S Monxana i»Taxe

622*5 University

M41hvin { Bozeman )•

19SS Hydro geochennical

t veget at ionai and
microbiological
effects of a

HYDROGEOCHEMICAL, VEGETATIONAL AND

MICROBIOLOGICAL EFFECTS OF A NATURAL

AND A CONSTRUCTED WETLAND ON

THE CONTROL OF ACID MINE DRAINAGE

by

Reclamation Research Unit and IMS! Detoxification Inc.

Montana State University 900 Technology Blvd
Bozeman, Montana 59717 Bozeman, Montana 59715

STATE DOCUMENTS COLLECTIO^

SEP 4 1980

HELENA, MONTANA 59620

'repared for

The Montana Department of State Lands, Abandoned Mine Reclamation Bureau, Helena, MT
Reclamation Research Publication 88-04 June 1988

sta n '-, >i MONTANA STATE LIBRARY

<Jsti — d—d — H^'^ij SS22.SM4tlnm19>8c.1

ifitr -i ri ■, ^ Hydrogeocli«iiiic«i, vegetatlonal and micro

^^'^17 2Q02 APOM 00062713 6

W ' DISCLAIMER NOTICE

''The views and conclusions contained in this document are those
of the authors and should not be interpreted as necessarily
representing the official policies or recommendations of the
Montana Department of State Lands.

Reference to specific brands, equipment or trade names in this
report is made to facilitate understanding and does not imply
endorsement by either the Montana Department of State Lands,
Montana State University or MSI Detoxification Incorporated.

ACKNOWLEDGMENTS

The authors wish to express their appreciation to Lincoln,
Montana resident Mr. D.R. Rundell whose able service greatly
facilitated project installation and data acquisition, and to
Mr. Mike Hiel, Montana Department of State Lands, for helpful
suggestions and support. The Lincoln Ranger District of the
Helena National Forest has provided valuable aerial photography
and climatic data which expedited portions of this study.

Dr. William P. Inskeep (MSU) , Dr. Eugene Madsen (MDI) ,
Christie Nordstrom (MSU student) , and Sean Koenig (MSU student)
provided technical assistance during the microbiological
investigation and their help is appreciated. Thanks is extended
to Dr. Edward R. Dratz (MSU) who provided a Spectronic 20
colorimeter for use in the study.

PREFACE

Staff of the Reclamation Research Unit were responsible for
investigation of the wetland hydrology, geochemistry, water
quality and vegetation. Dr. Douglas J. Dollhopf (Soil Scien-
tist) was the Principal Investigator, John D. Goering was the
Geohydrologist, Robert B. Rennick was the Vegetation/Wildlife
Scientist and Robert B. Morton was the Graduate Research
Assistant (Geohydrologist) . These investigators prepared
Sections 1.0 through 9.0.

Staff from MSI Detoxification Incorporated (MDI) conducted the
wetland microbiological investigation as reported in Section
iO.O. Mr. Peter M. Jones was Project Manager, Dr. W. Kennedy
Gauger was Senior Microbiologist, Dr. Jim B. Guckert was the
Microbial Ecologist, Dr. Keith C. Cooksey was the Lipid Biochem-
ist, Karen Bucklin was the Environmental Engineer, Rebecca Weed
was the Geochemist and Margaret M. Lehman was the Toxicologist.

Kmn

REPORT DOCUMENTATION PAGE

Title and Subtitle

Hydrochemical, Vegetational and Microbiological Effects
of a Natural and a Constructed Wetland on the Control
of Acid Mine Drainage

Authors D.J. Dollhopf, J.D. Goering, R.B. Rennick, R.B.
Morton (MSU) ; W.K. Gauger, J.B. Guckert, P.M. Jones,
K.C. Cooksey, K.E. Bucklin, Rebecca Weed, M.M. Lehmar

Report. Date

June 1988

-.ehman
(MDI)

Performing Organization
Report No.

RRU 8804

Performing Organization Name and Address

Reclamation Research Unit

Montana State University

106 Linfield Hall

Bozeman, MT 59717
(406^ 994-4821

MSI Detoxification, Inc.
900 Technology Boulevard
Bozeman, MT 59715
(406) 586-4885

Contract No.

Grants/Contracts
2-6000-230

Sponsoring Organization Name and Address

Montana Abandoned Mine Reclamation Bureau
Montana Department of State Lands
Capitol Station
Helena, MT 59620 (406) 444-2074

Type of Report
and Period Covered

Final Report
March 1987 -
April 1988

Abstract An evaluation of a natural wetland in western Montana, which has receiv-
ed acid mine drainage (AMD) from an abandoned lead/zinc mine for at least 40
years, indicated that the system is effective in removing iron from influent
water, but is less effective in removing manganese and other metals. Approxi-
mately 550 metric tons of iron have been deposited in the wetland, while only
about 1 metric ton of manganese has been deposited.

Water velocity through the Swamp Gulch wetland was most accurately deter-
mined using the bromide tracer method, rather than measurements of hydraulic
conductivity. Mean velocity was 2.6x10"^ cm/s(74 ft/day) and ranged from 3.1
xl0~^ to 6.2x10"^ cm/s.

Dominant plant species included Carex rostrata , Betula glandulosa , Salix
boothi i and Sphagnum tenellum . Tissues of these plants exhibited elevated
metal levels. Metal concentrations were generally well below phytotoxic level
however, levels of cadmium in some plant tissues exceeded the maximum dietary
intake level for domestic animals. This indicates a potential risk to animals
whose diet contain large proportions of these plants.

Anaerobic microorganisms, as well as abiotic processes are responsible for
acid reduction in the Swamp Gulch wetland. Sulfate-reducing bacteria are func
tioning at the natural Swamp Gulch wetland, but not at the constructed Sand
Coulee wetland. This is most probably a function of adequate food and energy
sources at the natural wetland. These studies strongly suggest that poor per-
formance by the constructed wetland is due to reduced microbial activity rather
than its relatively small size. The microbial component of natural wetlands
should be transferrable to artificially constructed systems.

Keywords

Acid Mine Drainage, Heavy Metals, Wetlands

11

TABLE OF CONTENTS

PAGE

Disclaimer/Acknowledgements/Preface i

Report Documentation Page. ii

Table of Contents iii

List of Tables vii

List of Figures ix

1.0 INTRODUCTION 1

1.1 ACID MINE DRAINAGE 1

1.2 STUDY OBJECTIVES 2

1.3 CONCLUSIONS 2

1.3.1 Wetland Hvdroaeochemistry 2

1.3.2 Wetland Veaetation/Wildlife 4

1.3.3 Wetland Microbiology 5

1.4 RECOMMENDATIONS 6

1.4.1 Wetland Hvdroqeochemstry 6

1.4.2 Wetland Veaetation/Wildlife 7

1.4.3 Wetland Microbiology 7

2.0 NATURAL SYSTEMS FOR WATER POLLUTION CONTROL 9

2.1 WETLAND TREATMENT SYSTEMS 9

2.2 BIOSORPTION OF ELEMENTS BY PLANTS 10

2.3 METALLOPHYTES AND LAND/WATER RECLAMATION 11

2.4 CHEMICAL PROCESSES OF ACID MINE DRAINAGE 11

2.4.1 Acid Water Formation 11

2.4.2 Bacterial Mediation 11

2.4.3 Influence of pH 11

2.4.4 Pvrite Oxidation Control 12

3.0 SITE DESCRIPTION 13

3.1 LOCATION 13

3.2 NATURAL ENVIRONMENT 13

3.3 NATURAL WETLANDS 15

3.3.1 Definition and Classification 15

3.3.2 Classification of the Swamp Gulch Mine

Wetland Study Site 16

4.0 SURFACE HYDROLOGY OF A NATURAL WETLAND 18

4.1 METHODS 18

4 . 2 RESULTS AND DISCUSSION 19

5.0 HYDROGEOLOGY OF A NATURAL WETLAND 24

5.1 LITERATURE REVIEW 24

111

PAGE

5.1.1 Hydraulic Principles of Tracer Use 24

5.1.2 Tracer Flow Retardation and Tracer Dilution . . 26

5.1.3 Tracer Methods 28

5.1.4 Tracer Problems 29

5.1.5 Bromide as a Tracer 30

5.1.6 Wetland Hydrology .34

5.2 MATERIALS AND METHODS 38

5.2.1 Site Instrumentation 38

5.2.2 Aquifer Characteristics 38

5.2.3 Hydraulic Conductivity 38

5.2.4 Tracer Experiment . 41

5.3 RESULTS AND DISCUSSION 42

5.3.1 Aquifer Characteristics 42

5.3.2 Hydraulic Conductivity 47

5.3.3 Tracer Experiment 52

5.4 SUMMARY 63

6.0 WETLAND INFLUENCE ON MINE DRAINAGE WATER QUALITY 65

6.1 SAMPLE COLLECTION AND PROCESSING PROCEDURES 65

6.2 ANALYSES METHODS 67

6.3 QUALITY CONTROL/QUALITY ASSURANCE 67

6.3.1 Accuracy 67

6.3.2 Precision 70

6.3.3 Cross Contamination 70

6.4 SURFACE WATER QUALITY 70

6.4.1 Swamp Gulch Above the Carbonate Mine 70

6.4.2 Swamp Gulch Below the Carbonate Mine 72

6.4.3 Blackfoot River 74

6.4.4 Background Statiscal Comparisons 74

6.4.5 Wetland Surface Water 75

6.5 GROUND WATER QUALITY 76

6.5.1 Shallow Ground-water System 76

6.5.2 Deep Ground-water System 82

6.5.3 Acid Mine Drainage Seepage Wells 85

6.6 SUMMARY 85

7.0 DISSOLVED METAL LOADING OF A NATURAL WETLAND. ........ 87

7.1 METHODS . 87

7.2 RESULTS AND DISCUSSION 88

8.0 WETLAND GEOCHEMISTRY AND THE CONTROL OF ACID MINE DRAINAGE. . 89

8.1 METHODS 89

8.2 RESULTS AND DISCUSSION 90

8.2.1 Areal Distribution of Selected Elements .... 90

8.2.1.1 Iron 90

8.2.1.2 Manganese 92

8.2.1.3 Aluminum 93

8.2.1.4 Copper . 94

IV

PAGE

8.2.1.5 Lead 94

8.2.1.6 Zinc 97

8.2.1.7 Arsenic 98

8.2.1.8 Barium and Beryllium 98

8.2.1.9 Cadmium 98

8.2.1.10 Cation Exchange Capacity 99

8.2.2 Distribution of Elements with Depth 99

8.3 SUMMARY 101

9.0 WETLAND VEGETATION AND THE CONTROL OF ACID MINE DRAINAGE. . . 103

9.1 SAMPLE COLLECTION, PREPARATION AND ANALYSIS 103

9.2 DATA QUALITY ASSURANCE/ QUALITY CONTROL 104

9.3 RESUL-^S AND DISCUSSION 105

9.3.1 Vegetation Patterns 105

9.3.2 Canopy Coverage and Production 105

9.3.3 Element Concentrations in Plant Material . . . . 108

9.3.3.1 Effects of Carex rostrata in
Remediating AMD lio

9.3.3.2 Effects of Salix boothii in
Remediating AMD 115

9.3.3.3 Effects of Betula glandulosa in
Remediating AMD 118

9.3.3.4 Effects of Bryophytes in Remediating

AMD 118

9.3.4 Summary of Element Enrichment in Plants .... 121

9.3.5 Vegetation Element Levels and Wildlife

Hazards 123

9.3.5.1 Dietary Hazard Levels for Animals . . . 123

9.3.5.2 Wildlife Observed and at Risk 124

9.3.5.3 Hazards to Wildlife at the Swamp

Gulch Wetland . 125

10.0 MICROORGANISM AND THE IMMOBILIZATION OF HEAVY METALS IN

WETLANDS 126

10.1 BACKGROUND 126

10.2 PHASE I: SWAMP GULCH WETLAND BUFFERING CAPACITY AND

THE EFFECT OF SOIL STERILIZATION ON WETLAND PH. . . . . 127

10.2.1 Introduction 127

10.2.2 Technical Approach . 127

10.2.3 Materials and Methods . . ...... 127

10.2.3.1 Soil Sampling . 127

10.2.3.2 Titration Experiment ........ 127

10.2.3.3 Microbiological Experiment 128

10.2.4 Results and Discussion 129

10.2.4.1 Titration Experiment . . 129

10.2.4.2 Microbiological Experiment 129

10.3 PHASE II: COMPARISON OF NATURAL AND CONSTRUCTED
WETLAND SOILS IN THE REMOVAL OF ACIDITY, SULFATE AND

IRON 131

L

PAGE

10 . 3 . 1 Introduction 131

10.3.2 Technical Approach . 131

10.3.3 Materials and Methods 132

10.3.3.1 Field sampling . . 132

10.3.3.2 Soil core flow characteristics . . . 135

10.3.3.3 Sampling and preparation of AMD . . . 135

10.3.3.4 Sulfate and iron determinations . . . 135

10.3.4 Results and Discussion 136

10.3.4.1 Acidity determinations 136

10.3.4.2 Sulfate determinations 140

10.3.4.3 Iron determinations 145

10.4 PHASE III: MICROBIAL BIOMASS, COMMUNITY STRUCTURE AND
PHYSIOLOGICAL STATUS ASSESSMENT OF NATURAL AND

CONSTRUCTED WETLANDS 149

10.4.1 Introduction 149

10.4.2 Technical Approach 149

10.4.3 Materials and Methods . . 150

10.4.3.1 Field sampling 150

10.4.3.2 Statistical analyses . . 150

10.4.3.3 Lipid nomenclature . . 150

10.4.4 Results 151

10.4.4.1 Total Biomass 151

10.4.4.2 Community structure . . 151

10.4.4.3 Physiological stress ........ 156

10.4.5 Discussion . 159

11.0 LITERATURE CITED 162

APPENDICES 178

APPENDIX A: Analytical Methods and Quality Assurance/ Quality

Control Statistics 178

APPENDIX B: Influent Flow Rates, Precipitation and Water Chemistry

Data from the Swamp Gulch Wetland 187

APPENDIX C: Chemical and Physical Data from Swamp Gulch Wetland

Sediments 199

APPENDIX D: Precision and Accuracy of Vegetation Data, and Plant

Plant Species Observed at the Swamp Gulch Wetland . . .210

yi

LIST OF TABLES

TABLE PAGE

4-1 Beaver dam tree ring counts 21

5-1 Range of important physical characteristics of fibric,

hemic, and sapric plant materials from northern Minnesota

bogs 37

5-2 Wetland stratigraphy and well completion data. ....... 45

5-3 Field hydraulic conductivity (K) , effective porosity (ne) ,
average hydraulic gradient (dh/dl) and calculated average
velocity (V) for each site • 51

5-4 Average hydraulic conductivity (K) from field results and

calculated average velocity (V) 52

5-5 Bromide background concentrations in the tracer study area

of the Swamp Gulch Mine Wetland . 53

5-6 Tracer velocities determined for each site, time and

distance from tracer inputs and peak concentrations of

bromide 60

5-7 Tracer velocity comparisons 61

6-1 Accuracy results based on blind field standards 68

6-2 Accuracy results based on laboratory spikes. . 69

6-3 Precision, based on laboratory duplicates, at the 90%

confidence level . 71

6-4 Precision, based on field replicates, at the 90 confidence

level 72

6-5 Field blank. 73

7-1 Total dissolved metal loading on wetland, April 1987

through March 1988 88

8-1 Mean selected elemental concentration changes with depth

for background and AMD impacted areas . 100

8-2 Acrotelm metal masses and the relative efficiency of wetland

metal removal 101

9-1 Mean percent canopy coverage and percent frequency for

plant species 107

vii

TABLE 2^^

9-2 Peak standing crop (kg/ha) at the Swamp Gulch site and at

the Hardscrabble background site 108

9-3 Concentration (ug/s) of elements in Carex rostrata.

109

9-4 Water quality in Swamp Gulch above and below the Carbonate
mine site, and below the Swamp Gulch wetland at station
F-i 110

9-5 Range and mean element concentrations (ug/g) in aquatic

forbs and grasses (from Hutchinson 1975) HI

9-6 Concentration (ug/g) of elements in Salix boothii . 116

9-7 Concentration (ug/g) of elements in Betula glandulosa. . . . 119

9-8 Accumulation of elements by bryophytes in various

environments ^20

9-9 Concentration (ug/g) of elements in bryophytes 122

9-10 Plant species demonstrating element enrichment (i.e.

