Cl] C, t 01 10 0 0 It) cc In 0% cc 0 cc to cc a a, In c-i Op-I 'I-lW cc cc I-IC-] CII Cfl Cfl en Cl] .10 00 000 00 0 00 dna 0% p-I 0 00 000 00 0 00 -1 Cli o • I" 0.0 000 00 0 00 Cl C '-I 00 000 00 0 00 Ifl cc 0 in cc cc 00 .-I 0000 10CM CC a' to en cc CO CO to In on = O p-I It) It)

00 0 00 000 00 0 00 000 a, en en 00 00 0 00 - on to 10 in CM C", C 0' tO 0 COCO 0 cc en In 00 In 0 tO C- 00 01

3-i 'SV Ifl 0 I', 0 p-I O on In p-i tO C, p-I-fl - p-i enCM]p-Ifl1 In to CII — 5,4 f-I 3-lU) 00 000 00 0 00 00 0 as 00 000 00 0 00 .-I — 0 O4- 0 cc . ri ('S 4 C 4) 4.} I :i 0 a C] c-leo 'is C C- C) 1c1 C C C 0', Co cc C-I C • w.c 00 0) 0) 30 (nPJj4J 0 U) C C I

00 01-4 lI-I 11 - *II 5,.IFI In 10

I pp 'o '0 0 a, no cc a, on cc. cc 0 01 Cl] CO a, in to to cc -- 'C tO cc E-' 0 4) U

C LU LU LU LU C' LU LU LU . . - o C - C C C LU LU U) C C C CO C) LU UI C C C C IC) (N ) C C 0

') 1-1 (N C LU CO r LU C - ' LU C> -. CO C LI LU (N s' C C 0

Cfl -lC COLCU- C C) U-U-LU L

> U LU LU LU IC) LU LU LU LU LU IC) • •'CC -C

C C) U- 0) LU LU C (N U- 4') (N 4 LU (N (N LU (N 1-4 4 .4 ('I ,-4 .1 LI 4') (N

LU - 0 —4- - LU - . 0 . • . • . C) -' o , " v e V 0 i '.- ' Ci 0 . U)

Ic,

Cl ) LU 0' mr 10 t- C IC) C (7 LU CO C LU LU U— LU LU 0' CO — .-1 %, (N sj - C 1.1 'i '-' 'V 'V

(N CO 0' 0 .-4 C LU LU C CO 4) C LU C U-U) -4 U-- ' U) '- 4') U) 'V C - -I U4 OD Cli c. .1 1 a,

C LU 0) LU 10 (N (N C () 4') .-4 I 4 (N 4') LU 4') co -1 C 1-V- C U) 4') C LUU)C LU LU CO LU - U)

4') 4 C LU LU 4 .-4 (N (N LU CO U) CO QQ U— LU CO LU 14 U) LU C' LU U) LU LU C U- U) LU 4 4', 4') CCU)-4 -4'U• U] (N (N .- (N -' -' 4 .1 C *1 -I U] U) U) •• -

C C C C C C C C C C U— C C C C C C C C C (N 0) C' - CO M 0) U) • 0'. LU LU U) —4 C' LU .-i C -4 C C (N 4 C (N .-4 .4 C) U- LU -4 4') Ui .I 4 .-4 .-1 .-I -4 -4 -t -4 C 14 .1 —4 (N (N (N (N (N —

C C C C C C C C C' C C C C C C' C C C' C' C C C C C C C C C C CO LU C C C' C C' C' C C UI ('I U) C C)) U— -4 C U) 0) Cl) LU C 4'- (N U- U— LU '4 LU 0) (N LU LU CO 10 C LU LU C U- LU LU CO U- LI U) (N U) (N (N (N (N (N (N C (N (N -p CC Lb U) (N (N U— IC) (7) CO C C C C C C LU 4') LU (N C (N CO .1 C C U— U— C C C C C C C C C LU IC) IC) U- C LU LU C C C C C

C C C C C C C C C C 0) LU -1 C C' C C C C C 4- U- (N LU CO (N C U) CO (N LU 0) (N (N C U) . LU CO (N 40 C' LU mr U- 4') U) Cl) wrIS- C IC) (N C' C' C- U) Co C' U— U U) 44 C"

0' U— 0)U)C' LDCN U- Cl) co -44 C 0) 10 C IC) (N U) LU LU C 0) ON LU '-. C LU CO CU CC' U) C) C LU 0 C'J C LU U) C') U- C CC) 0) 10 U) 0' Ui Cl) LU 0 0 U) C -' 1-4 LU 4 - ,-4 U) 44 -I -1 to -4

LU crC LU LU U) (N C' LU CO DIC LU 'C CU LU U- CO LU LU LU U- LU LU to x C 0) 0) 0) C C C td o W

Ej C - CO LU - (N - 0) 0 '0 . (N Ui CO LU CO C - C LU C (N -.- LU U- O U] U) U- (N1 Cl) 4') 10 4') U) 4') -1 C') 10 0 ' C COO) -4 4 ,' " ' 'i ') (N (N r- 8 WA LU C 4- 4') 0) CC CO t- (N CU LU U) 0) (3) 4'- CO ...... •' >1 Cu 4- U) 4- 4- U— C- 4-U-- t- CO U- U— 0 U— C', CO U— 4--CO U--

Os (N 0) LU C' (N C- (N (N Co ON Os CO

CU )C'06O) CO,-lCO C4) .-4 CCC CC 44 -I-' LI (N (N (N (N (N (N (N U) ('3 (N (N (N (N -4-' C) (N C C C C C o

C C 0 CO C 04 C) - C' CO - - 0 C' C C) C' . - C C -I 4 0) C' E- 0)E- CI) 0) - 0) 0) - 04 CC CO ' CO C 0' En -. C (. DL4-4 .-1 -4 -.EX4 £4 0 ---4 •).0 _•.Q 0 .4 C) C C C' • . . . U U -4 .4 (N (N ('3 • . - —4 --4 —4 - (N C (N 4)) . (N ...... C) . C a) . -I (NC'iCOtC-4U- • .4-ICC 40 .-4 COCOC)))(Y) 0)0) (N U)

40 .4 -4 -4- 41) C 0) 40 UI CII -4 40 -c (N U) -4 (N U] (4)