greater than background levels) at the Swamp Gulch wetland

site 123

9-11 Maximum tolerable levels of dietary minerals for domestic

animals 124

10-1 pH and iron values for AMD incubated under various

treatment conditions after a period of nine days. ..... 130

10-2 Swamp Gulch sulfate concentrations vs. flow rate 144

10-3 Sand Coulee sulfate concentrations vs. flow rate. ..... 144

10-4 Sand Coulee iron concentration vs. total flow volume eluted

at end of study • • • 149

10-5 Fatty Acid (FA) mole % for the various sampling sites. . . . 154

10-6 Description of phospholipid ester-linked fatty acids (PLFA)

used to define microbial groups for data interpretation. . . 155

10-7 Significant difference map generated from Tukey's HSD test. 157

Vlll

LIST OF FIGURES

FIGURE PAGE

3-1 Location of Swamp Gulch Wetland Study site in Montana. ... 14

4-1 Swamp Gulch Wetland Study site (with water sampling

stations) 19

4-2 Topographic map with 0.5 m contours. . 20

5-1 Breakthrough curves of four different tracer types. .... 25

5-2 One dimensional example of movement by molecular diffusion. 26

5-3 One dimensional example of hydrodynamic dispersion for

tracer particles A-G 27

5-4 Ranges of published field data on K (hydraulic conductivity)

of peat. 34

5-5 Example of study site instrumentation and lithology. ... 39

5-6 Map of auger hole grid 40

5-7 Acrotelm isopach map with 0.2m (0.7 ft) contours 43

5-8 Catotelm isopach map with 0.5 m (1.6 ft) contours. . . . . . 44

5-9 Deep piezometer map on 7/27/87 with 0.5 m contours. .... 48

5-10 Shallow piezometer map on 7/27/87 with 0.5 m contours. ... 49

5-11 Potentiometric surface map on 8/14/87 with 0.025 m

contours 50

5-12 Typical calibration curve generated for the tracer study. . 55

5-13 Tracer study breakthrough curve for site AH-14, 6.1 m from

tracer input showing little effect of sorption 56

5-14 Example of infrequent sampling effect on breakthrough curve

for site AH-9, 6.1 m from tracer input. 57

5-15 Breakthrough curve representing two tracer plumes at site

AH-19, 9.1 m from tracer input 58

5-16 Steep breakthrough curve indicating adequate tracer mixing

at site AH-3, 3 m from tracer input 58

5-17 Bromide concentration isopach map, 0.95 to 1.5 hours from

tracer start (20 mg/L contours) 62

ix

FIGURE p^

6-1 Shallow ground-water types for the wetland system 77

6-2 Iron concentrations for the wetland shallow ground-water

system. , 7g

6-3 Manganese concentrations for the wetland shallow ground-
water system. 79

6-4 Sulfate concentrations for the wetland shallow ground-water

system 80

6-5 Total dissolved solids concentration for the wetland

shallow ground-water system 81

6-6 Iron concentrations for the wetland deep ground-water

system g3

6-7 Manganese concentrations for the wetland deep ground-
water system. ^ 84

8-1 Iron concentrations in the wetland acrotelem 91

8-2 Manganese concentrations in the wetland acrotelem 92

8-3 Aluminum concentrations in the wetland acrotelem 93

8-4 Copper concentrations in the wetland acrotelem. ...... 95

8-5 Lead concentrations in the wetland acrotelem. 95

8-6 Zinc concentrations in the wetland acrotelem. 97

9-1 Vegetation associations and canopy coverage transects. . . 106

10-1 Location of wetland sites 128

10-2 Titration of Swamp Gulch soil with AMD 129

10-3 Diagram depicting soil core containers used in Phase II

greenhouse study 2.33

10-4 Sand Coulee constructed wetland sampling site (Heil and

Kerins 1988) 134

10-5 Influent (eluent) AMD pH measured over the study period. . . 137

10-6 Swamp Gulch soil core's effluent (eluate) pH measured over

study period 3^38

FIGURE PAGE

10-7 Sand Coulee soil core's effluent (eluate) pH measured over

study period 139

10-8 Influent (eluent) AMD iron and sulfate concentrations

measured over study period 141

10-9 Swamp Gulch and Sand Coulee soil core effluent (eluate)

sulfate concentrations measured over study period. .... 143

10-10 Swamp Gulch soil core effluent (eluate) dissolved total and

ferrous iron concentrations measured over study period. . . 146

10-11 Sand Coulee soil core effluent (eluate) dissolved total and

ferrous iron concentrations measured over study period. . . 148

10-12 Total microbial biomass as measured by analysis of

phospholipid ester-linked fatty acids. . . ... 152

10-13 Sul fate-reducing bacterial phospholipid fatty acid
biomarker expressed as density per gram of soil and
proportion of total microbial biomass. 153

10-14 Microbial community trans/cis and cyclopropyl/cis

phospholipid ester-linked fatty acids ratios. ....... 158

XI

1.0 INTRODOCTION

1.1 ACID MINE DRAINAGE

Acid mine drainage (AMD) , from coal mining in the East and
both coal and hard-rock mining in the West, is one of the most
persistent environmental pollution problems in the United States.
Acid water draining from these sites lowers ground and surface
water quality, impacts aquatic and terrestrial biota, and
degrades domestic water supplies. An estimated 11,000 miles of
rivers and streams in Appalachia have been negatively impacted by
AMD (Brodie et al. 1986). More than 5,000 miles of waterways in
the East and mid-West fail to meet federal water quality
standards owing to acidic mine water (Kim et al. 1982). In
Colorado alone, an estimated 10,000 abandoned and inactive metal
mines are point sources of AMD, and at least 25 watersheds and
450 stream miles have been rendered barren of aquatic life
(Guertin et al. 1985).

Artificial wetlands have been proposed as an approach to
provide low-maintenance, broadly applicable solutions to AMD
problems and numerous state agencies are currently investing
considerable sums of money in the construction of wetlands to
treat AMD. The efficacy of these projects is variable (Burr is
1984) . Some wetlands improve water quality to meet drinking
water standards. Others are less effective and provide only
temporary metal uptake, selective metal uptake, seasonal
effectiveness, or even complete abatement failure. Poor wetland
performance may be attributed to inappropriate hydrology or
vegetation. Flow channeling may cause polluted discharge to
bypass remediation stages; transplanted aquatic plants may be
unable to tolerate AMD loading or to take up sufficient metals to
significantly abate AMD (McHerron 1985, Wieder and Lang 1986) .
However, some wetlands with improved hydrologic conditions and
well-established metal-tolerant plants still do not perform well,
suggesting that other factors, such as microbial activity and
sediment geochemistry, are of overriding importance at some sites
(Kleinmann and Girts 1986, Lang and Wieder 1985, Peccia &
Associates 1987) .

The wetlands characterized in this study afforded a unique
opportunity to observe the effects of more than 55 years of
mining related effluent and AMD upon a natural wetland (Swamp
Gulch wetland) . In addition, this wetland system could be
compared to a non-polluted system, and to a recently constructed
AMD wetland treatment system located near Sand Coulee, Montana.
The findings will be valuable in the design of future artificial
AMD wetland treatment systems by the Montana Abandoned Mine
Reclamation Bureau.

1.2 STUDY OBJECTIVES

Specific objectives of this study were to:

1) determine the areal and vertical gradients of metal
loading in the natural wetland;

2) assess contaminant transit time in the natural wetland;

3) quantitatively describe plant associations and
determine production by dominant species in the
natural wetland;

4) assess metal loading of dominant vegetation and potential
impacts to wildlife in the natural wetland;

5) estimate the effective longevity of the wetland to
control the acid mine drainage in the natural wetland;

6) correlate wetland mass and volume as a function of the
quality and quantity of acid mine drainage in the natural
wetland; and

7) compare the relative biological activity of the natural
wetland to the artificially constructed wetland.

1.3 CONCLUSIONS

1.3.1 Wetland Hydroaeochemistry

1. The Swamp Gulch natural wetland is characterized by three
basic stratigraphic units: 1) the surficial material
(average thickness 0.6 m) composed of undecomposed organic
debris, referred to in this report as the acrotelm; 2) the
intermediate gray muck zone (average thickness 2.2 m)
consisting of decomposed organic materials, silt and clay
with occasional sandy gravel lens, generally referred to in
this report as the catotelm; and 3) the underlying alluvial
gravel zone.

2. Trace metal concentrations were highest in the acrotelm near
the Swamp Gulch discharge point and decreased both laterally
and vertically away from this point.

3. The only catotelm interval notably impacted by the acid mine
drainage was near the discharge point, with trace metal
levels in the remaining catotelm being approximately equal
to measured background concentrations.

L

4. Trace element concentration distribution in the underlying
alluvial gravel material showed only a nominal association
with the Swamp Gulch acid mine discharge.

5. The present iron treatment area of acid mine drainage is
confined to approximately 3.9 ha (9.6 acres) and represents
an acrotelm volume of about 23,000 m^ with an average
thickness of 0.6 m. This volume treats a daily acid mine
drainage inflow of 45 m-^ containing a volume weighted
average of 21 mg/L iron, 3.9 mg/L manganese, 6.5 mg/L
aluminum, 2.7 mg/L zinc and 0.71 mg/L copper at a pH of 4.0.
The system has in the past, been effective in reducing iron
levels in effluent waters.

6. Approximately 550 metric tons of iron have been deposited in
the wetland acrotelm from the Swamp Gulch source during the
past 55 years. Under present (1987) loading rates (338
kg/year) this represents a 1600 year accumulation, which
strongly suggests past loading rates were much higher than
those observed during the study period.

7. Conditions conducive to the physio-chemical precipitation of
iron may not be conducive to creating optimum conditions for
microbial buffering of pH and sulfate reduction. This may
explain why Swamp Gulch cores in the laboratory were
ineffective in removing iron but, at the same time, the
above mentioned iron was present in the acrotelm.

8 . Lead is apparently effectively removed from the AMD within
the wetland. The AMD attributed acrotelm lead mass is 1.3
metric tons which represents 2,350 years of deposition at
present loading rates.

9 . The acrotelm copper mass due to AMD is approximately four
metric tons which represents 325 years at the observed
annual loading rate. This metal is apparently only
partially ameliorated within the wetland.

10. Zinc, manganese and cadmium acrotelm masses due to AMD are
6.0, 1.1 and 0.004 metric tons, respectively. These
represent deposition intervals of 137, 16.9 and 6.7 years,
respectively. It is clear that the wetland efficiency in
removal of these metals is limited.

11. Using lead as a basis of comparison of the Swamp Gulch
wetland efficiency in removing metals from AMD, the removal
rates are: iron, 70 percent; copper, 14 percent; zinc, 5.8
percent; manganese, 0.7 percent; and cadmium, 0.3 percent.

12 . Bromide was a good choice for tracing water movement in the
natural wetland studied. Employing a natural flow tracer
technique and the bromide specific ion electrode method of
analysis, sufficient bromide concentrations were detected.

13. The most representative flow velocity of contaminated Swamp
Gulch water in the wetland is that determined from a bromide
tracer test. The tracer determined mean flow velocity is
2.6 X 10"2 cm/s (73.7 ft/day) and ranged from 3.1 x 10"^ to
6.2 X 10"^ cm/s. These estimated velocities are quite
variable and represent a low flow regime (Autumn) .

14. Field hydraulic conductivity (K) tests were performed on
auger holes and wells in the area of expected metal loading
and tracer movement. These hydraulic conductivity estima-
tions ranged from 1.0 x 10"^ to 1.5 x 10"^ cm/s with a mean
of 8.3 x 10~4 cm/s. The mean water flow velocity determined
from the field hydraulic conductivity values is 7.1 x 10"^
cm/s.

15. The large difference (three orders of magnitude) between
mean hydraulic conductivity determined flow velocity and
tracer determined flow velocity infers the measurement of
two different flow systems. The tracer velocity represents
preferential flow paths (channelling) within the shallow
flow system (acrotelm) and the hydraulic conductivity
determined velocity represents that flow found at greater
depth (catotelm) .

16. The distance the AMD water travels, before problem ameliora-
tion, as estimated by the iron impacted area, is 200m.
Using this 200m distance and the average tracer velocity,
the average residence time of AMD water in contact with the
wetland soils is estimated to be nine days.

1.3.2 Wetland Veaetation/Wildlife

17. The Swamp Gulch wetland was dominated by sedge ( Carex
rostrata) . followed by birch ( Betula alandulosa var hallii ) ,
lodgepole pine ( Pinus contorta) , willow ( Salix boothii ) and
itioss ( Sphagnum tenellum ) with percent canopy coverage of
55.5, 11.3, 6.3, 5.5 and 4.7, respectively. Mesic sites
were dominated by nearly pure stands of Carex rostrata while
the more xeric sites were characterized by Douglas fir
( Pseudotsuqa menzeisii ) , Englemann spruce ( Picea enael-
mannii) and subalpine fir ( Abies lasiocarpa ) and other
species typical of the surrounding upland forest.

18. The composition of plant specieiB in the Swamp Gulch wetland
was similar to that of the Hardscrabble Creek background
wetland .

19. Above-ground production at the Swamp Gulch wetland (3820
kg/ha) was not significantly different than the Hardscrabble
Creek background site (3750 kg/ha).

20. Metal levels were generally elevated above background in the
tissues of the dominant plant species. Levels were usually
higher in below-ground rather than above-ground plant
material .

21. Carex rostrata accumulated elements in the following order:
Cu>Pb>Al>Fe>Cd>Zn. This species apparently possesses an
exclusionary mechanism which prevents the absorption of Mn
and Ni .

22. Of the shrubs species studied, Salix boothii accumulated
elements to a greater concentration than did Betula
glandulosa . Manganese and Ni were readily absorbed by these
species as were Cd, Cu, Fe, Pb, Mn, Ni and Zn.

23. The bryophyte Isopterygium pulchellum had tissue metal
concentrations that exceeded background levels for Al, Cd,
Cu, Fe, Pb, Mn, Ni and Zn.

24. Some metal concentrations exceeded the maximum tolerable
dietary intake levels for domestic animals, suggesting a
potential problem for animals that consume these plants.
The greatest risk was from high levels of Cd.

1.3.3 Wetland Microbiology

25. Removal of acidity in Swamp Gulch natural wetland sediment
occured through the action of anaerobic microorganisms and
was not solely a function of abiotic soil characteristics.

26. Based on laboratory and greenhouse experiments, acidity
remediation occured in the Swamp Gulch natural wetland but
not in the Sand Coulee constructed wetland.

27. Sul fate-reducing bacteria were present at all experimental
sites, and appeared to be reducing sulfate in the Swamp
Gulch sediments, but not the Sand Coulee wetland sediments.

28. Adequate carbon and energy levels to support the growth of a
variety of indigenous microorganisms including sul fate-
reducing bacteria are evident in the Swamp Gulch sediments.

29. In the laboratory experiments, iron removal from AMD was not
evident in either the Swamp Gulch or Sand Coulee wetland
cores .

30. The total microbial biomass and densities of sul fate-reduc-
ing bacteria were low in the Sand Coulee constructed wetland
sediments, probably because of insufficient levels of carbon
and other nutrients.

31. The microbial populations at both Swamp Gulch and Sand
Coulee wetlands were physiologically stressed. This may
account for the inability of the Sand Coulee sediments to
facilitate acid, iron or sulfate removal. Furthermore, it
may be that the Swamp Gulch sediment microbiota are not able
to achieve their maximum physiological capabilities for AMD
amelioration.

32. It should not be automatically assumed that poor performance
in a constructed wetland is due solely to wetland size or
hydrology. There is a significant microbiological contribu-
tion in working wetlands that should be transferable to
non-working wetlands.

1.4 RECOMMENDATIONS

1.4.1 Wetland Hydroaeochemistry

1. These data suggest that any constructed wetland should be
made as large as practically possible. It is doubtful that
any constructed wetland could be built as large as the
impacted area of this natural wetland (3.9 ha).

2. The catotelm (muck zone) was generally ineffective in
removing metals from AMD, but a limited thickness of this
material (probably less than 7 cm) may be needed for a
proper siobstrate for some types of wetland vegetation.

3. The thickness of the natural wetland acrotelm averages 0.6
m. This thickness may be desirable in constructed wetlands
operating under similar circumstances.

4. The specific ion electrode method of bromide tracer analysis
could be improved by periodic double sampling and analysis
with a different method (titration) , especially at low
bromide concentrations.

5. The acquisition of limited additional data from the Swamp
Gulch wetland site would allow a determination of the
system's present efficiency in removing metals from AMD.
This will help determine the effective treatment longevity
of the wetland.

6. Precipitation of iron by physio-chemical processes may be a
very important mechanism for removing iron from AMD (Wieder
and Lang 1986) and likely played an important role in the
Swamp Gulch wetland. Further evaluation of the wetland iron
is recommended to determine if this material is organically
derived or inorganic oxide precipitation.

1.4.2 Wetland Vegetation/Wildlife

7. Carex rostrata . Betula alandulosa and Salix boothii should
be strongly considered for transplantation to constructed
wetlands. These plants absorbed substantial quantities of
Al, Cd, Cu, Fe, Pb, Mn, Ni, and Zn without ill effect. These
plants are common to many parts of Montana and should be
relatively easy to transplant. It is therefore recommended
that a screening study be implemented to determine the
ability of these species to tolerate transplantation and to
proliferate at a constructed wetland.

8. Potential impacts to wildlife from high plant tissue metal
levels should be investigated. The concern has been
expressed that in constructing wetlands for the amelioration
of AMD one is increasing the risk of poisoning the wildlife
resource.

9 . The best indication of whether animals are at risk at the
Swamp Gulch wetland is to analyze tissue samples from those
animals that are most likely to be poisoned. It is therefore
recommended that voles r Microtus spp.) be trapped and their
kidneys be analyzed for metals, especially Cd.

10. Until it can be demonstrated that animals will not be
poisoned by consuming vegetation at constructed AMD
wetlands, it is recommended that these systems be fenced to
limit wildlife access.

1.4.3 Wetland Microbiology

11. Procedures need to be established to transfer microbio-
logical activity of a working wetland to an ineffective
wetland.

12. The ester-linked phospholipid fatty-acid technique is an
important tool for assessing the impacts of AMD on in-
digenous microbiota and in evaluating the efficacy of
constructed wetlands.

13. The addition of carbon and other nutrient sources to
wetlands would diversify the microbial community structure
and be especially beneficial to the sul fate-reducing
bacterial populations.

14. Wetland constiruction practices should provide as much
cation exchange capacity as possible to aid biological
removal acidity and metals. Acidity control through the
application of appropriate amendments should also be
considered.

15. Wetland construction practices should provide favorable
conditions, such as anaerobiosis (e.g., by varying con-
structed wetland depth) , for the proliferation of sulfate-
reducing bacteria. If these bacteria are able to reduce
sulfate to sulfide, metals should be immobilized through the
formation of metal sulfide precipitates.

16. Seasonal effectiveness of wetlands needs to be assessed in
the context of the microbiota and site geochemistries.

17. Emphasis on the exact nature of physiological stresses
imposed on microbial populations in wetlands need to be
elucidated .

18. Vertical and lateral zones of microbial activity need to be
mapped in both natural and constructed wetlands in order to
develop engineering criteria for wetland construction.

2.0 NATURAL SYSTEMS FOR AMD CONTROL

2 . 1 WETLAND TREATMENT SYSTEMS

Research has confirmed the theory that natural wetlands can
improve the quality of acid mine water flowing through them
(Wieder and Lang 1984, 1986). However, since natural wetlands
are seldom located adjacent to acid mine effluent and because of
legal barriers regarding the discharge of pollutants into
wetlands, they are seldom utilized for AMD abatement. Construct-
ing artificial wetlands is the obvious alternative to using
natural wetlands.

Artificial systems, both in the laboratory and in the field,
have demonstrated the ability to improve water quality by
removing certain metal ions (Kleinmann et al, 1983, Burris et al.
1984, Gerber et al. 1985, Wieder et al. 1985 and Brodie 1986).
These systems utilize bryophytes, particularly Sphagnum spp.,
which have generally been successful in removing Fe, Mg, Mn and
other metals from solution. Simola (1977) and Chaney and
Hundemann (1979) studied the ability of Sphagnum to absorb and
tolerate Cd. Bryophyte species of the genera Rhvnchostegium .
Rlytidiadelphus ^ Amblvstegium and Fontinalis have demonstrated
metallophytic properties (Wehr and Whitten 1983, Brown and
Beckett 1985) . Emergent vascular plant species (Typha. Carex,
Eleocharis and Juncus ) and floating plants f Eichhornia . Salvenia
and Lenuma) frequently occur in areas with acidic water and/or
high metal concentrations. Haag and Covert (1987) reported Typha
angustifolia and several grass species growing in AMD water in
Missouri. Dinges (1982) reviews natural and artificial systems
for improvement of polluted waters. Species of Typha are able to
tolerate high metal levels, and therefore have been extensively
studied (Taylor and Crowder 1983a, 1983b, 1984, McNaughton et al.
1974, Mayer and Gorham 1951, Boyd 1970 and Bayly and O'Neill
1972). In terms of biofiltration of AMD, vascular plants have
not been studied as intensively as bryophytes. The ability of
vascular plants to thrive in AMD water indicates that more
research is needed in this area.