• C) C) cc cc .-i cc cc .-i cc cc - - cc cc cc t=i o cc w

cn a' cc Lf cc cc cc b cc .- a' cc.. cc cc cc cc cc cc cc cc • ., ...... cc cc cc cc cc cc cc cc cc cc C) cc C) cc cc cc cc Ln LO LC C)cc cc cc cc cc cc cc cc cc C) cc cc cc r1 cc - • • . . -cc • . 11 cc cc cc cc cc cc cc cc cc cc cc cc cc cc cc C cc cc cc cc cc cc cc cc cc 0 - cc 0 cc - cc 0 - C'J cc cc C) C) cc cc cc cc cc cc cc CD • . . . . . cc cc cc cc cc cc cc cc cc cc cc cc cc cc 0 cc' cc cc cc cc cc CD cc cc Rp cc cc cc cc cc cc ) cc cc . C'J cc • . .•-i ..- 4) cc cc cc cc cc cc 0 C) cc cc cc cc cc cc cc cc cc cc cc-I CD -. cc cc cc cc cc cc cc cc - cc cc cc cc cc cc cc cc cc. cc cc cc cc cc cc cc cc - cc cc EE a) V

o LC) cc cc cc cc cc cc cc Ln cc cc Lfl in cc cc cc Ul 'V 'V

- .-4 .-t ------V (U cc cc cc cc cc cc cc cc cc cc cc cc

.-.I cc cc cc - .-i C'.1 .1 0 '1 C' ' 'V 0 (U

cc a' cc cc cc cc cc r cc cc cc cc cc cc cc U) - 1 -4 .-I .-I %d 'V 'V 'V 'V .1 ', •-1 cc cc cc cc cc cc cc i-. cc cc cc cc cc cc cc - cc cc cc i- cc cc Cl) cc cc - i '.' '.' .-4 'V 'V .-4 .-4 U)

cc ON cc cc cc cc cc cc cc I- cc cc cc cc cc cc cc i-.- C -r cc cc I'.- cc .- - cc cc cc '-1 cc r-4 a' cc cc .-4 cc cc cc cc cc U)

icc cc cc r-.- cc in cc cc cc cc cc cc cc cc 'cc cc r- cc cc cc cc r- . cc cc cc cc cc cc cc cc cc cc cc cc qw cc 'a (U cc cc cc q& cc cc cc ' ,_ cc cc cc cc cc cc cc . cc cc - cc cc cc cc cc cc cc cc cc cci -4 cc cc U) •r1 cc cc cc cc cc a' to cc a cc cc cc P cc F- I".- a' cc cc- p- a' cccc cc cc Co .-. cccc U) P1 cc cca' cc cc cc cc cc cc ' cc cc cc cc cc cc cc cc cc cc cc cc cc cc cc cc cc cc cc cc cc cc cc cc a' a' cc cc cc cc '- cccc cc cc cc cc cc a' a' cc 4) U) cc cc cc cc cc a' cc cc cc cc cc cc cc a' c-i CD - cc cccc cc cc en Cl) wrcc- Cl) cc cc cc "J cc

41) cc cc cc cc cc cc cc a' cc cc ' cc cc- ' _ Co cccc 4) OD cc cc cc cc CD cc OD cc cc cc cc -cc cc cc a' 'V cc cc cc cc cc cc ci cc cc cc C cc (U -I cc cc cc cc cc cc cc cc cc cc cc cc cc cc cc cc cc CD cc CD cc cc cc cc cc a' cc cc cc cc cccc cc cc cc a' cc a' cc cc cc cc cc.. cc r-4 04 qj

- v) cc cc cc cc r oD a' cc cc a' 041) cC .- .-i cc cc cc r4 4-1

cc cc cc cc cc- cci cc cc .- cc cccc cc cc-. _ cc cccc cc cc cc cc a' cc o ,- 44

cc --. cc Cl) in cc d. cc cc cc cc - cc .- cc Co cc cc- o . .. cc .-I ç'J cc cc ' cc cc cc cc cc cci cc c-i

C cc cc cc cc cc cc cc cc .' cc cc CD cc cc cc cc ...... C. C). '- cc cc cc in cc cc cc cc cc cc cc cc cc cc cc cc cc cc. cc cc cc . cc cc cc cc c-i - cc cc cc cc - cc cc cc ccC -Q occcc cc(-a0 cc cc cc cc cc cc cc cc cc cc cc cc cc cc cc cc 4-1 Iii o

U]. Ng Wa) 4) -I C) [ U]4J C.) - - >inj

.0 .cc cc cc cc cc 0CC cc 4-'- C) 55 OC Ccc 4)4) ii C)' C)' cc' 0 I/i (0 0 0) .CQ) CccC4)0 00a'Q 0 cc cc cc (0 cc 0 . a' 01 w CD cc cc .-O • 0) Q) Co Ifl

cc - Wa) cc cc CD • • "C' 'C 'C' 'C C r-I 14 cc. cc cc in cc in cccc- cccc ON - cc cc cc cc -1 -.- -. _- - - .- -.-- cc - -'-- - — cc cc cc cc cc cc 'C c-a cc 'C cc cc ,-' cc cc cc cc cc cc cc cc cc cc cc cc cc cc cc cc cc cc cc cc cc cc

Co en '-C: 'C: en 'C: C: 'C) 'a,) Li') C: C: 'I-, CoCo to cci Co • . .1 • ' Co to c-a o -'C)'C: . 'C: 'a,] . 'C: 'C) - . . •C) C) C: C: C) 'C) C) (C) o en c-i to c' c-i C: C: C,) 'C) -'C)'C)'W'C:'C: I • , 'C: C) 'C) C) 'C) 'C) • . 'C: 'C) 'C: 'C) CO C) C: to Li) CO en '-I to C: (C) C C: • C 'C: •'C:'C)'C:'C:'C: 'C: 'C) C) ' C) 'C) . , 'C) 'C) 'a,, C Ci) C: C: C:

to to to to IC) to C: • 'C: •'C) •'C:'C: 'C) 'C) . ' 'C) • . 'C: 'C: 'C) C: C en '-I CI P1 P1 C-i en (C) to C-'- 8 to 'C: 'C) CO Co CO tO (0 '0 '0 Co 'C: 'C) 'C) 'C) tO U') 'C: '0'0 V. CO Cli C:C:

CO C-i to C-'- (C) c-i c-i 03 C-i V. en (0 C-i V. Cl) LO I-i

'V Co IC) • - I C: P1 c-i en en en en C: en en .-i 1 - - - • • ' 'C: . . • C.) 'V 0 C: C C: 'V 'V v 0