Mechanisms thought to operate in pollution control include
cation exchange of metal ions on the surface of organic matter,
physical filtration, chelation or complexation by organic
compounds, dilution, and biological sorption by bacteria, algae,
fungi, animals and plants. Guertin et al. (1985) presents a
review of each of these mechanisms.

Field testing of artificial wetlands are underway at over 20
sites in the eastern United States (Grits and Kleinmann 1986) ,
with the possibility of hundreds of wetlands being constructed in

Appalachia in the next few years. Guertin et al. (1985) discuss
a wetland system installed in Colorado and Heil and Kerins (1988)
report results from an artificial wetland in central Montana.

Primary design criteria for construction of artificial
wetlands are based on estimates of system loading rates, since
these directly effect maximum treatment efficiency and capacity
of the wetland. Influent volxime and chemical composition in
conjunction with treatment rate and required effluent concentra-
tions provide the basis for determining the optimal AMD to
wetland volume ratio, flow path length, and detention time.
Baseline data on flow rate and chemical composition need to
reflect yearly extremes. For a review of construction parameters
and essential baseline information, consult Girts and Kleinmann
(1986), Pesavento (1984) and Brooks (1984). Girts and Kleinmann
(1986) present important criteria used for selecting, planting
and transplanting various plant species.

2.2 BIOSORPTION OF ELEMENTS BY PLANTS

The idea that plant species absorb nutrients from the soil
was first suggested by Glauber in 1656. Prior to this, it was
generally accepted that plants accpaired all their sustenance from
air and water. In 1804, de Saussure demonstrated that plants
acquire a portion of their constituents from the medium in which
they are rooted. Hewitt (1975) reported that higher plants
absorb a portion of all the elements that are present in the
rhizosphere solution, but exhibit absorption selectivity.

Vascular plants characteristic of soils with high metal
concentrations have been known for centuries. In 1588, Thalius
noted that Minuartia verna was an indicator of metal presence in
soil (Ernst 1965) . Metallophytes (plants that accumulate metal
ions) often have tissue concentrations of ions several times
higher than the medium in which they are growing. These types of
plants have been used in geologic prospecting to indicate ore
bodies (Cannon 1960) .

Some bryophytes have also been identified as metallophytes,
accumulating elements well in excess of their structural or
metabolic requirements. The high ion exchange capacity, simple
cellular morphology, and compact low growth form of many mosses
can result in the rapid accumulation of metal ions, either by
extracellular ion exchange or by entrapment of particulates. Lee
et al. (1983) studied the brilliant lime green mosses that
flourished in hillsides blanket bogs fed by spring water enriched
with Zn, Pb and Cu. Bryophytes belonging to the Bryaceae and
Bartramiaceae families were reported useful in identifying Cu and
U deposits (Whitehead and Brooks 1969) .

10

2.3 METALLOPHYTES AND LAND/WATER RECLAMATION

Metallophytes have proven useful in the reclamation of
metalliferous waste sites (Gadgil 1968, Johnson et al. 1977) , and
in the treatment of wastewater (Dinges 1982). The mechanisms of
wastewater treatment by natural systems is discussed in detail by
Dinges (1982) . The use of metallophytes to improve the quality
of acid mine water is being studied extensively throughout the
United States.

2.4 CHEMICAL PROCESSES OF ACID MINE DRAINAGE

2.4.1 Acid Water Formation

Acid water occurs when pyrite (FeS2) is oxidized in air and
water to form sulfuric acid (H2SO4) and iron hydroxide (Fe(0H)3).
Jaynes et al. (1984) and Guertin et al. (1985) explain the
stoichiometry of acid formation in detail. The overall chemical
reaction was summarized by Schumate et al. (1971) as:

FeS2 + 3.75 O2 + 3.5 H2O > H2SO4 + Fe(0H)3

Acidic water is often indicated by the presence of a white to
reddish iron hydroxide precipitate known as 'yellowboy'.

2.4.2 Bacterial Mediation

The kinetics of acid formation are dependent on the
availability of oxygen, the surface area of the pyrite, the pH of
the influent water, and the activity of iron-oxidizing bacteria.
The principal bacteria involved in accelerating pyrite oxidation
are the chemoautotrophic Thiobacillus thiooxidans (Kleinmann and
Crerar 1979). Since these bacteria can increase the rate of this
reaction by a factor of over 100 times, they can play a crucial
role in acid production. Singer and Strum (1970) concluded that
this bacterial ly mediated reaction was the rate determining step
in acid formation. Walsh and Mitchell (1972) believe that the
production of acid can be controlled by breaking the Fe-bacteria
succession in the pH range of 4.0 to 4.5. Bacteria of the genus
Metal loaenuiro have been found to catalyze the acid producing
reaction (Parizik 1985) , while other microorganisms such as Tjl
thiooxidans are important in bringing iron into solution from
soil and organic matter (Oborn and Hem 1962).

2.4.3 Influence of pH

In general, metal ions increase in solubility several orders
of magnitude with each unit decrease in pH. The solubility and
stability of metals in solution is also a function of Eh, or
oxidation-reduction potential . Hem and Cropper (1959) present
data on the relationship between Fe ions, pH and Eh. In terms of

11

AMD treatment systems, Eh-pH diagrams are useful for obtaining
rough estimates of the dominant ions in the effluent (Guertin et
al. 1985)

Traditional methods for treating acidic water are to
oxygenate the water to remove volatile acids such as carbonic
acid. Bases such as sodium hydroxide, limestone or sodium
carbonate are often used as neutralizing agents. Injecting
alkaline materials into acid producing spoil material to prevent
AMD from occurring was reviewed by Ladwig et al. (1985).

2.4.4 Pvrite Oxidation Control

Methods for controlling pyrite oxidation include the use of
bactericides (Barnes and Romberger 1968, Shearer et al. 1970),
the use of bromo-substituted phenols to inhibit metabolism
(Aleems 1972) , the introduction of heterotrophic bacteria to prey
on the iron and sulfur oxidizing bacteria, removal of attachments
for the stalked bacterium Metal loaenium . and increase the pH in
the neighborhood of the bacteria (Walsh and Mitchell 1972). Some
of these techniques have been tested in the laboratory while
others have merely been suggested.

Anionic surfactants, primarily sodium dodecyl sulfate (SDS) ,
alters the cytoplasmic membrane of the bacteria, allowing
hydrogen ions access to the cell. Two organic compounds, sodium
benzoate and potassium sorbate, have been demonstrated as
effective as SDS (Erickson et al. 1985). Pelletized bactericide
has raised the pH of spoil material, allowing it to be revege-
tated (Zaburnov 1987) .

12

3.0 SITE DESCRIPTION

3.1 LOCATION

The Swamp Gulch Wetland study site is located 14 miles (22.4
km) east of the town of Lincoln, Montana in Lewis and Clark
County (section 2 0, TIN, R6W) (Figure 3-1). The site is bordered on
the north by Montana highway 200 and on the south by the
Blackfoot River. The entire wetland covers approximately 72.9 ha
(180 acres) ; however, the study area proper covers about two
thirds of that area of which 3.9 ha (9.6 acres) is apparently
impacted by AMD.

3.2 NATURAL ENVIRONMENT

The wetland study site is in the headwaters of the Blackfoot
River at an elevation of about 1573 m (5160 ft) . Elevations in
the Swamp Gulch drainage which feeds a portion of the study area
range from about 1574 m (5165 ft) to 1908 m (6260 ft) . The Swamp
Gulch watershed contains 72.6 ha (179 acres) and is composed of
generally well vegetated southeast to southwest aspects with a
total length of about 1.6 km (1 mi). The site is on the west
side of the continental divide and is therefore influenced more
by Pacific weather patterns than by the continental climate in
the region east of the divide. However, moisture from the Gulf
of Mexico occasionally spills over the divide during the summer
producing intense thunderstorms. Mean annual precipitation is
approximately 73.7 cm (29 in) (SCS 1981) much of which occurs as
snow (Coffin and Wilke 1971) . Precipitation during the 1987
water year and annual year averaged 60 percent of normal at the
Lincoln Ranger Station. Record (1949-1987) maximum and minimum
annual precipitation at the Lincoln Ranger Station are 91.49 cm
(36.02 in) and 31.14 cm (12.26 in), which occurred in 1975 and
1973, respectively. Average January and July temperatures are -
7.2°C (19.1°F) and 16.4°C (61.6°F), respectively (NOAA 1987).

The Blackfoot River drainage was significantly influenced by
glaciation. Sediments deposited by retreating valley glaciers
created a series a moraines and outwash plains in the Lincoln
area, allowing for the formation of wetlands in drainage impeded
areas on the upslope sides of the moraines (Alt and Hyndman
1972). The lithology of the study area is dominated by the
Spokane Shale Formation, part of the Belt series super group.
This formation consists of red or red-purple shale with numerous
green beds locally with quartzite (Ross et al. 1955). Parent
material of the valley floor apparently consists of a thin layer
of glacial till with fluvial gravel deposits.

13

-P

-H
W

>.

-p

W

Ti
C
(0

rH

4J
0)

u

r— I

O

B

(0

s

CO

iw
o

c
o

•H

-p

(0
D
O
J

I

0)
M

14

Swamp Gulch and many other drainages of the region have been
impacted by mining. The Carbonate Mine is within the Heddleston
mining district and in 1933 included an adit about 850 feet long
and a shaft approximately 300 feet deep (Pardee and Schrader
1933) . A 75 ton mill was erected in the summer of 1944 and it is
reported that 2500 tons of lead-zinc ore was taken out in 1945
(Shea 1947) .

The geology of the Carbonate Mine is similar to that of the
Mike Horse Mine which is the largest mine in the district (Shea
1947). The mine workings are in a diorite sill that intruded
argillite of the Spokane Formation, Belt Precambrian sedimentary
rocks (McClernan 1983) . Intruding into the sill and Belt rocks
is a decomposed rhyolite dike (Shea 1947) . Pardee and Schrader
(1933) describe the lode as consisting of a sheared and altered
diorite partly replaced by sulphides. The mine lode contains
abundant pyrite in several forms including fine grained, coarse
grained and radial. Ore minerals are galena, sphalerite, silver
and gold. Gangue minerals occur primarily as sericite, pyrite
and quartz and also include "a carbonate containing calcium, iron
and a little manganese" (Pardee and Schrader 1933) .

The dominant soil in the study area is a Hydric Borofibrist
(Soil Survey Staff 1975) . The gravels most often encountered at
depth are fluvial in origin as evidenced by many subrounded
pebbles. These were deposited by the meandering river following
glaciation and were likely derived from glacial outwash mate-
rials.

Vegetation within the Swamp Gulch wetland is dominated by
sedge (Carex rostrata) . This species is sometimes associated
with shrubs ( Betula qlandulosa and Salix boothii ) or lodgepole
pine ( Pinus contorta ) , but is most often encountered in nearly
pure stands. Where the water table is relatively low, plant
species more indicative of the surrounding forest are found.

Upland forest vegetation is predominantly subalpine fir
( Abies lasiocarpa ) , Douglas-fir ( Pseudotsuqa menzeisii) and
ponderosa pine ( Pinus ponderosa ) climax forest (Ross and Hunter
1976) . Understory species include pinegrass ( Calamaarost is
rubescens ) , huckleberry f Vaccinium s pp. ) , bluebunch wheatgrass
(Agropyrgn spicatum ) , mountain brome ( Bromus marainatus ) and
Columbia needlegrass ( Stipa columbiana ) .

3.3 NATURAL WETLANDS

3.3.1 Definition and Classification

Marshes, swamps and bogs have been well known terms for
centuries, but only recently have attempts been made to group
these ecological systems under the single term "wetland" . This

15

change has developed out of a need to understand and describe the
characteristics and values of all types of land so that they can
be wisely and effectively managed (Cowardin et al. 1979).

In general, wetlands are lands where the soil or substrate
is at least periodically saturated with or covered with water.
Saturation with water is the dominant factor in soil development
and in determining the types of plants and animals that are in
the soil and on its surface.

The most widely used wetland classification system in the
United States is that of Martin et al. (1953), which was
republished in the U.S. Fish and Wildlife Service Circular 39
(Shaw and Frederick 1956) . Wetland types were based on criteria
such as water depth, water permanence, water chemistry, vegeta-
tion life form and dominant plant species. The system defined by
Golet and Larson (1974) refined freshwater wetland types of
Circular 39 for the glaciated Northeast, but did not recognize
the coastal (tidal) freshwater wetlands. Stewart and Kantrud
(1971) devised a system for classifying wetlands in the glaciated
prairies. This system also relied on water and vegetation
characteristics .

In 1979, a wetland classification scheme was developed as a
tool for managers to inventory wetlands and deep water habitats
of the United States (Cowardin et al. 1979). This system defines
wetlands by plants (hydrophytes) , soils (hydric soils) and
frequency of flooding. The structure of the system is hierar-
chical, progressing from system to class, subclass and dominance
types. Modifiers for water regime, water chemistry and soils are
applied at the class, subclass and dominance levels. Special
modifiers are available for man or beaver modified wetlands.
This scheme was used to classify the Swamp Gulch study area.

3.3.2 Classification of the Swamp Gulch Wetland Study Site

Based on the Cowardin et al. (1979) scheme, the Swamp Gulch
site can be classified as a "Palustrine" system. The Palustrine
system includes all nontidal wetlands dominated by shrubs, trees,
perennial emergents, or emergent mosses or lichens. This
category was developed to group the vegetated wetlands tradition-
ally called marshes, swamps, bogs, fens, prairies or ponds.
Palustrine systems may be situated almost anywhere on the
landscape.

In terms of the subclass, the majority of the Swamp Gulch
site was classified as a "Persistent Emergent Wetland" because of
the dominance by erect, rooted and persistent herbaceous
hydrophytes. This subclass is typified by such plant types as
rush ( Juncus ) , cattails (Typha) , sedge ( Carex ) , bulrush ( Scirpus )
and certain species of true grasses. The portion of the Swamp
Gulch site dominated by shrubs was classified as a "Palustrine

16

Broad-leaved Deciduous Scrub - Shrub Wetland". This subclass is
typified by alder f Alnus ) . willow (Salix) , and birch (Betula) •
The western one-third of the site was classified as "Palustrine
Needle-leaved Evergreen Forested Wetland" . This subclass is
reserved for wetlands having trees taller than 6 meters.

To completely describe the Swamp Gulch site (and any other
wetland) certain modifiers at the class level need to be applied.
These modifiers are for water regime, water chemistry (Eh and
pH) , soil factors, and special modifiers for disturbed wetlands.
For simplicity, we will limit our discussion of special modifiers
to the dominant subclass present at the Swamp Gulch site, the
'Palustrine Persistent Emergent Wetland'. The water regime for
this subclass was 'nontidal, semipermanently or seasonably
flooded ' . These descriptors cover the range from saturated
surface soil to areas consisting of open water during the growing
season. The salinity modifier was 'fresh' (i.e. less than 800
umhos/cm) , the pH modifier was 'acid' (i.e. less than 5.5), and
the soil modifier was 'organic'. The special modifier was
'impounded', due to the activities of beaver.

The full wetland classification for the majority of the
Swamp Gulch site according to Cowardin et al. (1979) is,
"Palustrine, Persistent, Emergent Wetland, nontidal, semiper-
manent or seasonably flooded, fresh, acid, organic, impounded".

17

4.0 SURFACE HYDROLOGY OF A NATURAL WETLAND

The Swamp Gulch surface wetland water hydrology was
evaluated: 1) to determine the amount of AMD inflow and hence the
calculated metal loading input to the wetland (see Section 7.0);
2) to evaluate the general surface flow pattern to determine the
area of impact and if channeling was evident; and 3) to evaluate
the effect of AMD dilution by other unpolluted water sources
flowing into the wetland. The following sections present these
aspects of the study.

4 . 1 METHODS

The measurement of the AMD into the wetland was determined
using a 22.9 cm (0.75 ft) throat Parshall type flume equipped
with a Stevens Type F recorder. This unit was installed June 15,
1987 in the upstream toe of the tailings dam on Swamp Gulch. The
area beneath the flume was saturated and a mixture of tailings
and bentonite was used to seal the area. The flume discharged
directly into a wooden box culvert through the dam. The box
culvert had sufficient cross-sectional area and slope so that no
submersion of the flume outlet was observed. Complete records
were obtained from June 15 through November 17, 1987, at which
time the unit froze. The rate of flow was calculated using the
equation:

Q = 4WHa 1«522 yi 0.026 (Parshall 1950)

where: Q = discharge, cubic feet per second (cfs)
W = throat width (feet)
Ha = upper gage head (feet)

The rate calculations were integrated over time intervals varying
from 0.25 to 24 hours as determined from the recorder charts for
total daily flow.

Precipitation was measured with a Belfort #5-780-6 universal
recording rain gage installed near piezometer site E-1 on June 2-
3, 1987 (Figure 4-1). The gage was mounted on 3.05 m (10 ft)
long by 1.9 cm (0.75 in) inside diameter flanged pipes driven
into the wetland approximately 2 m (7 ft) . The installation is
equipped with an Alter (1937) type wind screen mounted at the
orifice level, which is approximately 2.1m (7 ft) above the
wetland surface. The gage was calibrated in 2.54 cm (1 in)
increments from to 15.24 cm (0-6 in) following installation.
Records are complete to date (Appendix B-2) , but some snowfall
may have been missed from approximately December 15, 1987 through
January 11, 1988, due to snow bridging.

18

CARBONATE MINEX-'^ //

f% // II

A WELL SITE

O SURFACE WATER SAMPLE

SITE
D SURVEY CONTROL STATION
M FLUME
O PRECIPITATION GAGE

SEC20, TI5N,R6W
LEWIS AND CLARK CO .MONTANA

FEET
200

50
METERS

Figure 4-1. Swamp Gulch Wetland Study site (with water sampling
stations) .

4.2 RESULTS AND DISCUSSION

Inflow Of surface water to the wetland site is from two
general sources: Swamp Gulch and the Blackfoot River. The
Blackfoot River water is dispersed by several beaver dam systems
below the confluence with Pass Creek (Figure 4-1) . Direct
drainage from these dams flows in moderately well-defined
channels southwestward past site AAA-3 into the study area. The
Blackfoot River water drains into the beaver pond between sites
C-3 and c-4 and then, via a well defined channel, flows past site
C-3 to the pond directly east of site D-2. The flow at site C-3
may approach 1 cfs at times and may average better than 0.25 cfs.
Outflow from the pond east of site D-2 drains westward to a small
pond adjacent to site E-2 and then through two ponds north of

19

site F-2. The pond immediately northeast of F-2 receives
additional Blackf oot River water diverted via a channel near
site E-3. Drainage from the pond northwest of F-2 flows
approximately 37 m (120 ft) in a well defined channel into the
Blackf oot River. This system likely determines the current
southern extent of the Swamp Gulch acid drainage impact area.
The pond northeast of site F-2 is also the likely point of
accumulation of wetland drainage resulting from Swamp Gulch. The
area topography generally confines this flow to the area
immediately south of Highway 200 (Figure 4-2) .

It should be noted that fish. Eastern Brook Trout, have been
occasionally observed in this system from site AAA-3 to the pond
near site E-2. These were observed frequently in the channel

CARBONATE MINE//

/\ //

60('

/\ WLLl SITE

O SURFACE WATER SAMPLE

SITE
O SURVEY CONTROL STATION

jm Flume

O PRECIPITATION GAGE

SEC 20, TI5N,R6W

LEWIS AND CLARK CO .MONTANA

FEET
200

50 100

METERS

Figure 4-2. Topographic map with 0.5 m contours.

20

from AAA-3 to the pond between C-3 and C-4, and only once at the
pond near E-2 (during the April, 1987 sampling trip).

The area surrounding the Swamp Gulch AMD discharge point is
relatively uniform with no distinct channeling. The discharge is
spread in a wide arc of roughly 180 degrees.

The age of the present wetland drainage system is important
m relation to mining and post-mining periods in determining the
actual extent of the area potentially impacted from this source
and if there have been any changes in the impacted area. A
comparison of aerial photography from the years 1939, 1966, 1972
and 1987 indicates that the present beaver dam system was in
place before July 15, 1939, and that apparently mature Pinus
contorta were established at that time in the same beaver dam
areas where they are presently found. Several dead specimens
were obtained and tree ring counts indicated ages from 22 to 70
years (Table 4-1) . The oldest specimen exhibited intact bark and
some needles suggesting it has been dead less than five years.
Other samples had deeply checked surfaces with no bark with tree
ring counts ranging from 22-63, which suggests these are some of
the dead trees evident in the 1972 photographs. Based on these
data, the minimum age for dam systems would be 78 years, having
been established sometime prior to 1909. It is likely that the
lodgepole pine could not have become well established on these
dam areas until unsaturated surface soil conditions developed,
possibly following abandonment. It is, therefore, probable that
the system has remained essentially unchanged since &t least 1900

Table 4-1. Beaver dam tree ring counts.

Sample Ring

Number Counts Condition Location

1 70 Bark intact, some needles, 30 ft SW C-3

12 inch diameter.

2 22 No bark, deeply checked, 25 ft SW C-3

3.5 inch diameter.

3 47 No bark, deeply checked, 30 ft S D-2

5.5 inch diameter.

4 27 No bark, deeply checked, 25 ft W D-2

7.5 inch diameter.

5 63 No bark, deeply checked, 15 ft N C-4

8.5 inch diameter.

21

and the mining impact area has not changed during this interval.
The 1939 aerial photographs also indicate that no tailings dam
was present adjacent to the highway at that time, which supports
the apparent 1944 construction date for this feature.

Flume records indicate flow rates varied from a minimum of
0.1 L/s (0.004 cfs) on November 15, 1987, to a maximum of 14 L/s
(0.5 cfs). The peak flow occurred July 17-18, 1987, in response
to 6.6 cm (2.60 in) of precipitation received at the site during
the two day period. Daily minimum and maximum flows were 12. 3m-^
(432 ft-*) and 734m3 (25,840 ft^), respectively. The average
daily flow was 45m3 (1,588 ft^) . A comparison of flume hydro-
graphs and precipitation gage charts suggests the time of
concentration (the time interval from initial storm precipitation
to the storm hydrograph peak) at the fltime is about ten hours
under conditions of low antecedent precipitation and moderately
large storms. The maximum intensity precipitation rate was
recorded on July 22, 1987, when 1.4 cm (0.55 in) fell within one
hour. The maximum daily total (24 hours) was 5.72 cm (2.25 in)
which occurred July 17, 1987 (Appendix B-1) . This event was
close to the reported ten year, 24 hour storm event of 5.6 to 6.1
cm (2.2 to 2.4 in) (Miller et al. 1973). It is apparent that
even the moderately intense storms do not produce peak runoff
sufficient to cause any channeling in the wetland. Extreme
events would be attenuated by the capacity of the highway culvert
and highway overtopping would likely spread the flow to reduce
wetland channeling. There is no evidence that this type of event
has occurred in the last few years. Normal peak flow would be
expected during spring snow melt and no data has yet been
recorded for this time interval.

Precipitation records from the site (Appendix B-2) suggest
warm season precipitation is, in general, similar to that
observed at the Lincoln Ranger Station with the exception of
isolated storms such as the July 1987 event. The limited record
is insufficient to determine if these isolated storms have a
significant effect on the annual long term precipitation. The
November 1987 through February 1988 records indicate consistently
more snowfall occurs at the wetland than occurs at the Lincoln
Ranger Station. This would be expected due to the higher
elevation of the wetland. The insufficient record length at the
present time prevents an accurate estimate of average annual
precipitation, but it is apparent that the site does receive
amounts in excess of the 48.1 cm (18.93 in) received at the
Lincoln Ranger Station. A linear regression of monthly data from
the two sites for the June 1987 through February 1988 records
gives the following relationship.

WP (cm) = (1.45) LRSP - .47 cm

where: WP = wetland precipitation

LRSP = Lincoln Ranger Station precipitation

r2 =0.69 (r=0.83)

22

Using the regression equation with the average annual precipi-
tation at the Lincoln Ranger Station gives a 69.3 cm (27.27 in)
estimate of the annual precipitation at the wetland. This is
somewhat less than the 73.66 (29 in) estimated from SCS (1981)
isolines. A rough free water surface evaporation estimate as
determined from Kohler et al. (1959) is approximately 66 cm (26
in) .

In summary, the Swamp Gulch AMD daily average flow is
estimated to be 45m-» with a range of 12.3 to 734m3. The Swamp
Gulch discharge into the wetland is generally well distributed
without any notable localized channels, although there are
preferred flow paths (see Section 5) . Blackfoot River and Pass
Creek surface water is introduced into the wetland via beaver
dams and two well-defined channels which have apparently been in
place during the complete history of the mine. The channels have
likely limited the extent of the AMD impact area and provide a
sufficient unpolluted water volume to dilute the measured AMD
volume to low levels. The Swamp Gulch wetland receives precipi-
tation in excess of that known for the Lincoln Ranger Station.
The precipitation is likely also slightly in excess of the
estimated free water surface evaporation at the site and hence,
no concentration of dissolved solids in wetland water likely
occurs from evaporation.

23

5.0 HYDR06E0IOGY OF A NATURAL WETLAND

The purpose of this part of the Swamp Gulch wetland study
was to determine if residence time of mine-contaminated water in
a natural wetland could be determined using a bromide tracer.
Specific objectives of the study were as follows:

1) Investigate the ability of bromide to serve as a tracer
in the wetland.

2) Identify water flow velocity in the wetland.

3) Relate water flow velocity to wetland physical
properties (hydraulic conductivity) .

4) Determine residence time in the wetland area affected
by AMD.

5.1 LITERATURE REVIEW

5.1.1 Hydraulic Principles of Tracer Use

In hydrology, a tracer is either matter or energy carried by
water which will give information concerning the direction and/or
velocity of the water. When sufficient data are collected,
tracers can assist in the determination of hydraulic conduc-
tivity, porosity, dispersivity and other hydrogeologic parameters
(Davis et al. 1980). Introducing a tracer at one point in the
flow field and observing its arrival at other points, is the most
direct method of determining groundwater velocity (Freeze and
Cherry, 1979) .

One factor which controls velocity is hydraulic conductivity
which represents the ease with which water moves through the soil
or aquifer. Hydraulic conductivity, "K", is often used to
characterize an aquifer because it includes the properties of the
fluid and the field of gravity as well as the properties of the
porous medium such as permeability. Hydraulic conductivity has
the dimensions of length/time (L/T) or velocity (Fetter 1980).
Hydraulic conductivity, porosity and permeability values can vary
widely in space and time. The relationship between velocity and
hydraulic conductivity is derived from Darcy's Law and is shown
in the following equation:

V = (-K/ne) (dh/dl) (Eq. 5-1)

Where: V = average linear velocity
K = hydraulic conductivity
ne = average effective porosity

24

dh/dl = average hydraulic gradient
dh = change in head
dl = change in length

Interpretation of the results of tracer tests involves
plotting the concentration of a tracer as a function of time or
volume of water passing through the aquifer. In the resulting
"breakthrough curve" the concentration is commonly given as a
ratio of the measured concentration at the observation well, "C",
to the initial tracer concentration injected, "Co". The average
travel time of a non-reactive (conservative) tracer can be
determined from a breakthrough curve for transport from the
injection point to the observation point. The first arrival time
of a tracer as it moves through the system represents the maximum
velocity of the groundwater. The peak concentration of the
tracer represents the average transit time of groundwater through
the system, if a conservative tracer is used. Retardation of the
transit time of a tracer is related to the breath of the
breakthrough curve. Figure 5-1 gives hypothetical examples of
breakthrough curves for a mixture of tracers injected as a single
slug into an aquifer. Notice the change in the curve shape when
the tracer is not conservative and interacts with the aquifer
system.

Breakthrough curves of four different tracer
types: (a) is conservative, (b) some effect of
sorption, (c) large effect of sorption, and (d)
precipitated or destroyed (From Davis et al,
1985) .

25

5.1.2 Tracer Flow Retardation and Tracer Dilution

Tracers are not perfectly conservative and the concentration
distribution of a water soluble substance which is transported in
a porous medium by groundwater is affected by sorption, molecular
diffusion and hydrodynamic dispersion (Gustafsson and Klockars
1981) . Sorption includes adsorption and absorption processes.
The many chemical processes which contribute to sorption result
in retardation of tracer movement. Thus, velocity of the tracer
is slower than that of

the groundwater. Therefore in order to design a meaningful
tracer experiment, the sorptive characteristics of the tracer
must be known (Davis et al. 1985). Davis et al (1985) and
Gustafsson and Klockars (1981) give equations to describe the
effect of tracer sorption in relation to groundwater flow.

Hydrodynamic dispersipn and molecular diffusion have the
effect of diluting the concentrations of artificially induced
tracers. Hydrodynamic dispersion generally affects short term
tracer tests and molecular diffusion affects the concentration of
slow moving tracers in heterogeneous materials (Davis et al.
1985). Freeze and Cherry (1979) also state that molecular
diffusion is important only at low flow velocities. Figure 5-2
depicts movement by molecular diffusion. Note that no water
movement is required for the dye to spread out (diffuse) in a
direction tending to equalize concentrations in the blotter.

spot of dye

soaked blotter-

(no water movement)

initial conditions

three hours

one hour

one day

Figure 5-2.

One dimensional example of movement by molecular
diffusion (from Davis et al. 1985).

26

Hydrodynamic dispersion is the spreading which occurs both
perpendicular to and in the direction of groundwater flow, of a
water-soluble substance that is transported with the groundwater.
The transit velocity of specific water-soluble substances may be
higher or lower than the average groundwater velocity. Hydrody-
namic dispersion is dependent on the velocity distribution in the
medium and on molecular diffusion. Bear (1972) gives a quite
detailed evaluation of hydrodynamic dispersion. Dispersion of a
solute requires groundwater flow in a medium with a system of
pores or channels. An example of hydrodynamic dispersion which
is caused by unequal velocities of the ground water is shown in
Figure 5-3. Tracer particles released at the same time and
carried by the groundwater have different flow paths. This
results in a more widespread distribution of tracer particles
with time.

General direction of water motion

W Initial position

^ Position after one. hour

[(£))) Position -ofter two hours

INITIAL
-DISTRIBUTION
OF PARTICLES

OISTRISUTION OF
.PARTICLES AFTER
/ONE HOUR

DISTRIBUTION OF
PARTICLES AFTER
TWO HOURS

DISTANCE

Figure 5-3.

One dimensional example of hydrodynamic dispersion
for tracer particles A - G (from Davis et al.
1985) .

27

5.1.3 Tracer Methods

The initial step in conducting a tracer study is to collect
as much hydrologic information as possible about the study site
(Davis et al. 1985). This infonnation should include homogeneity
of the aquifer, layers present, fracture patterns, porosity, flow
system boundaries, hydraulic gradient and hydraulic conductivity
(Davis et al. 1985). This hydrologic information is used to
assess the groundwater flow direction and velocity. Direction
and velocity are usually estimated by the use of Darcy's law or
by performing a preliminary tracer test (Davis et al. 1985). The
second step is to determine the best tracer to use for the
conditions and objectives at the site. The third step in
conducting a tracer study is to determine the correct amount of
tracer to be used. This amount is based on dilution expected,
natural background concentration and detection limit possible
for the tracer. The next step is to determine the correct tracer
method.

One method involves the use of an environmental tracer which
is a substance that exists in the soil before the investigation
begins. This tracer can be artificial (man induced), semi-
artificial (tritium) or natural (natural radioisotope) . A good
environmental tracer must be free from chemical reactions, such
as ion exchange and precipitation and must not react with the
medium (Fried 1975) . Initial and boundary conditions must be
known, chiefly by knowing the amount of tracer added and its
history. The wetland system under investigation has the
characteristic of a high organic matter content (large exchange
capacity) . Trace metals (environmental tracers) commonly react
with the organic material, making them unacceptable in this
situation.

Freeze and Cherry (1979) describe four main types of field
dispersivity (tracer) tests. These are (1) single-well
withdrawal -inject ion tests, (2) natural-gradient tracer tests,
(3) two-well recirculating withdrawal -inject ion tests and (4)
two-well pulse tests. In a natural gradient tracer study, the
direction and velocity of the groundwater flow are very important
(Davis et al. 1985). In the natural -gradient test, the tracer is
introduced into the system and its migration is then monitored at
one or more sampling points. Dispersivity values are obtained by
fitting an analytical or numerical model to the experimental
data. Davis et al. (1985) notes that:

"It is not at all uncommon to inject a tracer in a well and
not be able to intercept that tracer in a well just a few
meters away, particularly if the tracer flows under the
natural hydraulic gradient which is not disturbed by
pumping."

28

A tracer may be injected as a slug or as a continuous source
input. Mixing of the tracer during injection is important in
most types of tests and can be as simple as pouring it in the
water to be studied (Davis et al. 1985). For shallow wells a
plunger can be surged up and down in the hole or the tracer can
be released through a pipe with many perforations . The ideal
condition is to inject the tracer into the water instantaneously
as a slug.

Taking field measurements of electrical conductance within,
ahead of, and behind a tracer slug can minimize the number of
field samples kept for laboratory analysis (Lee et al. 1980).
Electrical conductance is the simplest and most inexpensive
detection and analysis technique for ionic tracers and can be
used as a break-through indicator (Slichter 1902) . In the
wetland system being studied, the naturally high ionic content
may make electrical conductance data difficult to interpret.

5.1.4 Tracer Problems

Many water tracing attempts are unsuccessful. Aley and
Fletcher (1978) have found that the major causes of failure are:
(a) insufficient hydrological field work before the tracer is
injected, (b) tracing attempts during low flow conditions, and
(c) failure to allocate sufficient time for tracing effort.
Davis et al. (1980) found that tracer test failures are most
commonly a result of incorrect choice of tracers, insufficient
concentrations of tracers and a lack of understanding of the
hydrogeologic system being tested.

For a given head drop, expected travel time is a function of
the distance squared, and therefore increases very rapidly with
the distance. Davis et al. (1985) note that this relationship
causes one of the most common errors in tracer tests, which is to
conduct tests between points which are separated by too great a
distance. Freeze and Cherry (1979) describe four main disad-
vantages to the determination of groundwater velocity by the
direct tracer method.

1) Undesirably long periods of time, because of the fact
that groundwater velocities are usually low, are
normally required for tracers to move significant
distances through flow systems.

2) Numerous observation points are usually required to
adequately monitor passage of the tracer through the
study area because geological materials are typically
quite heterogeneous.

3) Because of (1), only a small and possibly nonrepresent-
ative sample of the flow field is tested.

29

4) Because of (2), the flow field may be significantly
distorted by the measuring devices.

The concentration of ion to be injected should be well above
the natural background concentration level found at the test site
and high enough to ensure detectable levels in observation wells
(Davis et al. 1985). The ion concentration injected must also be
kept low enough so that density effects do not effect flow of the
tracer. An insufficient quantity of tracer will result in an
unsuccessful trace: too much tracer wastes materials and can
degrade water quality. Dilutions of a tracer in transit from
injection to sampling wells are almost always at least tenfold
for "slug" injections and dilutions of ten thousandfold are
common (Davis et al. 1980) .

Lenda and Zuber (1970) describe a method to estimate the
adequate amount of tracer and carrier to be injected. The
effects of adsorption and the uncertainty from whether the
observation well is exactly at the center of the tracer path
should be accounted for by the use of a safety factor on the
order of 10. Skibitzke and Robinson (1963) used tracers to show
that solid particles (sand grains) retard diffusion in a porous
medium. Biggar and Nielson (1962) concluded that the mere
presence of a tracer downstream from the point of injection is a
poor indicator of the velocity of the fluid. Pore geometry,
water content changes and the magnitude of the interaction
between the tracer and the porous medium are important in
determining an accurate estimate of the fluid velocity.

5.1.5 Bromide as a Tracer

There is no such thing as the perfect tracer but Davis et
al. (1980) note that the ideal ground-water tracer is nontoxic,
inexpensive, moves with the water, is easy to detect in trace
amounts, does not alter the natural direction of the flow of
water, is chemically stable for a desired length of time, is not
present in large amounts in the water being studied and is
neither filtered nor sorbed by the solid medium through which the
water moves. Different types of tracers include: water tempera-
ture, solid particles, ionized substances (Br-) , stable isotopes,
radioactive tracers, organic dyes, gases, and f luorocarbons .
Davis et al. (1980) report that some of the most useful general
tracers are bromide, chloride, rhodamine WT, and various f luoro-
carbons. Most tracers have relatively limited or specialized
uses.

Tracer selection should be based on purpose of the study,
type of aquifer system, aquifer characteristics, natural
background concentration of the ions in the groundwater, and
analytical techniques available (Davis et al. 1985) . In most

30

cases, anions are not affected by the aquifer medium (Davis et al
1985) but the characteristics of some aquifers will cause
retention or exclusion of anions moving through the system.