C:! to (C) (C) (C') ' 0" en CO en Co 0) V. V. (C) V. CO to to C- C'- Li) en - t 'V en en en - c-a C-i C',) (n 'V C-i C-i V 'V

P1 CO (C) co en uI CO I (0 , (n IC) 0) U') 0) U') C' C-i Co C: (C) (C) Co C-i c)'V C',) C-- Co e- V. V' 'V V. c-i en en 0) C'i to c-i

C"J (C) ' r-, C: en C-- (C) C: to C U) co C: C V. C- (C) C C-i to C: en ' Cl) c-i ' ' w enen c-i P1 co '- P1 Co en to r. I-)

C: C:0 -4 'C) (C) 'C 'C) U'I Ii) (C) 'C) 'C: IC) '0 '0 'C: '0 (C) V.'0 'C: 'C: (C) C: 'V

C'- C: C CO (C) C'- P1 1 Co C') C-I en Co en LI) to C-i ,en P1 5.4 C C: ON Co '- - - C--i Co tO IC) co Co C - Co C-- to C-- Co Cl) CO c-i en - en C-i en en en en C C: C: C: C: C-- C: C Ci) C: en Cl) C: C: C: C: C: C: C: LI) Co en en c-i c-i c-i e I CO V. to C-- Co V. C-i to V. to C'] en to Co en C--i en - Qt Co Co 0) Co -i en 0) e- C: Co I-) C-i C',] Co CN = en en en en - en en en C: C C: C: C: C: C C: C: C: C-iC C C: C: C: C: C C: C-- C C C: C: C C: C C C: C C: C: to en C-- (0 zr C-i I () C: to Cl ) (C) to C-,] V. C- (C) C c-i C: C-) co (C) C- C- (0 I Co C: en C-a C Ii) P1 en C: C CC) CO 110 tO V. Li en Li) ,-1 to C-i C-i en C') C) C-] '-) Ci C'-,] C') en

(C) en en C: Co ' C) 'C: C: C-i 'C: V. en 'C: '0 C-- 'C: '0 'C: 'C: (C] CO P1 CO CO C: C: C') C: C: C) C: C-) Ct] C: C: C: C-i

-4 C:C: C: C C: C: C: I C C: C: C: C C: C: C: C: C: C: C C: C C) 'i' ' Co c-i (C) to ' (0 Co V. Cl) C)) 0) Ci IC) en to V. C') C" V. C--i en Co c-i en en en Co I V. V. V. V. to V. C- 0) tO en ) U) en (0 V.

'-I en Co V. V' V. I - - to - Co to Co Co CC) C-i (C) to C-) 'C) en en en ' 'o C: -4

p.'-. V. V. CC) 0) 0) -a Co c-i en c-a V. C") Co cci to V. ,-I rrtOr-r-- I Ci o ae C:

to (C) '-I en en l-a to tO C en V. to C-i C- Co en (C) LI) Co (C) (C) (C) to to C: Co to (0 tO C-) (0 Li) to C Co LI) P1 V. CO en c-i tO c-i en en en ,-, ,-t ,-c

Co V. CC) to C-- to V. (0 V. c-i tO 03 C: en U) V. (0 Co cn Co CO .1 Cl. tO tO----- I (0 (0 (0 (0 LCI '0 to e- to to - - - to c-i (0 Co C-) U) CO C C: Co Co C) to C'i Li) V. C',] C-) DGCoC'- Ct] 33 0) '0 C-- 03 '0 '0 '0 CoCotOtOCoCi C-a C-i C-i C-i C,) C-i C',] C-i Ct] C: C: C: C',] Ct] c-i I-I ci) en 0) en en C: (3) ci)

COCci 0) I o 'ci C: -1 - 818,014 .0 - -en ,,-. C: C: C) cci CO - .-4 0) 0) C: i)- CO Cd '-4 Cd '0 .] Cd Cd en en --S -4 en Li enct- I en en C). D. Cd C. C). (C 3, 3, 3, Cd Cd -en CM4 -4 - I C). C) 'C: '0 -. - 'C) '0 0 0 0. 'C: '0 0 0 0 0) 0) 0) 0) C)) ls 0'en I 0 0 0 0 0 'C: 0 C: C:.0 0 C. 0 0) ci) 0) C: 0) : a,, --I (I o c' en en .0 0 C) 0 0 0) en -I.0 .C.4 0) en en04,04-Q0Qo, 0. 4-. I4-.Q .en0)0) .0 '- en' 3-' C: ". en cci en en cn C: . C: CO 4-. CO en - -i - . en .1 o - '--C:-- '--0 Cls C: (0 C:.0 C: ' C: C: C-,] en.0 en 0) en 0) en w • w en .-. en en ,-4 0) ,- 0) C') . en en c-i en a) -V. •V.V. - .0) .0) . 0) , 0)'0' 0) - . C). • .0) '0' . U_] 'toLC) C- Co(0'0(0tOCd •tdCo 0i'C:CoC:C:(C C: 'CiV.CoC: en e' en en - I c-i'0 (-i 0. c-a C:. Ci 0. en 0. C-i C-i C. c-a ---1 P1 0, en en C-I c-a ,-S en - - - - — - - I 0) '-- C) - 0) - C) - 0 -- 0 - (0 - C) - - -'- - - C--) C--] en a,i c-a C)) 0) Ct] C) en en ,-. I-i en CI) :Sicen en V' I nic (0 (/) (0 'I-s I4- 'C: Cl)

'0 '0 at 10O '0'0 rS o o 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

Ln cc -tv a '0 CD Co 0 0 0 0 o 0

- - CD CD CD • 'O '0 - '0'0 '0 '0 '0 '0 '0 '0 '0 '0 '0 '0'0 '0'0 '0'0'0 '0 CD CD CD 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 CD 0 - = 0 00 • 'O'0 . '0'0 '0 '0'0 CD 0 CD 0' 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

0 0 0 0 0 • 'O'0 • '0 - '0 '0 '0 '0'0 '0 '0 '0 - '0 -'0 '0'0 '0'0 - - CD 0 00' CD 0 CD 0 0 0 0 0 0

'-4 •— 0 0 Co . '0'0 0'0 '0'0 '0'0'0'0 '0 '0 '0 '0 '0 '0'0 '0'0 '0'0'0 '0 cc 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