Anionic tracers such as bromide (Br"") and chloride (Cl~) are
particularly useful because of their low susceptibility to
adsorption or ion exchange processes of natural aquifer material-
s. Bromide does not appear to be lost by precipitation,
adsorption or absorption and is biologically stable (Schmotzer et
al. 1973) and therefore can be considered a conservative tracer.
Bromide offers one of the best possibilities as a general tracer
for groundwater studies (Davis et al. 1980). Most bromide
compounds also have relatively low toxicities. Davis et al.
(1980) note that bromide samples, being nonvolatile can be stored
indefinitely without concern for tracer loss to the atmosphere
and sampling can be done using inexpensive air-lift pumps.

Davis et al. (1985) relate that bromide is perhaps the most
commonly used ion tracer. These authors list advantageous
characteristics of the use of bromide as a tracer as inexpensive,
stable, low limit of detection, low background concentrations,
low toxicity and no sorption. Bromide as a tracer is commonly
injected as NaBr, CaBr2/ or KBr.

The concentration of bromide in natural ground waters is
roughly 1/300 that of chloride and usually <1 mg/L (Davis et al.
1980; Vinogradov 1959). Detection of bromide is relatively
simple with a specific ion electrode which has a lower limit of
detection of about 0.4 mg/L. If natural water has 30 mg/L of
chloride (suggesting the natural presence of 0.1 mg/L of Br~) and
if the bromide tracer is introduced with a concentration of 1000
mg/L, then a dilution factor of 10^ is possible before it is
masked by the natural background (Davis et al. 1980). Concentra-
tions in our introduced chloride tracer should not exceed about
3000 mg/L because of increased density of the solution (Davis et
al 1980) . Since the halides bromide and chloride behave
similarly (Bohn et al. 1985) it is thought that bromide should
not exceed the 3000 mg/L concentration.

Davis et al. (1985) state that an advantage of anions, such
as bromide, used as tracers is that they do not decompose and are
not lost from the system. This statement is generally true but
under certain circumstances anions such as bromide may be
affected by anion exclusion and/ or anion exchange. As anions
move through soil they do not come in contact with all of the
soil water. This is termed anion exclusion and occurs in
response to the fluid flow rate and the fact that water near
negatively charged soil surfaces is relatively immobile. The
result of this exclusion is that anions can move through the soil
faster than one would predict on the basis of uniform association
with all the soil water (Smith and Davis 1974) . The association
between anion exclusion and cation exchange capacity is strong.

31

Anion exclusion is a manifestation of the unequal ion distribu-
tion in the diffuse double layer surrounding charged colloid
surfaces (Bohn et al. 1985) . Factors affecting anion repulsion
(exclusion) include: 1) anion charge and concentration, 2)
species of exchangeable cation, 3) pH, 4) presence of other
anions, and 5) nature and charge of the colloid surface (Bohn et
al. 1985). Thomas and Swaboda (1970) suggest that in soils with
high cation exchange capacities anion exclusion causes anions to
move much faster than they would if no interaction with clays
were to occur. This suggestion agrees with the theory that anion
exclusion is a function of negative charge.

Grim (1968) mentions two types of anion exchange in clay
minerals. One is the replacement of hydroxide ions and the other
factor is related to geometry of the anion. The geometry of the
bromide anion does not fit that of silica tetrahedral sheets,
thus it cannot be so absorbed. Anion exchange would take place
around the edges of the clay minerals, not on the basal plane
surface. A factor which complicates anion exchange studies is
that any free or exchangeable iron, aluminum or alkaline earth
elements present in the clay may form insoluble salts with the
anions (Grim 1968) . Mattson (1929) has shown that the adsorption
of anions was found to be negative in neutral and alkaline
solutions and that as pH decreases the capacity of clay minerals
for holding anions increases. The interaction of cations with
clays is much more frequent than with anions. Berg and Thomas
(1959) found that sulfate and chloride anions are adsorbed in
soils high in kaolin clays and aluminum and iron oxides.
Chloride will desorb readily at pH values found in most field
conditions, but at low pH values, chloride ions were not easily
desorbed. They also found that sulfate is held much more
tightly to these soil types than is chloride.

The study site is in the headwaters area of the Blackfoot
River and bromide toxicity was a concern. Alexander et al.
(1981) studied the effect of sodium bromide on Fathead minnows.
Sodium bromide has a low toxicity to Fathead minnows with the
average LC50 (lethal concentration for 50% of the test popula-
tion) for a 96 hour period being 16,479 mg/L. Tests by Barnes et
al. (1981) show that the mean number of organisms (thirty-five
species of algae and zooplankton were studied) increased with
time due to the addition of 1000 ppb potassium bromide and
nutrients to the system. Species diversity decreased slightly in
this situation. Schmotzer et al. (1973) reported that bromide
has a very low toxicity in humans at 50-100 mg of bromide/100 ml
of blood. This translates to a human having to drink 12 liters
of 200 mg/L bromide to be toxic. The oral LD50 (lethal dose)
in rats is 3.5 g/kg (Merck Index 1983).

32

Martin et al. (1956) concluded that plant tolerance to
bromide is quite variable. Carrot tops contained 2.5% (25,000
ppm) bromide with no reduction in growth but citrus trees
containing 0.2% (2000 ppm) demonstrated reduced plant growth.

The desirable characteristics of using bromide as a soil
water tracer were demonstrated by Onken et al. (1977) to be easy
detection, unlikely contamination of the environment and lack of
reaction with soil and soil constituents. Schmotzer et al.
(1973) conducted a fairly extensive study of using bromide as a
groundwater tracer and found bromide to successfully fulfill the
requirements devised by Schmotzer to be as close as possible to
being the ideal tracer. The favorable characteristics of bromide
include low toxicity, high sensitivity of detection, little loss
through precipitation, adsorption and absorption, high stability,
low background concentrations, small sample size requirements
(post sampling activation analysis) , low cost, government
approval is relatively easily attained and bromide is biologi-
cally stable.

Vinogradov (1959) found that bromide correlated well with
iodide content and both bromide and iodide content was propor-
tional to the amount of organic material in the soils studied.
The concentration of bromide is greater in humic soils and there
IS practically no dependence of chloride content on soil organic
matter content. The studies by Vinogradov (1959) suggest that
bromide is sorbed by peats and that the amount sorbed decreases
somewhat with aging of the peats. Vinogradov (1959) found that
as soil organic carbon content increased the bromide and iodide
content increased, but no effect was observed on the chloride
content .

Smith and Davis (1974) found that bromide is a good tracer
for mimicking the movement of nitrate (NO3) through subsoils.
Merrill et al. (1985) used bromide as KBr to trace NO3-N movement
and indicate water flux in a study to develop an understanding of
plant growth response to soil thickness over sodic minespoils.
Onken et al. (1977) used sodium bromide and sodium nitrate to
show that nitrate and bromide move together in the soil profile.
They note that both nitrate and bromide are readily absorbed by
plants but the rates of removal from soil are different.

Tennyson and Settergren (1980) used a sodium bromide tracer
to evaluate percolating water and ion movement in an irrigation
saturated surface soil. Background levels of bromide in soil
water, groundwater, and precipitation were measured and bromide
movement was quantified by soil water sampling and post-sampling
neutron activation analysis. They suggested that laboratory
measured hydraulic conductivity was not adequate in evaluation of
the site because the bromide tracer moved through the soil much
more rapidly than the hydraulic conductivity suggested. Peak
concentrations of the bromide tracer moved through 0.9 m of these

33

soils within 3.75 hours after field application. Tennyson and
Settergren (1980) found indication of bromide retention occurring
because bromide concentrations above background levels were
present in the soils studied three weeks after application of the
tracer.

5.1.6 Wetland Hvdroloav

Wetland systems can be considered quite heterogeneous.
Figure 5-4 is shown to exemplify this point. Notice that not
only is there seven orders of magnitude of hydraulic conductivity
variation reported in the literature shown but there is often
quite a large range of conductivity within a study site. Freeze
and Cherry (1979) and Davis et al. (1985) note that a problem

•z

o
o

111 1
le II 13 13

IS

IT

REFERENCE

Figure 5-4.

Ranges of published field data on K (hydraulic
conductivity) of peat: 1. Baden and Eggelsman
(1961, 1963, 1964); 2. Eggelsman and Makela
(1964); 3. Boelter (1965); 4. Ingram (1967); 5.
Galvin and Hanranahan (1968); 6. Romanov (1968)
7. Sturger (1968); 8. Dowling (1969) ; 9. Irwin
(1970); 10. Yamamoto (1970); 11. Knight et al.
(1971); 14. Paivanen (1973); 15. Galvin (1976);
16. Dasberg and Neuman (1977) ; 17. Chason and
Siegel (1986) . Adapted from Chason and Siegel
(1986).

34

arises in determining the direction of water movement and travel
time when local differences in hydraulic conductivity amount to
several orders of magnitude. The following paragraphs describe
some of the findings of specific authors in regard to wetland
characteristics.

Hydrogeologic factors (Winter and Carr 1980) that affect
wetland ground water flow include:

1) geometry of the geologic framework through which ground
water flows, including flow boundaries;

2) hydraulic conductivity of the geologic materials,
including anisotropy — the ratio of vertical to
horizontal hydraulic conductivity;

3) recharge and discharge of the ground water system.

In particular. Winter and Carr (1980) found three major inter-
relationships between wetlands and ground water:

1) some wetlands appear to recharge ground water;

2) some wetlands are flow-through types where ground water
enters one side and surface water seeps into the ground
on the other side;

3) some wetlands are discharge areas for ground water.

These interrelationships are further complicated by changing
throughout the year.

One of the more important physical properties that affect
the hydrologic features of bog areas is the hydrologic conduc-
tivity of the peat soil horizons (Boelter 1965) . Dai and
Sparling (1973) suggested using the piezometer method to
determine the velocity of flow from the hydraulic conductivity.
Ingram et al. (1974) used field experiments to find that
estimates of hydraulic conductivity in humified peat increased
with head, which is not in accordance with Darcy's law. It is
suggested that Darcy's law may only apply in peat of low
humidification. Boelter (1965) used field measurements of
hydraulic conductivity (K) and found they covered a wide range of
values. Chason and Siegel (1986) found that horizontal K was
significantly greater than vertical K.

Sturger (1968) studied the hydrologic properties of a
mountain bog in Wyoming. Undecomposed surface peat had the
lowest bulk density, 0.160 g/cc, and much of its volume consisted

35

of large voids which emptied at suctions of less than 0.10 bar.
Greater suctions were required to drain the smaller pores of the
higher bulk density, decomposed material found at depth.

In 1969 Boelter found that the classification of peat
materials based on degree of decomposition as measured by fiber
content (> 0.1 mm) and bulk density can relate significant
infoirmation about the hydraulic conductivity, water retention,
and water yield coefficient. Regression equations were employed
to determine a range in hydraulic conductivity and other physical
properties for fibric, hemic and sapric peat materials (Table 5-
1) . In studying a Minnesota spring fen-raised bog complex Chason
and Siegel (1986) found that hydraulic conductivity, bulk density
and humicity all vary widely through the peat profile and in most
cases were not mutually dependent. There are two major hydro-
logic zones within a peat soil column (Ingram 1983):

1) the acrotelm, a thin upper aerated zone of fluctuating
water conditions composed of undecomposed dead and live
vegetation,

2) The catotelm, an underlying anaerobic zone in a constant
waterlogged state composed of more humified peat.

Some authors (Dasberg and Neuman 1977; Ingram 1967; Sturger
1968) note erosion channels or fractures at depths within wetland
systems. The results of field and laboratory studies show that
the properties of peat change drastically when it becomes
partially desaturated (Dasberg and Neuman, 1977). Two layers
result from this change, a permanently saturated layer below the
zone of water table fluctuation, and an overlying partially
unsaturated layer with properties that vary with depth. The
saturated peat layer studied had a high porosity (90%) ; organic
matter content (60% by weight) ; specific yield (20 to 30%) ; and
a very high specific storage (0.7 x lO^^, to 1.7 x 10"^). The
unsaturated peat layer was found to have a high bulk density, up
to 0.8 g/cc near the surface. The hydraulic conductivity of the
unsaturated layer is generally higher than 1.0 mm/h (3.0 x 10"^
cm/s) , but much more variable than the saturated peat layer. In
summer, shallow cracks often develop which may increase the
hydraulic conductivity by several orders of magnitude in the
unsaturated surface layer.

Ingram (1967) described a "water track" as a mire surface
feature having a higher rate of water movement due to more
steeply inclined water tables and/ or peat with higher hydraulic
conductivity. These water tracks are associated with the
presence of more eutrophic plant communities and are sites of
greater ion supply. Erosion channels occur both on the surface
and at depth in peat deposits with the associated vegetation
often being unlike the surrounding vegetation. Subterranean
erosion tunnels are frequently encountered in blanket bogs but

36

Table 5-1. Range of important physical characteristics of
fibric, hemic, and sapric peat materials from
northern Minnesota bogs (From Boelter 1969) .

Bulk Total 0.1 bar Hydraulic

Organic density porosity H2O content conductivity

Material fa/cc) C%J f%^ fi o'^cm/sec)

Fibric <.075 >90 <48 >180

Hemic .075-. 195 85-90 48-70 2.1-180

Sapric >.195 <85 >70 <2.1

little is known of these concealed free drainage systems (Ingram
1967) . Wetland hydraulic conductivity measured in situ by the
piezometer method (Sturger 1968) at depths of 46 and 91 cm was
23.9 X 10"3 and 16.1 X 10"^ cm per day respectively. Two of five
piezometers at the 91 cm depth showed much higher hydraulic
conductivity, indicating the piezometers terminated in or near a
fissure. The fissures running through the deeper peat are filled
with water under positive pressure and there are no surface
physical features to indicate fissure location. Overall wetland
hydrologic characteristics are found to be complicated by the
fact that they vary greatly between wetlands, within wetlands and
with time.

Tracers have seldom been applied to wetland hydrology,
"primarily because of the difficulty in choosing an appropriate
tracer and the disturbance of the experimental area which results
from frequent sampling (Girts 1986)." Knight et al. (1971)
successfully used a tritixim tracer to reveal variable patterns of
movement in a wetland under different drainage conditions. The
flow rate variation was found to be attributed to factors within
the peat rather than to analytical procedure. Girts (1986)
successfully used bromide and chloride tracers to yield flow
rates within the range of published values (see Figure 5-4) .

Bowmer (1987) used bromide and dye tracers to study a man
made wetland used to treat sewage. In this study theoretical
detention time of effluent was predicted using the pore volume of
the system. Preferential flow paths and "dead zones" were found
using the tracers and a substantial proportion of the effluent
was found to travel through the system faster than predicted by
the theoretical retention time. The hydrology of the system was
further complicated by occasionally high evapotranspiration rates
and diurnal changes in water consumption.

37

5.2 MATERIALS AND METHODS
5.2.1 Site Instrumentation

Thirty-eight shallow small diameter wells (piezometers) were
installed throughout the study area during the spring and summer
of 1987. Twenty-eight of these wells are paired, with one well
shallow and one deep resulting in 24 total well sites (Figure 4-
1) . The shallow wells were completed to a total depth between
1.2 to 1.5 m (4 to 5 ft) and deep wells to a total depth between
1.8 and 4.9 m (6 and 16 ft). Wells were constructed of 2.5 cm (1
m) diameter PVC pipe and Timco 0.025 cm (0.01 in) slotted,
capped PVC screen. All wells were completed with a 20-30 grain
sand pack and bentonite pellets. Figure 5-5 depicts a typical
paired well site construction. All of these wells were developed
by surging and overpumping.

Well holes were constructed using 5.1 cm (2 in) diameter
Giddmgs soil core barrels hand driven into the ground. A 7.6 cm
(3 m) silage auger (Arts 1987) modified with a core cutter/hol-
der was used to obtain soil samples in the to 1.2 m (0 to 4 ft)
depth at each well site. An additional 32 auger holes were
excavated near the highway culvert outflow to a depth ranging
between .5 and .9 m (1.6 and 3 ft). These additional auger holes
were positioned (Figure 5-6) so as to facilitate tracer sampling.
A Parshall flume equipped with a Stevens Type F recorder was
installed June 15, 1987 on Swamp Gulch below the acid mine seep
to quantify flow rates into the wetland. A Belfort weighing
precipitation gauge was also installed at the site in June 1987.
The position of both the flume and precipitation gauge are noted
m Figure 4-1.

5.2.2 Acfuifer Characteristics

Porosity, bulk density and particle density were determined
on soil samples obtained during excavation of auger holes. Bulk
density was determined using the Core Method described by Blake
and Hartge (1986a). The Pycnometer Method was used to determine
particle density (Blake and Hartge 1986b) . Total porosity, n,
was calculated with the following equation (Danielson and
Sutherland 1986) :

n = (1 - pb/pp) , (Eq. 5-2)

where pb = bulk density

pp = particle density

5.2.3 Hvdraulic Conductivity

Saturated hydraulic conductivity was measured in the field
using both auger hole and seepage tube (piezometer) methods. A

38

variable head method modified from the hydrostatic time-lag
method of Hvorslev (1951), as reviewed by Cedergren (1977), was
used to calculate saturated field hydraulic conductivity. This
method is based on the rate at which water rises or falls in a
hole after a known volume (slug) has been removed or added. In
the case of the auger holes a bailer was used to remove a volume
of water and the rate of water rise was measured using equipment
designed similar to that described by Beers (1983) . For the
piezometers a volume of a solid steel rod was added and the rate
of water fall was measured using an electric tape.

I -Horizontal Not to Scale

2- Water Levels (WL) on 8/13/87

Auger Hole
C-2-AH ,

Shallow Deep
Well Well

/0-2-S ,C-2-D

r /

Catotelm
grey -black
silt clay
muck

*" ^ _ Top of Casing

_ Deep Piezometric

"shallow" Piezometric

WL
_ Ground Surface

^ Potentiometric
WL

Bentonite
Seal

ELEVATION

feet /meters
5174

Gravel
2-5cm
subrounded
to angular
Belt Fm-
rocK fragments

Sand Pock

Screened
or
Open
Interval

5173
5172
5I7J^
5170
5169.
5168
5167
5166
5165.
5164
5163
5162^
5I6J^
5160
5159
5158

1522

1521

1520

1519

1518

Bedrock
mineralized
Belt Fm-

Figure 5-5. Example of study site instrumentation and lithology.

39

C4

I

o

Ol

<
I

I
J
u

CD

c
111

t-

Ui <

^ *

-J Ui u^
-J (5 CC
lU 3 3

i < W

o o )»

M

K

Ui H
111 UJ

O __CM
t 1-

S + 'O

o._

,i,

z
<

o

<3..

o
o

X

<

o
o

T3
•H
M

cn
0)

O

0)
ty

D
(0

>w
O

Q.
(0

I

in

3
■i-i

40

Hydraulic conductivity was calculated for auger holes and
piezometers, respectively, using the following equations:

and

K = R X fho - h i) (Eq. 5-3)

16DS (t2 - ti)

K = r^ In L lnjChi/h2l (Eq- 5-4)

2L ( R ) t2 - ti

Where: R = radius of cavity

D = depth of cavity below static water table

S = shape factor coefficient

h^ = head at time one

h2 = head at time two

ti = time one

t2 = time two

r = radius of the well

L = length of screen

Hydraulic conductivity was converted to velocity (V) to
obtain a preliminary estimate on which to base the tracer test
sample timing. These velocities determined from hydraulic
conductivity also facilitate comparison with velocities deter-
mined by the tracer test. The formula used to make the K-V
conversion is shown in equation 5-1.

5.2.4 Tracer Experiment

A preliminary dye tracer test was run using fluorescein dye.
This experiment showed preferential flow direction, helped
determine the tracer input point and resulted in an even better
idea of flow rates near the tracer input point. In this
experiment approximately 7 grams of fluorescein powder was added
to 18.9 liters (5 gal) of water, mixed and introduced at
different points along the mine drainage flow path. Tracer
movement was monitored visually.

Bromide was chosen as the tracer in the experiment for use
in determining water flow velocities. Prior to adding bromide to
the wetland, background (natural) bromide concentrations were
measured. These background concentrations were measured using
the bromide specific ion electrode as described below and by the
EPA (1979) titrimetric Method 320.1. In the titration method the
EPA (1979) found bromide recoveries from 83% to 99% and standard
deviations from +.13 to +.44 for sample concentrations ranging
from 0.3 to 20.3 mg/1. The titration method as used in this
study showed 100% accuracy based on two laboratory spikes.

The natural gradient tracer experiment utilized 208 liters
(55 gal) of wetland water at a concentration of 1000 mg/L
bromide. This water was added at the point of highway culvert
outflow at 10:30 AM on September 9, 1987 (time zero). Fluor-

41

escein dye (18.9 liters - 5 gal) was added at ^his time as a
visual aid in bromide sampling. The tracer was added to the
system as an extended slug input at a rate of 0.39 1/s {0.bz:>
gpm).

Flow of the tracer plume was monitored by sampling the auger
holes and wells and analyzing for bromide. Water sampling was
done using portable peristaltic pumps. Care was taken in
sampling so that a representative water sample was acquired, but
not so much water that the flow field was greatly affected. This
care usually amounted to pulling one well volume to ^inse the
sample bottle three times and taking the 100 ml sample with the
second well volume.

Water samples were then analyzed for bromide in the field
using a bromide selective ion electrode in conDunction with a
pH/millivolt (mv) meter. Calibration curves (mv vs mg/L) were
generated using standards measured under the same conditions as
the samples (Orion Research 1982). Using the bromide electrode
the concentration range is from 0.4 to 79,900 mg/L with a ±2%
reproducability .

The sampling intervals at the start of the experiment were
nearly continuous near the input point. Sampling had decreased
to once every other day, by September 14 (124 hours) . From
September 26 (412 hours), to the end of the field experiment,
November 21, 1987 (1755 hours), sample sets were taken once each
week, water sampling was terminated after the tracer Pl^^^ had
passed a particular well or auger hole. November water sampling
was hampered by freezing conditions. The passing of a tracer
plume was determined from the bromide breakthrough curves.

Bromide breakthrough curves are plotted as bromide con-
centration in milligrams per liter (mg/L) vs time m hours,
concentrations were generated from the sample analysis data from
a particular site and calibration curves from the time of sample
analysis. Times represent the time a particular sample was taken
from the wetland. The average velocity of water flow was taken
to be the distance from the tracer input point to a particular
well or auger hole divided by the time of peak concentration for
that site breakthrough curve.

5.3 RESULTS AND DISCUSSION
5.3.1 Aquifer Character istics

Field evaluation of auger hole and well cores showed a
fairly distinct boundary between the upper, undecomposed (fibric)
oraanic matter (sedge peat) and the underlying moderately
decomposed (hemic) organic matter. Little distinction was made
between moderately and well decomposed zones which are grey ro

42

black organic matter muck with occasional sand and/or gravel (<5
cm) stringers and rotted wood pieces (<10 cm) . The horizoniza-
tion scheme used distinguishes between two soil types 1) upper,
undecomposed organic matter (acrotelm) , and 2) underlying muck
(catotelm) (Figure 5-5) . The upper undecomposed zone is aerated
and subject to fluctuating water conditions and water is
transmitted comparatively faster in the upper zone compared to
the underlying muck zone which is anaerobic and constantly
waterlogged. In the wetland being studied, the acrotelm
thickness ranges from 0.24 m (0.8 ft) to 1.04 m (3.4 ft) with an
average of 0.58 m (1.9 ft) (Figure 5-7) and catotelm thickness
ranges from 0.58 m (1.9 ft) to 4.30 m (14.1 ft) with an average
of 2.16 m (7.1 ft) (Figure 5-8). This information is summarized
in Table 5-2. Note from figures 5-7 and 5-8 that the greatest
thickness for both zones are near the highway culvert outflow.
These greater thicknesses may be due to sediment deposition from
Swamp Gulch or metal precipitation originating from AMD water.

CARBONATE IVIINE/>

/% II

A WELL SITE

O SURFACE WATER SAMPLE

SITE
Q SURVEY CONTROL STATION
]■ FLUME
O PRECIPITATION GAGE

SEC 20, TI5N,R6W
LEWIS AND CLARK CQ .MONTiANA

FEET
200

50
METERS

Figure 5-7. Acrotelm isopach map with 0.2 m (0.7 ft) contours.

43

CARBONATE MINE/> //

A WELL SITE

O SURFACE WATER SAMPLE

SITE
D SURVEY CONTROL STATION
H FLUME
O PRECIPITATION GAGE

SEC20, TI5N,R6W
LEWIS AND CLARK CO .MONTANA

FEET
200

50
METERS

Figure 5-8. Catotelm isopach map with 0.5 m (1.6 ft) contours.

Both sediment deposition and metal precipitation are
indicated in the wetland soils data (Appendix C) . Sediment
deposition m the acrotelm near the highway culvert outflow site,
C-1 (0 to 91 cm), is indicated by a higher sand content (37%) and
lower organic carbon content (12.4%) than the other three wetland
acrotelm sites analyzed for these parameters. Acrotelm metal
loading at this site is indicated by the highest observed iron
content m the study site at 415,000 ppm (41%). The sediment
deposition and metal loading at site C-1 has had an interesting
effect on cation exchange capacity (CEC) and anion exchange
capacity (AEC) . The low organic carbon content and high sand
content of the acrotelm at site C-1 h^ve likely contributed to
^tl^ °^f®^®^^i°^ ^EC of 4.9 meq/lOOg. : The resulting increase in
CEC with depth at this site is the opposite of that expected and

?pc

44

(0
4J
(0
t3

C
O

-H
4J
Q)
i-l

Oi

e

o
o

0)

no

C
(0

>1
x:

Ck

to

M
01
■<-i
4J
(C

4J
(fl

t)
C
(0
1— I
-p
(1)

I

in
(U

r-l

(0

12 u

4J ir.

C IT

in
in
0)

■c ^

K ^
IT U

C --I

If!

•□ in
oj a:
in c 1/!
□ ^ o
a. u m

6 -I I
D ^ D
u t- t-

X! * w

c r: D

D C

T3 m IT!
0) - IS

4J C CO

n X I

-1 Q
U U H
C 0)

m a
<r <L

in

ID

m
•-* I

— Q

3

0!
ID

nj

U-: IT

in

nj
in

IP
o

in m

m

in - -I
in - -<

in

m
tn

- 01

o
o

ID riE IE a)

UJ r ID UD ff)

nj

T

R

ru m m r-^

nj

'^

tr.

m
tn

w oD in t^ *-<
o ru n o —

nj

ru

n

ru

aj m m --1 n

—

c

tD Tn lf\ T-~ -T c o -^ oo
ID in Lriu3r*CLn tn oo

nj O! ^njOJCOj o oo

^

ID

IE —

in

-r- t-i

fli

r

3- IE

m

O 01

ID ID

ID

S

O
O

m ^ "
in ID ID

IE

r

O O

O

O

o

o c o

o

ID -^ '-' O '^ O
T ID ID O ID n

O O C O O

O O
C O

oaDmmruiimajDDmiEaJCEoajcccDiDCDCJDnJiEiricccmiEiEiEtDcmtriiDiEiEtD
rnSinininffitninininirininnuiuiLnrir. mininininu-inm

OOOOOOOOOOOOCDOCOCCOOCCOOODOOOOOOOOOOO

m

rt -• N tU

mo (T ru

o -■ o -

T {L 3'

01 — o

« O '-I
01 D3 U)

O O

in-HriijrncuNrtuininnjinm-H!\j3-ri\j'-orrnoi--ri\jmmN[~-minrioippo
rinSriiniDiurajK;miC(\jNrn-njmtiirtorun.'OitnrnjiurumnjinN--Htu(bin

(Ti^mrt^nJ-tu--"3-«-'m-|'>J-^-'«fnnj-'rtrn-.nj«--i--nj-itu-i-HtuO

D in o ui

a in

c m

I I Q in □ in ru (^

mmii Dtniiiainirainaincinoini ainoin _..-., ^^

ic??riii,i7TTJHiJrl,r[,nnrrinirTTlirumn-iVm---.rurbmujuJ

gSISiiiy^^iuiuuuui:;iuu55iioiii.:.i.i^i^i:.ilntn

0!
C U

-H
- C

IC (C
^ Q

O b
I- D

• +

45

observed at the other wetland sites measured for CEC. The AEC
exhibits an inverse relationship with CEC and decreases with
depth at site C-1. The high AEC (90 meq/lOOg) at site C-1 may
have the effect of retarding the bromide anion flow velocity by
sorption of bromide on the exchange sites.

Values for bulk density, particle density and porosity were
measured on three surface soil samples (acrotelm, to 30 cm
depth) . Undisturbed samples were not possible below the 30 cm
depth due to the sampling methods used and the nature of the
wetland soils. The average bulk density (0.26 g/cc) is higher
than that found by other studies at this depth in a wetland
(Sturger, 1968 - 0.184 g/cc; Chason and Siegel, 1986 - 0.06 to
0.165 g/cc and; Boelter, 1969 - <.075 g/cc). The distinctness
between the measured bulk density and that from other studies may
be due to differing vegetation types and/or soil compaction
during the sampling procedure.

Particle density values range between 2.077 and 2.942 g/cc
with an average of 2.595 g/cc. This average density is less than
the 2.65 g/cc used as an approximation for many mineral soils
(Danielson and Sutherland 1986) , but higher than the usual case
for humus which is <1.5 g/cc (Blake and Hartge 1986b). The
relatively intermediate, average particle density is presumed to
be the effect of the high organic matter content and high heavy
metal loading in the wetland.

Bulk density and particle density were used to find the
total porosity (equation 5-2). Total porosity ranged from 87.35%
to 94.27% with an average of 90.30%. This average porosity is
similar to that found by Boelter (1969) which was >90% (Table 5-
1), but higher than that calculated by Girts (1986), 66.1% to
76.1%. The values found by Girts (1986) were for partially and
greatly decomposed peat and assumed no water was held within
plant tissues. Boelter (1969) noted that total porosity
decreased with degree of decomposition or depth in a wetland.

Effective porosity is less than total porosity because in
small pores the retention forces are greater than the weight of
water. Effective porosity in practice may be considered equal to
the specific yield of an unconfined aquifer (Kruseman and Ridder
1983). Generally, specific yield is the water yielded by gravity
drainage from water bearing material (Lohman 1979) . Boelter
(1965) found specific yields in undecomposed peat ranged between
0.79 and 0.33 cc/cc and that moderately to well decomposed peat
ranged from 0.22 to 0.10 cc/cc. The averages of the above
ranges, 0.54 and 0.186 cc/cc respectively, were used as the
effective porosity for converting field hydraulic conductivities
to velocities. The decrease in effective porosity with decom-
position is appropriate because the more decomposed peat
materials have more small pores which are not easily drained.
The screens of the deeper wells were completed within or near the

46

gravel zone lying below the wetland. The effective porosity of
deep wells completed in gravel, as used in conductivity to
velocity conversions, was 0.25 cc/cc (Driscoll 1986).

Results of water level measurements in both wells and auger
holes (Appendix B-4) indicate both upward and downward vertical
hydraulic gradients in parts of the wetland. The artesian
gradient (upward vertical) can be explained by two different
arguments: 1) a confined aquifer system at depth, or 2) the
wetland serves as a groundwater discharge area. Hydraulic
conductivities which will be discussed later, suggest that the
wetland system is capable of transmitting water throughout the
soil profile (not confined). Therefore, the wetland is con-
sidered both a groundwater discharge and recharge area. Field
observations, such as at the culvert outflow, suggested water
flowing into the wetland also seeped into the soil profile and
became groundwater. Surface flow channels which would disappear
into the ground and apparently to reappeared down gradient were
also noticed in the wetland. Situations where wetlands function
as groundwater discharge, recharge and flow-through areas are not
uncommon (Winter and Carr 1980) .

Deep and shallow piezometric maps and a potentiometric map
are shown in Figures 5-9, 5-10, and 5-11, respectively. A
general comparison of these maps reveals that the hydraulic
gradient does not change between water surfaces measured, and
there is a hydrologic and topographic (Figure 4-1) high near the
culvert outlet. The average hydraulic gradient across the
wetland where the hydraulic conductivity tests were performed is
0.025 m/m. The change in head within a shallow and deep well and
auger hole nest is generally less than 0.30 m (1 ft). This
relationship can be seen in Figure 5-5. Overall piezometric
surface levels stayed constant throughout the period measured.
The potentiometric surface fluctuated approximately 0.30 m for
short periods of time (days) during the tracer test. These
fluctuations may be attributed to increased water use by
phreatophytic plants (Mayboom 1967) during period of high air
temperatures and the lack of evapotranspiration during the
winter. This is further evidenced by slight diurnal fluctuations
in water flow rate at the Swamp Gulch flume. The rate of water
flow into the wetland from the culvert exhibited little change
during the tracer study as indicated by the flume recorder charts
(Appendix B-1) . Precipitation during the tracer study was very
low (Appendix B-2) .

5.3.2 Hydraulic Conductivity

Hydraulic conductivity tests were run on auger holes and
wells in the area of expected metal loading and expected tracer
movement. This area was determined from water chemistry data and
hydraulic gradients in the wetland. Field determined hydraulic
conductivity values and the calculated water velocities for each

47

CARBONATE MINE/>

f% // II

A WELL SITE

O SURFACE WATER SAMPLE

SITE
D SURVEY CONTROL STATION
B FLUME
O PRECIPITATION GAGE

SEC 20, TI5N,R6W
LEWIS AND CLARK CQ, MONTANA

FEET
200

50
METERS

Figure 5-9

Deep piezometric map on 7/27/87 with 0.5 m
contours .

auger hole and well tested are given in Table 5-3.

When field hydraulic conductivity values were grouped, some
differences were found between auger holes and wells (Table 5-4) .
Lowest conductivity values were found in the shallow wells with
their screened intervals within the catotelm (well decomposed
organic matter, muck) . Highest conductivity values were found in
the deep wells with their screened intervals within both the
catotelm and the underlying gravel. Auger holes which have their
open intervals within the acrotelm (undecomposed organic matter)
had conductivities intermediate in comparison to the wells (Table

48

CARBONATE MINEX- '/

A WELL SITE

O SURFACE WATER SAMPLE

SITE
O SURVEY CONTROL STATION
]■ FLUME
O PRECIPITATION GAGE

SEC20, TI5N,R6W
LEWIS AND CLARK CQ, MONTANA

FEET
200

50
METERS

Figure 5-10.

Shallow piezometric map on 7/27/87 with 0.5 m
contours .

5-4) . These values fall within the range of conductivities found
in other wetlands studied (Figure 5-4). The range of hydraulic
conductivities from the wetland, 1 x 10"^ to 1.5 x 10"^ cm/s,
compares to the range commonly found for silty sand (Freeze and
Cherry 1979) . The decrease in conductivity with depth,
excluding the deep wells completed in gravel, is a common
occurrence when moving from the acrotelm to the catotelm (Chason
and Siegel 1986) .

An argument to account for the small conductivity difference
between auger holes and wells is that the tests measured part of
the same zones. In viewing the depths of auger holes tested

49

in

CM

x:
-p

00

'3'

00

c

o

a

(0

g

(U

o

(0

IM

u

3

tn

o

•^^

u

■p

(U

e

•

O

CQ

■1-1

U

•4J

3

C

O

<U 4J

4J

c

O

o

a<

u

I

in

0!

50

Table 5-3.

Field

hydraulic

conductivit

:y (K), effec

;tive porosity (ne) ,

average hydraulic gradient

(dh/dl) and

calculated average

velocity (V) for each site.