0 CD 0 0 CD CD 0 0 CD 0 0 CD Co 0' - 0 D r-. 0 0 CD 0 0 0 ' - - % -•-, \/ , , 'V '1 'V 'V

0 0 f) cc Co 0 IS) 0 Co IC) 0 0 Co Co 0) 0 0 0) 0 0 s-- 0 s-I 0

o (-)

cc 00) IS) 0 Co 0' s-I 0 0) 0 0 r- 0 00 Co IS) Co C SO 10 s-(

0 IS) 0 0 ICI IS) 0 IS) 0 U•) IS) IC) IS) IS) 0 0' IS) IS) IS) U) cc IS) IC) IS) C '., a) a) -v 'J 0 'V 'V 'V 'V C) cli 0

s-I CD CD s-I s-I CD IS) F-, ('J F-. CD .- a) - o SD C- CO cc- o cli CD cc cc s-I a) 0 0 s-I(N C) IS) o a) cc cc 1-1 ci cc CO r. s-

0' CD CD 0 1 IC)'0 '0'0 '0 '0 '0'0 'C '0 occ

cc - Co cc- C- Cl) CD r- D c- o s- a) CD C- cc- o cc cc o cc 0 0 I-i - cc CO r- U) cc o s--i CO 'cc' 0' i-- .- CO s- 0 I'- 0 CD IS) CD En Cl s-I 0 .I s-I s-I s-I IS)

a) CD 0 CD C) CD 0' 00 CD 0' CD C CD CD CD 0' IT0 0 CD 0) CD C) C C) cc c-,j C) LO cc ON so 0) cc so cccc CO o cc s- ' SD 0' cc a) C) cc SO 0 U) C- C) 0 CD C) C") s-I 0) 50 CD 0' 'cc' s- o s-I 0 0 0 ------.-'I - - - •-I C') - s-I ------s-I CD' 00' 0 CD CD 0 CD' CD 0' CD C 0' CD C) C C 0' CDCDCD CD 00 C C) C CD C' CD' CD 0' CD CD C' CD 0' CD CD CD CD 00 C C) 0' 0) 0 CD o c- a) 0 0' CD SO IS) IC) 0) C') Cl) C') 13) a) CD 50 15) Cs) cc o C') cc o o' cc' cS) CD Co cli Co C') CD, a) cc C) - cc cc' C- 0 a) CD IS) c's) o o Cl) s-I s-I C) c'S) C) C) Cs) CD 0 c's) Cs) a) cc Cs) -

- 0' cc• C') '0'0 C') 0 '0 '0 '0 10'0'0'0 '0 '0 '0 '0 'CD 0 '0'0'0 0 13 CD 'V - C CD C C C C C C C C C C 0 C C C

s-I CD 0' C 0' 000' CD CD C CD C) CD CD C) 0, CD 0) C) 00 C CD 0 0 -'-,, C) s-I CD '' U) CD' 15) C <') a) 5)) 'cc' CD 0) SD I') -cc- SC) o CD s-I C") CD C-- I 0'. C') C- 0' C) C') C- '' ' a) 50 'cc' U) C) SD C- cc cc s-i cc CD SD CD C') Cl) s-I ,'-I s-I 0 C') 0 s-I s-i s-i Cs) s-I C') C') 0 C')

-I 0) 15) CD s-i o a) 0 5)) 0) IS) 50 d. 0' 50 C') SD IC) 1- C) U) 130 C C

50 cc 5-, o 15) 5--) 0 Co d' cc r"- C') - cc cc- c'- cc- c-a -. 15) = c-I C - C s- C') 0) a) 5--

C CD IC) a) 0 CD ON X) c-a cc C') 0) ,- 0 CS) C- 0" 15') o U) cc '0 U') U')

0) 00) cc' Cs) 'cc' ' CD U) s-I s-I 0 s-I ' 0' 0 0) r- 0' s-s 5- CD IS) a) ,- ,--, 5-- co r's-a)a) 00s-C-

U) c-a a) - IS) 15)0 0 '0' IS) C'- 0) 5))

C') C Cs-I CD Cl) 5-) 1') C') C') CD C r') CO

'-1 C) 13

cc -cc '0 Cl) cc cc cc C 3-s 13 5) ,13 cc CO ,_' cc cc cc '-cc cc cc 0 s--I 0 CD 3, 113 0 dl) 'V -. cc dl) cc 3, 3, a) CU cc c - cc cd '0--I cc '0 cc dl) cc cc cc 3, 13 C)(L) .-i -s--i 3- - •,-1 -s-I ---I -s-I - ,.-I -I31 -I -s-I I 0 It Di C >3-3-3-3- .3-3-3-3- 0-00- '0 530. 0(1300. E-IE-IL) C) (C(CCD'0cc'03-'0It'0'0 -I'O -s - c13 0 C CD CD CD CD CD CD rC C) CD CD 13) 0 '*- 0 0 00 13 U) 0. 5). 0) 0) 53 3, 13 (I) CD 0 0 0 '-4 C 113 0) - ,)4 ,lS (13 C) - - .0-S (3 • • • • • '-I ,-I C' 0 - '0 C) ,D) - cc) cc cci '00 'i- 513 sO. 0 ,-3 cO I,) s-) .,) .-1 E. 00 cs) 1-3 cc -3 - - 5-3 • • . . 0) CD - CD CD 0' CD C CI CD CD 00' C C 3, CD 3, C '0 0' 0 CD CD (13 CD - s-I C) - s-I ,-* .- s--a s-i s-i ,-I s-s cc C7, () '0. ' . • (3. • C C . I '0' • - - • - - U) 0 = . cc '0 '0' o Cl C',) SD CS) cc 10 s-I (0 s-I - 0' s-I s-I 000 0' '0 C') 53 C') 0. C) (3. CD 000 000 s.D CD 's--Ioos-')CD C")r')'0"- 55) -o o '.- cc 0 -

o •o (_•4 (N 0 0

LO ,-I 0 00 0 . . Co o o o -o • o -o 0 0 0 0 0 00 0 0 0 0 0 0 0 0 0 0 0 0 0 (N 0 00 0 - .0 0 .0 Co 0 0 0 0 0 0 0 00 0 0 0 0 0 0 0 0 0 0