Site

K 1

'cm/s) n«

rcc/cc) dh/dl (m/m)

V(

cm/s)

B-2-AH

1.06

X

10-3

.540

.025

4.91

X

10-5

BC-1-2-AH

5.60

X

10-4

.540

.025

2.59

X

10-5

r-

C-l-AH

7.02

X

10-4

.540

.025

3.25

X

10"5

C-1.5-AH

1.06

X

10-3

.540

.025

4.91

X

10-5

C-2-AH

9.53

X

10-4

.540

.025

4.41

X

10-5

CD-1-2-AH

1.18

X

10-3

.540

.025

5.46

X

10-5

CD-l-AH

4.54

X

10-4

.540

.025

2.10

X

10-5

D-l-AH

4.18

X

10-4

.540

.025

1.94

X

10-5

E-l-AH

3.40

X

10-4

.540

.025

1.57

X

10"5
10-5
10-5
10-3

r-

E-2-AH

1.08

X

10"3

.540

.025

5.00

X

F-l-AH

1.11

X

10-3

.540

.025

5.14

X

B-2-D

1.01

X

10"2

.250

.025

1.01

X

B-2-S

1.71

X

10-4

.186

.025

2.30

X

10-5
10"^
10-5

BC-1-2

3.89

X

10-5

.186

.025

5.20

X

C-l-D

2.81

X

10-4

.186

.025

3.78

X

C-l-S

6.02

X

10-4

.186

.025

8.09

X

10-5
10-5
10-5
10-5

rr

C-1.5

3.00

X

10-4

.186

.025

4.03

X

C-2-D

1.22

X

10-4

.250

.025

1.22

X

C-2-S

1.44

X

10-4

.186

.025

1.94

X

CD-1-2

2.22

X

10-4

.186

.025

2.98

X

10-5

CD-I

1.40

X

10-4

.186

.025

1.88

X

10-5
10-4

D-l-D

1.69

X

10-3

.250

.025

1.69

X

D-l-S

6.95

X

10-5

.186

.025

9.30

X

o in in L
o o o

H H 1-1

D-2-D

6.62

X

10-4

.186

.025

8.90

X

D-2-S

1.92

X

10-4

.186

.025

2.58

X

E-1

9.16

X

10-5

.186

.025

1.23

X

10-=

E-2

3.21

X

10-5

.186

.025

4.30

X

10"°
10-5

F-l-D

5.05

X

10-4

.186

.025

6.79

X

F-l-S

1.53

X

10-5

.186

.025

2.10

X

10-^

that the auger holes also measured the decomposed organic matter
zone (catotelm) layer thereby possibly reducing conductivity. A
similar argument is true of the shallow wells some of which had
the top of the screened interval and sand pack very near or
within the acrotelm. The effect of this situation is that well
tests measured a portion of the acrotelm in addition to the
catotelm and may have greatly overestimated the actual conduc-
tivity of the catotelm. The overall result is an average
conductivity from the water table to shallow well total depths,
but not the two distinct conductivity differences that might be
expected. The average conductivity of both the auger holes and
shallow wells is 4.9 x 10*4 cm/s and the velocity is 3.0 x 10"^
cm/s.

51

Table 5-4. Average hydraulic conductivity (K) from field
results and calculated average velocity (V) .

Site Type K fcm/s) V (cm/s)

Auger Holes 8.11 x 10-4 3.75 x lO'S

Shallow wells 1.68 X lO'^ 2.26 X 10 ^

Deep wells 2.22 x 10*3 2.31 x 10 ^

overall Mean 8.38 x lO'^ 7.14 x 10 =>

The hydraulic conductivity determined in this study is the
effect of both vertical and horizontal conductivity. No
distinction between vertical and horizontal conductivity was
possible with the methods used in the field. It is presumed from
other studies (Girts 1986; Chason and Siegel 1986; and Dai and
Sparling 1973) that horizontal conductivity would be greater than
vertical conductivity in the sedge peat wetland being studied.

Average velocities calculated using field hydraulic
conductivities and equation 5-1 are shown in Tables 5-3 and 5-4.
Velocities determined were lower than the associated field
hydraulic conductivity. This order of magnitude change (Table 5-
4) is explained, in part, by the effective porosity chosen for
the velocity calculation. From equation 5-1 it can be seen that
as effective porosity decreases the water flow velocity will
increase if the hydraulic conductivity and gradient remain
constant. Another explanation for the lower velocities deter-
mined for the field sites measured is the small average hydraulic
gradient in the wetland (0.025). The hydraulic gradient of the
potent iometric surface in the area of tracer input, culvert
outflow to C-1 is 0.058. Incorporating this increased gradient
into velocity calculations increases the average velocity from
auger hole sites to 8.7 x lO'^ cm/s and velocity from shallow
well sites to 5.2 x lO'^ cm/s.

5.3.3 Tracer Experiment

Water flow velocity was also determined in a bromide tracer
experiment. Bromide was chosen as the primary tracer in this
study for a number of reasons. Many authors have successfully
used bromide as a tracer in a variety of environmental conditions
(Onken et al. 1977; Smith and Davis 1974; Merrill et al. 1985;
and Tennyson and Settergren 1980; to name a few). In addition.
Girts (1986) and Bowmer (1987) successfully used bromide to trace
wetland water systems. Davis et al. (1985) note that bromide is
perhaps the most commonly used ion tracer. Schmotzer et al.
(1973) consider bromide to be as close as possible to the ideal
tracer. Bromide is a good tracer choice, relative to other

52

f

tracers, in the wetland system studied because it is inexpensive,
stable, has a low limit of field detection with a specific ion
electrode (0.4 mg/L) , has low background concentrations, exhibits
low toxicity and has low sorption. The relatively few disad-
vantages of bromide as a tracer are anion exclusion, anion
exchange and the presence of other ions which interfere with
operation of the electrode. Data from this study indicate
bromide is a good choice in the system studied.

Bromide background concentrations and the concentrations of
ions which interfere with the electrode were analyzed before the
tracer was introduced to the wetland. Background concentrations
were analyzed by both the specific ion electrode and by titration
(Table 5-5) . No auger hole background concentrations were above
the 0.4 mg/L limit of detection using the bromide specific
electrode. Well water which had background values above the
electrode lower limit of detection, 0.4 mg/L, had that value
subtracted before plotting the breakthrough curve. The
subtraction of background values determined from the bromide
electrode was necessary only for five wells. The average bromide
value subtracted was 0.47 mg/L with a range from 0.43 to 0.55
mg/L. This subtraction was found to change the shape of the
breakthrough curves, but not the placement of the concentration

Table 5-5.

Bromide background concentrations in the tracer
study area of the Swamp Gulch wetland.

Site

Bromide Concentr ation rma/L)

Determined by Ion Determined by

Electrode 9/8 - 9/87 Titration 7/28 - 31/87

AH-6

AH-14

AH-17

AH-22

AH-29

AH-32

BC-I-2I

BC-l-S

C-l-S^

C-1.5^

C-2-S

CD-l^

CD-I-2I

D-l-S

D-2-S

BLANK

BC-1-SURF

0.40
0.40
0.40
0.40
0.40
0.40
0.43
0.40
0.44
0.50
0.40
0.45
0.55
0.82
0.40
0.40

<

.10

<

.10

<

.10

0.

10

<

.10

0.

20

<

.10

0.

20

^Backgrounds values subtracted from breakthrough curve.

53

peak. Bromide background levels measured by titration were never
larger than 0.2 mg/L and 80% of all the wetland samples measured
were <0.1 mg/L (detection limit). Review of the two methods of
analysis suggests that the titrimetric method may be more
accurate than the field electrode method, especially at low
bromide concentrations. The advantages of using the field
millivolt meter and bromide electrode are quick, inexpensive
results at the site. This allows for rapid on-site sampling
decisions and no need to save a large number of samples for
expensive laboratory analysis.

Ions which interfere with the bromide specific electrode, in
order of increasing interference, include S""^, CN", I", NH3, Cl~
and OH" (Orion Research 1982). Sr''"^ (strontium) must be absent
(Cole-Parmer 1987). Water quality analysis (Appendix B) of the
acidic wetland system indicate that, of the interfering ions
above, only chloride and ^trontium may cause inaccurate electrode
readings. Williams (1979) relates that a chloride concentration
of 20 times the bromide concentration will cause error in the
bromide determination. At the lower limit of detection with the
bromide specific electrode (0.4 mg/L) a chloride concentration of
8 mg/L can cause error. Water analysis at well D-l-S and surface
water near BC-1 showed chloride concentrations of 16 mg/L and 8
mg/L respectively on 7/29/87 (Appendix B) . The most likely
source of this chloride is highway road salt runoff. All wetland
sites sampling strontium showed less than the lower limit of
detection (0.1 mg/L) except for well F-l-S which showed 0.1 mg/L.
Chloride and strontium may thus have created limited variation in
the readings for bromide during the tracer test.

The time involved in a uniform (natural) flow tracer method
can be quite long and the method is usually limited to short or
intermediate distances not more than 100 m (Davis et al. 1985).
The velocities determined from the field hydraulic conductivities
and preliminary tracer test suggested that the piezometer network
might not be close enough to the tracer input point. In addition
to variable velocities, conductivity was noticed to decrease with
depth and this suggested that flow would be predominantly in the
acrotelm. Therefore, a grid of shallow sampling sites was
installed nearer the tracer input point. Inexpensive auger holes
were used as the sample sites. The preliminary dye tracer test
exposed preferential water flow paths and auger holes 27 to 32
were placed along these paths (see Figure 5-6).

Bromide standards ranging from 0.1 to 1000 mg/L were run
during each sample session to determine the corresponding mv
value for that time period. An example of the resulting
calibration curve is shown in Figure 5-12. Regression equations
were determined for each calibration curve. The coefficient of
determination, (r^) , for each regression was maintained at 0.99.
Values which were below the 0.4 mg/L limit of detection were
excluded from the regression. Field mv values for each sample

54

E

<

o

CL

UJ
Q
O

q:

h-

o

1

BROMIDE CONCENTRATION (log mg/l)

Figure 5-12

Typical calibration curve generated for the tracer
study .

site were converted to mg/L values using the regression equation
for that corresponding time period. Distilled water blanks were
run during each sample session and all were less than the 0.4
mg/L limit of detection.

Bromide breakthrough curves were developed using actual
bromide concentrations on the y axis instead of the commonly used
concentration/concentration injected (c/co) . This was done to
facilitate calculations and has no effect on the shape of the
breakthrough curve. Changing the y axis to concentration/ concen-
tration injected is commonly done in tracer studies to facilitate
comparison of different types of tracers introduced at the same
time. An example of a bromide breakthrough curve generated is
presented for site AH-14 in Figure 5-13. Based on the work by
Davis et al. (1985) (Figure 5-1), the shape of this curve (steep
slopes) indicates little sorption of tracer in the porous medium
between the input point and sampling site. The curve shape also
exemplifies the expected result that bromide can be considered a
fairly conservative tracer in the wetland studied.

There are many possible explanations for shapes of the
breakthrough curves. Some of these explanations have been
discussed earlier such as retardation by sorption and dilution by
dispersion and diffusion. Site AH-25 generated data which

55

D SITE AH- 14

Z
O

t<

t-
z

o
z

o
o

0.9

0.8

0.7

0.6

0.5

0.4

40

80

120 160 200 240

TIME (hrs.)-fronn start of tracer input

280

320

Figure 5-13. Tracer study breakthrough curve for site AH-14, 6.1
m from tracer input showing little effect of
sorption.

resulted in only one point above the lower limit of detection.
All the neighboring sites, 1.5 to 3.7 m (5 to 12 ft) away
generated many bromide breakthrough curve points up to 4.8 mg/L.
This suggests zones of very low flow which are described as "dead
zones" by Bowmer (1987). At some sample sites the sampling
interval was not frequent enough to determine the exact break-
through peak. In the case of site AH-9 (Figure 5-14) the
resultant curve is flat topped between hours 169 and 220. Both
of these times were used to calculate velocity in addition to the
peaks at 30 and 54 hours. Subtraction of the background values
resulted in no detectable tracer peaks on the breakthrough curve
for some of the well sampling sites such as C-l-D, D-l-S, and D-
2-S.

Placement of the breakthrough curve peak is important in
that the associated time represents the passing of the tracer
plume. This time is also used for determination of average
velocity. Widely spaced sampling intervals may cause the exact
breakthrough curve peak to not be found and the velocities
determined may be too large or too small. Bromide values below
the 0.4 mg/L limit of detection were not considered as possible
peaks on the breakthrough curves. A common observation of the

56

\

z

o

<

o
z

o
o

1.7
1.6
1.5
1.4
1.3
1.2
1.1
1
0.9
0.8
0.7
0.5
0.5
0.4

\~iA

' ^3^

H r

- i V

- i 4

-^ \

^ ^

i ^^-^

y

r

y

10 30 50 70 90 110 130 150 170

TIME (hrs.)— from start of tracer input

190

210

230

Figure 5-14. Example of infrequent sampling effect on break-
through curve for site AH-9, 6.1 m from tracer
input .

breakthrough curves shows more than one peak of tracer concentra-
tion. An example of such a curve is given in Figure 5-15 for
site AH-19. This curve seems to indicate two bromide tracer
plumes moving past the site with the second plume being more
effected by sorption.

How breakthrough curves acquire more than one concentration
peak is not specifically discussed in the literature reviewed.
Possible explanations of multiple breakthrough peaks could
include: inadequate tracer mixing during input, a heterogenous
medium with layers of different flow velocity, periodic flushing
and/ or cross contamination of samples. The inadequate mixing
theory is negated by the breakthrough curve shape of the first
sampling site encountered (Figure 5-16). This site, AH-3, which
showed no multiple peaks and exhibited a high concentration and
short time interval to the breakthrough curve peak for this short
distance from tracer input. A heterogenous medium is an entirely
plausible explanation as layering was indicated during soil
sampling (Table 5-2) and preferential flow paths were found in
the wetland. These situations are also found in other wetlands
(Ingram 1967; Dasberg and Neuman 1977; Sturges 1968). The
possibility of samples being drawn from more than one zone of
flow is good because many sampling sites have their screened

57

z
o

<

o

z
o
o

2.7 -I

m

2.6 -

r

\

2.4 -

2.3 -

2.2 -

'

I

\

__—--'

k.

\

\,

/

\

\

/

^

\

/

/

\

/

I

/

\

/

^

\,

0.8 -

n/

B

N

c(

1

0.5

1

20

60

100

140

180

220

260

TIME (hrs.)— from start of tracer input

Figure 5-15. Breakthrough curve representing two tracer plumes at
site AH-19, 9.1 m from tracer input.

110

100

90

80

c

t-

70

z

o

&

60

a:

t-

z

50

O

Z

o
o

40

l_

JO

20

10

/

\

/

\

\,

/

\

\

/

s

\

/

\

/

\

/

\

N

\

\

s

1^

\

Vi

0.4

0.8

1.2

1.6

2.4

2.8

3.2

TIME (hrs.)— from start of tracer input

Figure 5-16. Steep breakthrough curve indicating adequate tracer
mixing at site AH-3, 3 m from tracer input.

58

interval in contact with more than one soil horizon (Figure 5-5
and Table 5-2). It is theorized that pulse flow could cause
multiple breakthrough peaks by flushing higher concentration
water which was previously held in pore spaces during different
flow rates. Periodic flushing is indicated by mildly fluctuating
water levels during the study period (Appendix B-4) . The diurnal
variation noticed in flume charts could also contribute to this
type of flow.

The possibility of cross contamination of samples, analysis
error and/or some other means of acquiring faulty values is one
which must be considered in any experiment. Blank samples
(distilled water) were run through the same analysis procedures
as the wetland samples. The bromide concentration in blanks
never was greater than the 0.4 mg/1 limit of detection and cross
contamination was not a problem. Another explanation for
multiple breakthrough curve peaks is error during analysis for
bromide in the field. Analysis methods were constant throughout
the study and should not have created the noted variation. The
final explanation for multiple peaks is a periodic contamination
by the interfering ions, chloride and strontium which seems
unlikely.

Given these possible explanations for multiple tracer peaks
it seems that they are true bromide values and have been
processed accordingly. The previously shown breakthrough curve
for site AH-14 (Figure 5-13) yields a velocity of 1.4 x 10"-* cm/s
(6.1 m/121.7 hr) . Table 5-6 shows the velocities associated with
the bromide breakthrough curve peak/peaks chosen for each site.
The velocities shown represent the rate at which a tracer plume
passed that particular site. Site locations are shown in Figure
5-6. The overall mean velocity represented by all of the sites
and peaks is 2.6 x IC^ cm/s with a range of 3.1 x lO"-*- to 6.2 x
10-4.

The velocities were compared in three different ways 1)
between auger holes and wells, 2) between the highest peak of the
breakthrough curve and the remaining peaks, and 3) between sample
sites within the grid area, <18.3 m, and sites outside the grid
area, 27 to 38 m (Figure 5-6) . These results are shown in Table
5-7. Several observations were apparent: 1) there is little
difference between tracer velocities determined at auger holes
and wells, 2) the velocities determined from other than the
highest peak of the breakthrough curve are notably slower than
from the remaining peaks, and 3) there is little difference (same
order of magnitude) in velocities determined near the tracer
input and those velocities from farther away.

The apparently similar velocities observed at auger holes
and wells can be explained by the fact that portions of the

59

n

w

c

c

<a

-r-l

0)

-P

e

1X5

-iH

v^

-U

4J

c

••

OJ

d)

u

-P

c

•r^

m

4:

M

m

OS

<u

0)

a

>4-l

C

e

U

0)

4J

Q)
to

t3
C
(0

M

D

a
c

d)
o

(0

e
o

4-J

d)
>

0)

0) -rH

u e

W c o

0) (B W

U 4J JQ
to 03

E-i TS O

I

ID

Eh

- E

u u

c

3

CD--.

u -u g
r 10

c -^

E I.

- E
U U

o

0)

■ c -^

c c ^

E - D)

C -^ E

:: a)

c in
E L.