Ui 0 Ui (N •0 0 .0 0 • .0 C.0 • 0 -0 0 0 0 0 0 0 o 0 0 0 0

0 — Ui 0 Ui O -

Cl) Co Ui Ui Co Ui Ui Ui (N C) .- (N (N ' '- - '. - Ui 00 Ui Ui c-- Ui Ui 0 N-

.-1 Ui (N CO CO 0 0' 0' 0) N— Co Co Co Co (N 0' - - •C (N Ui

00 • . . . 0 • . . . 0

C N— ON 0' 0 m 0) Ui 00 Ci 0 .-I Co N— Ui (N Ui (Uv Ui CO Ui 0 0 Co c-i (N -i c-) c-- (N .-1 r- 'V (N

) Ln c--s ... 0 kn Ui Ui I S.- \ — — •• Ui Ui Ui Ui Ui Ui 0 Ui — 'V 'V 'V N— S.' 'V s' 'V 'V 'V — 'V

0 0 Cl) 0 N- 0 Ui 0 0 0 Ui 0 0) If) -I N- Ui 0 .— ON Ci c-) N— Ui (U 0 -I (N 0 ON N— Ci N- (N C 0) N— 0 If) 0 i .1 0 c-i N- Oi (N (N (N

CO 00 0 - •0 q=o U) 0 -0 C C C C C c) C C

Qi Co N- Ci 0 (U N— 0 0) 0 0 0 Ui If) 0 (N 0 0 0 Ci 0 Ui (U' c-i Ci (U' If) Co Co Ui 0 r'- 0— (U' 0 (I) (N .1 . rI 4

0 c-i If) 0 Ui CO 0 Ui Ui 0) 00 N) 000 Co Ui 0 ,'-1 Co (N (U Ui C) - .-I Co 0 N) (U' CC 0 Ui 0 Co 0 Ci Ui (N N— 0) 0 (N N— N— 0 0

0 0 0 00 0 00 000 0 00000000 0 0 0 0) 0 ON 00 0 Ui 0 0 0 ON (n 00000 (N Ni 00 N- 0) N— N— 0 If) (N (U c-i U) 0 c--I c-i 0 c-- (U U) U) N— Ci Csj U) U) (N 0 N— L() 0) 0 C) C') (N (N N- Co Ni If) N) (N (N C') C') (N c-I .-1 c-I (N

(U' CO U) U) 13) 0 '0 '0 0 '0.0 00 '0.0.0 C C C C C C C 0 (N C C C C C C N) C C C 0000 000 -1 0 0 0 00 0 00 0 (U' 0 Cl 0 0 0) (N 0 (N 0 (U' '- 0 13) Co CO 0 Ni 0 C') (N (U' (U' 0 10 IL) 0) c-i 00 (U' N— U) 0' If) U) 'C' c-4 (N Cl Cl c--I c-I N— (U' c- c-.I c-I c--I c-I (U —

1 Ci N) (N In 0' Cl (N 'C' 0 ft .—. (N IC E Co C C

Cl N) 0 N) C') (Uv ON 0 r- Cl) Cl CO CO (3) Co N) ,-I (N c--I N- 00)0)0) (U'

130 N) ,-I 0) 0 on (U' C') U) o /) N— N— (U' (U' (N 0) (N c-I 'C' Ni o — — — — — — — N— Co (U' (U' U) U) — — C

(N N— 0 0 If) (U' N— 0) 0 0 (U' 0) 0) Co . . - . . . . . , - , . N--

N-Co(N CO N),-I0U)ClSD

C) 'C) C 0 00) 0) ON Co Co -1 Cl U) ,-I —I Ni .-1 0) C') N) N)

(1) 0' (U --I '-I ci) -I C) =

C C 0 (0(0 - 0 - (U C) () .O .0 c--I 1/) '-I C. 4-) I C C ci) +- -I-' -I- 4- - Co Co C) ,-4 00 C)C) (0 (flWci)ci)ci)C1) CU I-. CD 0 ci) 0 0' 0' I,) U) .0 .0 U) C)) C)) If) If) -- - — a) 0 0 - 0 (0 00 (Ii U) C.) C.) II-) C) (ci (/) 0' 0' CT) 0 - - 0 0 0) 0 0 0 0 0 Cci C)) (C) (U -1 0' 0) 0) +- -.- 0 0) 0) 0) 0) 0) Cci U) (C) It) ci) ci) C 0 0 ci) ci) C]) (I) CC) 1 U) (l) (l) Ci) ci) (1) ci) 0) CO o .-I 'C) (1) •..-) .,-4 I E 0 .i ,') U) ci) CC I-) L . .IU ,)U .)' C), I- 0 (U 0•- • C) (-I 0)0) 0 0.'-) • - - . , , , CC) (N (N Cs] VS NJ c—I C') (N (N (N 'C' (N (N (N , , - , ,C)...... —1 50 Co 0) C, cC - 00 c—C U) U)'0 C') --Zr (U' (U' U) U) U) (N (N c--i c—C Cs] (N 0 c-a c-a (N (N - 0 --- - - — - -0 -0 — - - '- - - — Ni I')- C') .—I C') 'C —I (N U) Co 'C' Ui (N IC Co Co -C Co Co 'l- Co Co Co Co Co Co Co 'i-C Co Co Co Co Co Co Co Co

en 0' 0

0 C (N '0 CC-) 000 0)

U1 1t Ln Ui Ln W< U-) U< U) 'V 'ci I-I en aS IQ en -i Cl — - - ',0 • 'V aS (N a' k.0 Lo 00 a' ON a' (N -4 0'. 4) 00 en -4 (-4 — - en UI

en W 00 C- LC CC-) CL') -1 <.0 CO tO c-i 'V 'I en - (N CO C 000 12 en a' CO D a '- c") CO C C <.0 CO Cl) en "0' en 00 C'- CO IC) en 00 00 en (N ,-4 ON In a' 0' fr' (N en U-I <-4 UI

en CO 0 C CO -I CO CO (N '.0 00 (N 0) IC) 10 U Ln en CL') U-) C-- CO '.0 o en (N -1 kD en en -v-I PC en (7' 4 (-4 V . U-, Lc- 00 ON C ¼D CO CO '0'en C .-I 00 5-4 C'- C'- L(•i 00 a' -'ci' C-i C- en rio (-4 .-4 (-4 -C i 00 00

CO CO CO C- 40 0 en 0) U') CO "0• 1$') 00 en en ON .-4 <.0 '

'CO (.0

(N (N CO 00 CO. t- 4-I v-laS

>i4 <"I 01 CO CO (a (04-I 0' 0. en en rI CO 0 0 en CO CO CO CO .4: — 00 00 00 00 4-1 .- :l • • - - to-' 0 0' 0' (C i <0 QH C'- .-q - 00000000 CO ---I — C - - <0 '0' CL) €1) CO 0 '-' -4 C 'ci' LC< (NC en [10 o- en BERLINER GEOWISSENSCHAFTLICHE ABHANDLUNGEN Freie Univcrsität Berlin Technische Universität Berlin Technische Fachhochschule Berlin

Noch erhältlich sind: Reihe A: Geologie und PaläontoLogie

Band 135/Hans-Ceorg Herbig: Das Pathogen am SUdrand des zentralen Hohen Atlas und im Mittleren Atlas Marokkos. Stratigraphie, Fazies, Palaogeographie und Paläotektonik. 289 S., 52 Abb., 40 Tat'.. I Tab.. 1991DM 112.—. Band 136 I Jurgen Gauer: Bodencntwicklung und Bodengesellschaftcn voni Mittelmeer zur Qaflara Depression in Nordwestagypten. 171 S.. 49 Abb., 39 Tab., 1991. DM 47,—. Band 137 / Axel Sternkopf: Einsatz der Fernerkundung zur Erkennung von Eisenspatvercrzungen am Steirischen Erzberg/Osterreich. 104 S., 38 Abb., 4 Tat'., 12 Tab., 1991. DM 48.—. Band 138 I LüdvIk Eckardt Gustaisson: Geology and Petrography of the Dyrfjoll Central Volcano, Eastern Iceland, 98 S., 73 Abb., 3 Taf., 1992. DM 68,—. Band 139 / Heinz Hilbrecht: Die fazielle und suukturclle Entwcklung der helvetischen Oberkreide im Allgau und Kleinwalsertal und ihre Beziehung zu Meerwasserspiegelschwankungen und Paläozeanographie. I80S..78Abb.,9Taf., 8Tab., 1991.DM72,—. Band 140/ Christian Sommer-von Jarmersted: Hydraulische und hydrochemische Aspekte der Uferfil- tration an der Unterhavel in Berlin. 169 S., 105 Abb., 9 Taf., 15 Tab., 1992. 1DM 48,—. Band 141 / Christoph Merz: Laboruntersuchungen zum Migrationsverhalten von Cadmium, Zink, Eisen und Mangan in Lockergesteinen unter Wahrung der natUrlichen Milicubedingungen mit Hilfe der Radionuklid- tracerung. 116 S., 75 Abb., 35 Taf., 1992. DM 43,—. Band 142 / Torsten Schwarz: Produkte und Prozesse exogener Fe-Akkumulation: Elsenoolithe und lateri- tische Eisenkrusten im Sudan. 186 S., 136 Abb., 24 Tab., 4 Taf. (meist farbig), 1992. DM 67,—. Band 143/ Erio Rahders: Verteilung von Gold und Silber in den Nebengesteinen der vulkanisch-sedimen- taren Sulfid- und Manganerzlagersthtten des SW-iberischen Pyritgurtels, Provinz Huelva, Spanien. 117 S., 35 Abb., 2 Tat'., 25 Tab., 1992. 1DM 47,—. Band 144 / Abdellah Milhl: Stratigraphie, Fazies und Paläogeographie des Jura am Südrand des zentralen 1-lohen Atlas (Marokko). 93 S., 28 Abb., 6 Tat'., 1992. 1DM 47,—. Band 145* I El Sheikh Mobamed Abdel Rabman: Geochemical and Geotectonic Controls of the M etal lo- genic Evolution of Selected Ophiolite Complexes from the Sudan. 175 pp.. 41 figs., 21 tabs., 4 pls., 1993. 1DM 64,—. Band 146*/Mohamud A. Arush: Sedimentological and Geochemical Studies of the Tisje/Yesornma Successions (Cretaceous / Tertiary of Somalia) and their Source Rocks. 116 pp., 32 figs.. 17 tabs., 1993. DM 56, Band 147* / Alaa El-Din Ramadan Mostafa: Organic Geochemistry of Source Rocks and Related Crude Oils in the Gulf of Suez Area, Egypt. 163 pp., 59 figs.. 24 tabs., 1993. 1DM 54,—. Band 148 / Matthias S. Tantow: Stratigraphie und seismisches Erscheinungsbild des Oberkarbons (Westfal, Stefan), Emsiand. 66 S., 35 Abb., 3 Tab., 13 AnI., 1993. DM 38,—. Band 149* / Wolfgang Mette; Stratigraphic und Fazjes des Jura von Nordsomalia. 125 S., 28 Abb., 15 Tat'., 1993. DM 78,—. Band 150* I Ahmed Ahmed Ali Refaat: Facies Development of the Coniacian-Santonian Sediments along the Gulf of Suez, Egypt. 146 pp.. 35 figs., 28 tabs., 3 pls.. 1993. DM 54.—. Band 151* /Johanna Kontny: Grundwasserverdunstung in ostsaharischen Senkengebieten unter besonderer Berticksichtigung der Transpiration wild wachsender Vegetation. 99 S., 15 Abb., 24 Tab.. 1993. DM 49.—. Band 152* / Evelyne Isaac Mbede: Tectonic Development of the Rukwa Rift Basin in SW Tanzania. 92 pp.. 30 figs., 5 tabs., 1993. 1DM 34,—. Band 153 / Rainer Enillin: Die Kreide des zentralen Mittleren Atlas und der Haute Moulouya, Marokko. Stratigraphie, Mikrofazies, Palaogeographie und Paläotektonik. 85 S., 22 Abb., 10 Taf., 1993. DM 49,—. Band 154 / Michael S. Quednau: Gold in der Kreuzeck- und Goldeck-Gruppe, Kärnten, Osterreich - Geochemie und Metallogenie -. 138 S., 70 Abb., 1993. DM 64,—. Band 155* / Arno Strouhal: Tongeologische Entwicklungstrends in kretazischen und tertiärcn Sedimenten Nordostafrikas: regionale Fallbeispiele. - 68 S., 27 Abb., 22 Tab., 1993. DM 32,50. Band 156/Jonas KIey: Der Ubergang vom Subandin zur Ostkordillere in Sjidbolivien (21°15 - 22° S): Geologische Struktur und Kinematik. - 88 S., 63 Abb., 1 Tab., 1 Karte, 3 Profiltaf., 1994. DM 64,—. Band 157* /Fevzi Oner: Quantitative Mineralbestinimung in Sedimentgesteinen: Anwendung IR-spektro- skopischer Methoden. 140 S., 60 Abb., 17 Tab., 1994. DM54,—.

Beiträge aus dem Sonderforschungsbcreich 69: Geowissenschafcl. Probleme in ariden Gebicten BERLINER GEOWISSENSCHAFTIJCHE ABI-IANDLUNGEN iieic lni\critat Berlin lechnkehe Unix eHtii Berlin Techniehe Fachhochchule Berlin

ocl er1ii1t1 ieli i nd:

Reihe A: Geologie und Paläontologie

Band 158 I Andrea Koschinsk: Geocheiikehe KmutenpioIile und seqLlClltiCIlC laiiguiig ersuelie an rvlan- eanerLLruten au Liem Zentralpazi I1I. mr K kirtiiig von (3enee end Elemenrasoziationen. I 92 S.. 69 Ahh.. 16 Tab.. 1994. DN1 62. . Band 159' I hdi \1()haIlI()LJd Siad : (ieoniatIieniitici1 e\ aluatioti of trace element paticnis in laterite soils above I Sate Proterotoic baenient unil of Niceria. \ket Africa. 122 pp .. figs.. 33 tabs.. 1994. DM 47-. Band 160 I ,Jörii Scheele: Vom frühnievokehen Ri Iterahen /Allll iiirakrnitineiitaleti (]ehiige : Konvergente Blatt\ erehiehungsteLtoni ill ,entralen Miltleren At10 MaiokLo . I 73 S.. 82 Ahh.. 2 lat.. 1994. DM 64.-. Band I 6 1 I N1oIianed Za'ed Awad : Siratigiapliie, paIr nologieal and paleoecological stLlClieS in the east- Lnn d SLtd in ( Kh utoum LOW Kosti U o.iiis ) I ie Rll-assiC to mid I .i ti a 16 lP 25 ti .s 6 t ihs 18 pis I9)4. L)v1 55--. Baud I 62 I Elke \Iages : VLirietiansportinodei lierong 011 [rltelgrLlnd dcs SEIJI iclien RaLlines von Berlin. 79 S.. 55 \bh.. I I lth.. 1994. OM 6.-. Baud 163 / Holger (ebIiardt: Die Tap-\iergel iii der Provint Alieante (Siiclostpanien. Langhiuiii his Tortoniuni): Biostnoigraphie. Palioeoaaphie und Pikioikologie. 175 S .. 57 Ahh.. 44 Iah.. 14 Tal.. I 9)4. L)M BaIl(I 164 I Ilias K. Gerolviiiatos: le1ainoiphose und Tekionih der Phvlli-QuarLIt-Serie und rler Tyros- Sehiehten aLil dern PCI0p011llCS Ulki Kr ilìi ra. I 0 I S ... .Abb.. 16 Tah.. 4 Tat., 1994. DM 45.-. Baud 165 I t,titz ThOflas: Hvdrogeoelieiiiisehe Untersuehungen an 011cldwiissern aus N\-Deitsehiiirl tind dent Oberrheingraben und ihre Model I ierung unter dern :\spekt der Uniwicklung clues Expertensvsteins für Fluid-Rock-lntei'actions lXPS FROCK!. 160 5.. 75 Abh .. 34 Tab.. 34 Ani.. 1994. DM 54,- Rand 166* / Eckhard Ludwig Wipiler: Geoclienusehe. strrikturelle rind erimikroskopische Llntersuchungen fLu I. LUst uttLnLntss uklun. d 5' st luLhL I \i 1 ib N if. ti h BL it RL cI SL I Pi iP in! N F SLId in 206 5 55 Ahb.. 9 Tab.. 1994. [)M 69.-. Rand 167 / Klaus Gerniann, Bernd Lehniann & Franz K. List ( Hrsg. ): F'estsehritl turn 70. Geburtstaui son Professor Di'. Hans-.loehen Sehneideu. VIII + 226 5.. 56 ,\hh.. 13 lab.. 1994. DM 52.-. Rand 168 / Johannes H. Schroeder 1 Hrsg.> : Fuurtschrilte in der Geologic son Rhdersclort. 375 5.. 102 Ahh.. 34 Taft. 30 Taf.. 1995. DM 55.-. Rand 169 / rxeI v. llillebrandt & \lanI'red GrUschke : Aununonuien aLls rlern Callos iLuni /Oxfordium-Grent- hereich son Nordchile. 40 S.,5 ,\bb.. 6 l'al.. 1995. DM 36,--. Band 170 / Beitrge zum IFP 16/02: l'i'ansportverhalten \oil anthropogenen Schadstoffen in tonigen Bar- rieren. 149 S.. 59 Abb .51 lab,. 1995. DM 47.--. Band 171 / Christian Thom: Konieptuonelle Planung und Prototr ping clues ss issenbaiei'ten S steiiis mr Ansprache puLntiLll flLuldtnhit nds I F oi in ltlonshLiLlchL im Ki st lllinaestLln 14 5 69 \hb 9 I ib 995. DM 49. Rand 172 / .Johst Wurl: Die geologischcn. 11\ draLuiischen nnd h droehernischen \erhLi1tnisse in den sudss est- lichen Stadtbeiirken son Berlin. 164 S.. 63 Abb...2 Tab.. 1995. D\1 69.-. Rand 173 / Flendrik Siegmund: Faties und Genee unterkarnhnseher Phosphorite und mariner Sedimente der Yangtte-Plattform. Shdchina. 114 S., 32 Abb.. 29 'lab.. It) Tal.. 1995. DM 54.-. Rand 174 / Birgit Niebuhr: Faties-Differentiei'ungen rind ihre Steuerungsf'aktoren in der hOhereri Oherkreide on S-Niedersachsen / Sachsen-Anhaii (N-Deritsehland . 13 I 5.. 43 Abb.. 6 'lab.. 12 I at.. 1995. DM 76.-. Band 175 / Roll Kohring & Thomas Schuiiter (eds.): Curient Geoseuentilie Research in Uganda and Tan- tania. 166 pp.. 6 pls.. IS tabs.. 92 ligs.. 1995. DM 80.-. Hand 176 / iörg Heftier: Bergbau und Unuss cit in Papua-Neugtiinea: DIe OR 'ledi-Mune rind das Fly River- Flriifiikosr stem. 152 S.. 29 .\hb.. 17 lab.. 2 Tal'.. 1995. DM Band 176-11 / .JUrg 1-lettler & Bernd Lehmann: En\ uronmental Impact of 1.arge-Scale Mining in Papua Ness Guinea: Mining Residue Disposal b\ the OR Tedu Copper-Gold Mine. 52 pp..31 figs.. 17 tabs.. 2 maps. 1995. DM51.--. hand 177' / Fekart Schrank & Mohamed I. A. Ibrahirn: Ci'etaceous Aptian-Maastrichtian ) palvno1og of Ioi nuinulLi i d Uvd ss us (KR\1 I \G 1 8) in noi thss Lstu, in Fs Pt 44 pp 10 IL \t 1ig t ibs 9 pls 1995. DM42.--. Band 178* / \Iumin \loiiamud God: Climatic conditions deduced from I .ate Pleistocene deposits at Karin gap (NE-Somalia. 145 PP. 47 tcxt-figs.. 20 tabs.. 1995. DM 53.-. Band 179 / Sun Yong Bin; Radiogene Isotopengeochensie des li'oodos-Ophioliths. Lntersuchuiigcn Haupt- rind Spurenclemente sosvie Si-. Nd- Lund Ph-Isotope I der iyprischen Ozeankruste. IX + 137 S.. 44 Abb.. 23 Tab.. 1995. DM 54.--.

Beitrüge ans dciii Sonderforsehungshei'eich 69: Geoss issenschaltl. Problerne in ariden Gehietcn BERLINER GEOWISSENSCHAFTLICHE ABHANDLUNGEN .' Freie Universität Berlin Technische Universität Berlin 'Technische Fachhochschule Berlin

Reihe B: Geophysik Band 20 I Michael Schmitz: Kollisionsstrukturen in den Zentralen Anden: Ergebnisse refraktionsseismischer Messungen und Modellierung krustaler Deformationen. S. 1-127, 78 Abb., 2 Tab., 1993. Wolf-Dieter Heinsohn: Druck- und Temperaturabhangigkeit der Geschwindigkeit-Dichte-Relation für extrem groBe Krustenmächtigkeiten. S. 13 1-226, 58 Abb., 9 Tab., 1993. DM78,—. Band 21 / Detlef Kruger: Modellierungen zur Struktur elektrisch leitfahiger Zonen in den sudlichen zentralen Anden. 91 S., 60 Ahb., 1994. DM 67,—. Band 22 / Christian Kiesper Die rechnergestutzte Modellierung eines 3D Flachenverbandes der Erftscholle (Niederrheinische Bucht). 117 S., 51 Abb., 1994. DM 69,—.

Reihe D: Geoinformatik Band 3 / Christiane C. Schmullius: Radarfernerkundung landwirtschaftlicher Flächen mit einem flugzeug- getragenen L-, C- und X-Band Scatterometer. 111 S., 77 Abb., 7 Tab., 1992. DM 42,—. Band 4 /Junfeng Luo: Konditionale Markovsimulation 2-dimensionaler geologischer Probleme. 103 S., 55 Abb., 38 Tab., 1993. DM 48,—. Band 5/Franz K. List & Peter Bankwitz (Hrsg.): Proceedings of the Fourth United NationslCDG International Training Course on Remote Sensing Applications to Geological Sciences, held at Potsdam and Berlin, September 28 to October 16, 1992. 120 S., 88 Abb., 4 Tab., 1993. DM 38.—. Band 6/Christian Bauer: Digitale Bildverarbeitung von groLraumigen Satellitenhild-Verbänden: Mosaik- bildung, multitemporale Analyse und Klassifizierung. 97 S., 35 Tab., 58 Abb., 1994. DM 48,—. Band 7 / Gerold Reusing: Contribution to the Risk Analysis of Hydrologic Droughts Based on Hydrologic Time Series of the River Nile. 125 5., 74 Abb., 24 Tab., 1994. DM54,—. Band 8/ Sabine Balmer-Heynisch: Erfahrungen bei der Erstellung von rechnergestUtzten Bewertungssyste- men für die Errnittlung des Grundwassergefährdungspotentials kontaminierter Altstandorte im Hinblick auf SanierungsinaBnahmen. 208 S., 1995. DM 74,—. Band 9 / Jörg Tietze: Geostatistisches Verfahren zur optimalen Erkundung und modellhaften Beschreibung des Untergrundes von Deponien. 96 S., 1995. DM51,—. Band lO/Andreas Neutze: Fin Beitrag zur geostatistischen Raum-Zeit-Prognose. Anwendungsbeispiel H bodennahes Ozon. 119 S., 1995. DM59,—.

Reihe E: Palaobiologie Band 12/Ulrike Kienel: Die Entwicklung der kalkigen Nannofossilien und der kalkigen Dinoflagellaten- Zysten an der Kreide/Tertiär-Grenze in Westbrandenburg im Vergleich mit Profilen in Nordjutland und Seeland (Däriemark), 87 S., 27 Abb., 15 Taf., 1994. DM 68,—. Band 13 / Roif Kohring & Thomas Martin (Hrsg.): Miscellanea Palaeontologica 3. 531 S., 67 Taf., 1994. DM248,—. Band 14/ Annette Richter: Lacertilia aus der Unteren Kreide von Uña und Galve (Spanien) und Annual (Marokko). 147 S., 65 Abb., 6 Taf., 1994. DM 48,—. Band 15 / Michael Steiner: Die neoproterozoischen Megaalgen SUdchinas. 146 5., 86 Abb., 20 Taf., 1994. DM 72.—. Band 16.1 / Roif Kohring & Markus Wilmsen (Hrsg.): Miscellanea Palaeontologica 4. 384 5., 56 Taf., 1995. DM 139,—. Band 16.2 / Rolf Kohring & Markus Wilmsen (Hrsg.): Miscellanea Palaeontologica 4. 406 S., 44 Taf., 1995. DM 158,—. Band 17/Fritz Neuweiler: Dynamische Sedimentationsvorgange, Diagenese und Biofazies unterkreta- zischer Plattformränder (Abt/A1b; Soba-Region, Prov. Cantabria, N-Spanien). 235 S., 95 Abb., 61 Taf., 1995. DM 179,—.

Das vollstiindige Verzeichnis der lieferbaren Titel der Reihen A, B, C, D und E ist erhältlich bei:

Freie Universität Berlin, Fachbereich Geowissenschaften, Selbstverlag Malteserstrasse 74-100 12249 Berlin. Germany