QDmo«r^ommmmmu3Uj"in-mf^mu3uia3m?-ix3irmmrnjm-rxj-gmm?:mrnjt--crrmr-.r
ir-moiinoornommmmrajtommjvuD-mLCocDcnmcri-Hmmj-innjooiommm^tncnipr-mmo

nocojoomooSoooooooooccocooooocqooooqccoooqooooooooo
oooooooooooooooooooooooooooooooooocooooooooococcoo

ajmnnjajfijajnjturnrnrTij-3-T'"^'-'^TTCD02'-'-^^'-'Lniri/iini/iLnini^NNr-njaj(XJiJi(r5[nf^Nf^NNrv
u^maDaJnJruLnmlna)tDc□lDU)U)UjU3lXlii3^^oou^u^lrlr^--^^^^N^^^l/ll^mL^^^^l^JWaJTr^J•Tr

IP o T o n tj! -<

o - o ru n m fu
ni in T

nj ID pj-^ -'-* io-Htrajrurnuiajruj-ajrnLnnjLn'-<i/irnLn--HrnLnrij--'niincu-^i/ifiJtci/)nj'-^'^Ln'ur\j--''-'
-Hrunjmrnrrur'-inj .-Hnj— en u) — •* ^cn^tu—tr ^ruTtn

D D I <r <r <r o: 111

xxxxt/itni/itnnjf^jnjnj ' i 1 truojtuiDnjnjtvj

(Da)tr!ooO'^^'-*rurL;ruLny]Ln<r<r<r<Ei 1 1 1 1 '• 1 iinLT. intjiiriLruii 1 i 1 i 1 1

cururumrnmrnrnrnmnrniiiifi

r X X I X X
<t C C C <! <t

t

I

1

I I

i I i t t

xxxxxixixxxii I i I I I luuijuLjuuui I i I I I iEJacnaDaaonnooDCD
<i<ic<i<ic<r<r<E<rccu'UU'_)UU(jmcaD2£DCDc:cEmij'iJLJUULiuuijLJUUuut_;uuc;uuuuu

LE'C^inui-'O — cruoiiTjaicm-^O!^— 't^'-^oot\jC^L£icnc£r^LnofULnc7!i/i?*Lr!TLninmLnLnr\JLnirirjrucDru
lEircajCDiXiinLDOnjruoiDCiDajOr-ruNcnmoc-'N-'O'^o — [DiD--u^In^"^aI-'^--' — NOicncDUJajm
nj^rbr^oj^^-^-'h-mivui^LnfTi^ornoirJCD— <[r!ir;Nrn-^-*nj'HOC'-'trjOT^ro^^-^"?"OCOOLDC

OCCSOOCCCfUOOJOOOOCOOC — mOT---rL'000000001ECCOOOOOOOCCCOOOO

oo-occocococcooooooonj-onj-'cooooooooooooooocoocoqcqoq

COOOOOOOOOOCOOOOCODOCOOOOCCOOOOOOOOOOOOOOOOOOOOOOOO

ocooGooo-'-<-i-^^---*-~-<-i«-<, — ^„_j^ — ^„_,_„_„-^„_,_rt — ^-<-.--- — ^-Hfunj
nmr^nrnnnr^. mu3U3:3U3tx)UDt£irtrtDLDiD[x:ii3ir'JUDLCiiitD0itTicnaicn(r;c^ci[noia5CnEnff3oiiTimCTcr)trrurv.'

fttcLT'^icmcntx.Oiiiirtcr--
— — pntDru-H-<nj-'3--'naj-<

'tD[^m-'tcErII)UDa^r'^u^cnooNr^tI)£^3C"^tCL^tXJa)Lna^l

r^ UD m 01 ^ -I

-H-iOc — cru^LOnjrc^n--^^coc.•
(Ii CD r^ —

:TtU'^'-C2-''-''\lTOOO-

cnrrc^tnojr-'rutDLDnjLrinircnrutrcruoo^inc— *LcrT-, cD'-'fijCCD — crntEnomoDOrnccf^aDr-r^orno

njin ojinj-LniD tn njUDajinuDruirinj m Lntxirurunj— <^ru uDi/i"ui/iU3ruLnnjruJiu3U)tD'--i/i-*

" -1 '-tfuaj ^ .-iw.^njmj-t\jrunj(unjajfua)njnjj'j*

oc ^-lajDnrmyiniDQCDr-CDmiTimco — rutunrruitDUSt^N

I I ) < I I I I 1 i I I j I ! < I I i I I t 1 I I : I i I 1 I r I I i 1 I I ' i < > I > ' I I < I I I

XXXXXIXXXIXXXIXXXXXXXXXIXXXIXXXXXXXIXIIXXXXXIXIXXIX

<c<E<i<i<i<i<t<r<i<t<r«<t<i<ta;c<tc<i<r<r<r<ic<rc<i<r<r<t<ic:<r<r<E<r<E<rcc;<rcc<i<r<r<r<r<i<i

60

Table 5-7. Tracer velocity comparison.

Comparison Number Mean

Type of Cases Velocity rcm/s)

Auger Holes 83 2.9 x 10"2

Wells 18 1.4 X 10~2

High Peaks 45 5.0 x IC^

Remaining Peaks 56 6.4 x 10"-^

Grid Area 70 3.1 x 10"2

Outside Grid 31 1.4 x lO"'^

-2

Overall 101 2.6 x 10

screened interval of most of the wells are within the same layers
of conducting material (Table 5-2 and Figure 5-6) as the auger
holes. The total depth of the shallow wells has little effect in
that they are able to pick up tracer water from upper layers
(acrotelm) . This "pick up" can occur through the well sand pack
which is usually 30 cm (1 ft) from ground surface, below the
bentonite seal. The auger holes are subject to the acrotelm
flowing water slightly sooner than the shallow wells as indicated
by the slightly faster velocity. The comparison of the high peak
and remaining peaks supports the theory of residual tracer water
being flushed through after the main tracer plume. These
remaining tracer peaks could represent tracer water flowing
through the less conductive (catatolem) peat layers hence
reducing flow velocity. The higher velocities determined in the
auger hole grid area, although within the same order of magni-
tude, may reflect the steeper gradient in the area near the
culvert outflow. Considering all the velocities from the
different comparisons it is concluded that the overall mean of
2.6 X 10"2 cm/s is the most representative of the wetland system
measured .

The presence of preferential flow paths is demonstrated by
results of the bromide tracer experiment. Breakthrough curves
peak at just 1.4 hours from tracer input for sites AH-28 and AH-
31, both 15.2 m (50 ft) from tracer input (Table 5-6). This
results in a tracer plume of the shape represented in Figure 5-
17 . These preferential flow paths indicate the tracer is moving
in both a south-southwesterly and easterly direction. The
bromide concentration at site AH-31 is more than twice as high
(52 mg/L) as site AH-28 (20.5 mg/L) at the same time (1.4 hrs)
and indicates that more tracer is flowing in the south-south-
westerly direction. This southwesterly skewed radial pattern is
expected as the culvert outflow is a topographic high point and

61

CD.

<3
Oi

<
I

<M

I
I

o
to

oc

h-
lU <

-J lU il
-"OK
"3D
$ < OT

<3 O \W

K
K Ul
111 t-
UJ i^
u. S

S + «o

o._

o
o

<
I

OI

O
U

tn

4J

CO

M

in

3

o\

O

•

■P

®

c

o

I

o

< a

J, 'O

IJ

2 ^^

"o x:

e

<3« ^

<s

W QitN

6 o

■ —

w

-1-1

4J

U

c

(0

o

-p

-l-l

to

4J

rO

M

U

(V

4J

o

C

(0

0)

u

U

4J

c

o

a

u

o

S>l

<u

«+-!

n

•fl

Cfl

B

M

o

D

u

O

CD 4=

I

in

u
o
a>
•1-1

62

the regional hydraulic gradient and topographic slope is to the
southwest. Preferential flow paths are found in man made
wetlands (Bowmer 1987) and in natural systems are sometimes
called water tracks (Ingram 1967) . Such paths may cause a
substantial proportion of the water (tracer and/or AMD polluted)
to travel through the system faster than predicted by the field
hydraulic conductivities. Brooks (1984) recommends man made
wetlands be designed to avoid channelization (preferential flow
paths) and encourage low velocity sheet flow by use of flow
obstructions. Burris et al. (1984) shows that as flow rate
decreases effective iron removal from AMD increases in a wetland.

Tracer flow velocities were faster than flow velocities
estimated from field hydraulic conductivity tests on auger holes
and wells. The relationship between tracer determined velocity
and field hydraulic conductivity determined velocity is compli-
cated by a number of factors. The measurement of the conduc-
tivities and the tracer test were done at different times. In
effect, the measures are of different systems, but in practice,
are assumed the same in view of the minimal change noticed in the
system over an extended period. The field conductivity deter-
mined velocities are more representative of the wetland at depth.
The tracer velocities, which include preferential flow paths, are
more representative of the true movement of AMD waters. The
amount of area and time represented by the tracer test is large
compared to the small area and time that the field conductivities
represent. This infers that the tracer velocities are closer to
the true value than the conductivity determined values.

5 . 4 SUMMARY

Bromide was a good choice for tracing water movement in the
wetland studied. Using the natural flow tracer technique,
sufficient bromide concentrations were detected up to a distance
of approximately 38 m from tracer input. The bromide specific
ion electrode method of analysis was found to be adequate. This
method could be improved by periodic double sampling and analysis
with another method (titration) especially at low bromide
concentrations .

The wide range of velocities determined from this study are
the effect of a very heterogeneous wetland flow system and the
methodology used in determining the velocities. Some of the
heterogeneity involved a change in hydraulic gradient, changing
flow rates and periodic flushing over time. Preferential flow
paths and "dead zones", the result of variable hydraulic
conductivity, also add to the heterogeneity. The most represen-
tative velocity of the AMD water in, the wetland was that

determined by the tracer experiment. These velocities were found

63

to be a reflection of the heterogeneous system and ranged between
3.1 X lO"-"- to 6.2 X 10"4 cm/s with a mean of 2.6 x 10"^ cm/s
(73.7 ft/day) .

Comparison of field determined hydraulic conductivity, from
tests of auger holes and wells, to tracer determined velocity was
accomplished by converting conductivity to velocity utilizing the
estimated values of effective porosity and hydraulic gradient.
These field conductivity determined velocities are reasonable in
that they estimate water flow velocity in the small area where
conductivity was determined. When comparing the conductivity
determined velocity to tracer velocity even the average velocity
of auger hole sites (fastest of the conductivity determined
velocities) was three orders of magnitude lower than average
tracer velocity. This order of magnitude relationship does not
change even if effective porosity is changed to 0.1 (low) and the
hydraulic gradient is changed to the steepest observed in the
study area (0.058 m/m) . This relationship infers the comparison
is that of two different flow systems. The tracer velocity
represents preferential flow paths within the shallow (acrotelm)
system and the conductivity represents that flow found at greater
depth (catotelm) .

The determination of water residence times in the wetland is
contingent upon the velocity of the water system to be considered
and the distance the water travels. The concern in this study
involves the AMD water which emanates from the culvert outflow
into the wetland. The velocity representing this source is that
of the tracer which was input at the same point that the AMD
water is introduced to the wetland. The distance the AMD water
travels, before complete problem amelioration, was not deter-
minable from the tracer methods employed, but is estimated to be
approximately 200m as observed in acrotelm soil iron concentra-
tions (Appendix C and Section 8) . The shallow groundwater iron
concentrations show a trend similar to the soils by decreasing in
the down-gradient direction and approaching background iron
concentrations near site F-1. With the 200m distance chosen and
using the mean tracer velocity, the average residence time of AMD
water in contact with the wetland acrotelm soils is estimated to
be 215 hours (9 days) . This estimate is considered to be quite
variable as noted earlier and represents a low flow regime
(Autumn) . This residence time estimate should be viewed with
caution as it pertains only to iron amelioration and the velocity
used represents only that measured at the time of the tracer
study for the very limited area of the wetland near the AMD
discharge point. Assuming the wetland studied is a viable system
for the amelioration of AMD, the challenge in the design of man
made wetland systems will be to determine how to emplace similar
hydrologic characteristics, and limit preferential flow paths.

64

6.0 WETLAND INFLUENCE ON MINE DRAINAGE WATER QUALITY

The water quality sampling program for the Swamp Gulch
study was initiated to determine: 1) the concentrations of
parameters in the AMD; 2) the background water quality in Swamp
Gulch above the Carbonate Mine and on the Blackfoot River above
the wetland; 3) the changes in AMD water quality as it pro-
gressed through the wetland and; 4) changes in the water quality
of the Blackfoot River below the wetland discharge point.

6.1 SAMPLE COLLECTION AND PROCESSING PROCEDURES

Water samples were collected by various means, each ap-
propriate to the sample source. All samples were initially
collected in 4 liter (1 gal) linear polyethylene (LPE) (Nalgene)
decontaminated bottles (acid/deionized water cleaned) . These
bottles were rinsed three times with sample water prior to fill-
ing. The filled 4 liter bottles were placed in coolers with ice
until divided into appropriate aliquots for preservation,
usually within minutes after collection.

Water level depths were recorded at all wells prior to
pumping. These measurements were referred to the top of the
well casing (as surveyed) and recorded to the nearest 0.01 foot
using a Soil Test Model 444-000 or Johnson UOP water level
indicator.

Well samples were acquired with a peristaltic pump (Master-
flex) using a length of silicone tube. The apparatus was rinsed
with distilled or deionized water after each sample, and if
organic deposition on the hose was evident, new hose was in-
stalled. A minimum of three well volumes (including sand pack
volume) of water was pumped before sampling at the higher yield
sites. At low yield sites, wells were pumped dry and the re-
charge was sampled.

Surface water samples were obtained by submersing the four
liter bottles below the water surface where sufficient water
depth was present or by filling the four liter bottles with a
peristaltic pump where small volumes of water required this
method. Flow measurements or estimates were made at the time of
sample collection for all flowing surface water sites.

Various aliquot portions were taken from the four liter
bottles for the required analyses. The samples were preserved
in three separate bottles: 1) 125 or 250 ml of 0.45 micron
filtered sample with 1 ml of concentrated HNO3 preservative
(metal analyses), 2) 125 or 250 ml of 0.45 micron filtered
sample with one ml of concentrated H2SO4 preservative (NO3
analysis) and 3) 250 or 500 ml of nonfiltered sample with no

65

added preservative (pH, electrical conductivity (EC) , major
cations, HCO3, F, P, CI, SO4 analyses). Special bottles were
prepared for the limited number of dissolved organic carbon and
cyanide analyses.

Samples were filtered utilizing a Horizon Ecology peristal-
tic pump (silicone hose) and a Geotech 102 mm filter apparatus.
The pump and filter assembly were rinsed thoroughly with deio-
nized water between samples. Filters and pref liters were
changed between all samples except field replicates. All sample
bottles were rinsed three times with sample water (filtered or
nonfiltered according to aliquot) prior to filling. A minimum
of one field blank was included with each sample collection and
inserted blind within the sample train. The blank was processed
in a manner identical to that of the normal samples. Blind
field replicates were processed from the same four liter initial
sample. Blind EPA field standards were processed in the same
manner as were field blanks. All processed samples were immedi-
ately placed in coolers with ice or refrigerated.

Temperature, conductivity (EC) and pH measurements were
made (in-situ where possible) at the time of sample collection
or, under adverse conditions (ambient temperatures below freez-
ing) , they were determined within one to three hours in a
sheltered environment. The temperature was determined with a
mercury thermometer, dial thermometer or a Cole-Parmer SL-5985-
80 pH meter with a digital temperature readout. These instru-
ments were evaluated with an ice bath and a laboratory mercury
thermometer for accuracy and, in all cases, were accurate within
1°C. Sample pH was determined using either a Cole-Parmer Model
5986-00 or a Cole Parmer Model SL-5985-80 pH meter. The latter
instrument was rated to maintain calibration up to one year. It
was calibrated at the beginning of each sample trip and checked
daily. The instrument was recalibrated if the pH varied more
than 0.2 pH units from standards. The Model 5986-00 pH meter
was standardized for each sample.

The EC of samples was measured with a Lamotte Chemical
Model DA-1 Conductivity Bridge or a Cole-Parmer Model R-1481-50
Conductivity Meter. The Lamotte unit was standardized for each
analysis. The Cole-Parmer unit was standardized at the begin-
ning of each sample trip and checked twice daily thereafter.
This unit is rated to maintain calibration up to 90 days.

All data were recorded in the project field book or on site
matrix sheets.

At the completion of sample collection and processing,
samples were stored under security until shipped by bus to
Energy Laboratories in Billings, Montana. Although strict EPA
custody procedures were not followed due to the expense, at no
time were samples left out of sight, unattended or unsecured.

66

The water sampling program for the Swamp Gulch wetland
study began with initial sampling during the April 29 - May 2,
1987 period, following completion of the project installation.
The largest surface flows through the wetland were observed at
this time, but the flume on Swamp Gulch had not yet been instal-
led so only a one point estimate of Swamp Gulch flow was noted
(estimated at 20 to 30 gpm) . The second set of samples was
collected from July 28 to July 31, 1987, and included samples
from piezometers F-l-S and F-l-D which were installed on July
28, 1987. The third set of samples, collected during the
January 11 to January 13, 1988 period, consisted of a partial
set confined largely to the area of estimated maximum acid
drainage impact with several appropriate background sites. Low
flow conditions prevailed during both the second and third
sampling trips.

6.2 ANALYSES METHODS

All analyses were made utilizing standard EPA procedures
(Appendix A) . These are detailed in Methods for Chemical
Analysis of Water and Wastes (EPA 1983).

6.3 QUALITY CONTROL/ QUALITY ASSURANCE

The water quality program was designed to evaluate the
validity of these data. Field sampling included blind field
replicates, blind field water blanks and blind field standards.
Laboratory procedures included duplicates and spikes. These
data have been utilized to generate accuracy and precision
statements following standard procedures for evaluating data
from the EPA Contract Laboratories Program (Appendix A) . These
data are summarized in the following sections.

6.3.1 Accuracv

An evaluation of five Environmental Protection Agency (EPA)
standards submitted with samples indicate laboratory analyses
somewhat overestimated Ca, Mg, SO4 , HCO3, TDS and NO3 (Table 6-
1) . The laboratory did not recover any phosphate as ortho-
phosphate from standards containing 0.50 and 0.27 mg/L
phosphate. This problem is the result of running phosphate
analyses on the unpreserved aliquot which did not contain the
PO4 standard. The laboratory spikes for PO4 suggest an adequate
recovery (96 percent) for this parameter. Three EPA standards
were submitted for trace element analyses. These data generally
exhibit low standard deviation values and little bias (Table 6-
2). The variation between the number of samples was due to the
changing of the suite of analyses to be performed as the project
continued.

67

m

M
(0

•a

c

(0

•a

(U

c

c
o

0)

m
m

Xi

CO
4J

iH
D

to
<u
>-l

>1
o
ns

D

o
u

<

I

0)

t-H

m

Eh

OP

CO
(0
0)

c

0)

JJ

Ou

e
o
u

c
o

-H
0) -P

(0 o

CO

IM

O

W

0)

M

i-l

(U

cu

-Q

e

a

(0

D

CO

2

c

'V o
(0 -P

C
(0

CO Q

O C

•H

0) (0

cn-M
c u
m Q)

oi o

rH c

Q)
>

o

c

fO
Q) U

o\o

M
<U
JJ

e

(0
M
(0

SlCSlSSliS)CS<Sl(Sl

I II II

|CSl(S<9(Sl(SltStSll<SS<SlCS<SICSI IS

isi(s<a(si<siGi<siisi<SiS!<sd«S!i lo

lrHrHiHrHrHI.Hr-(|rHr-lr-lrHr-lrHI IrH

m •^ en is in
• • c • • • c c

(SS O'HtSSi o o

xxzxxxzz

CN '*

1 00 00 0) <U (U

0) <u

1 (U 0) (U

Q) Q) Q) \

1 Q)

1 • • c c c

c c

C C C

C C C 1

1 C

1 IS s o o o

o o

O O O

O O O 1

1 O

1 X X Z 2 2

2 2

12 2 2

2 2 2 1

1 2

u^Lnl/1u^lnlrlLnlnsl'*l^)Lnlnonisc^lf^^^^^lcsl^ocN^oc^l(SlC^c^J

r~ CN

r^ VD

r~- "* r- in CM LD

r-l t^ iH <n 00 s

in CN 00 CN •«a<

CM

ro

00 r- IS CM rH
CO rH CS CO CO

en IS CM rH CM
CN rH

I •^ rH I

V£)

r~~ CM VO ro CM rH

IS rH vo 00 in r-

r~ CN VD CN CN IS

I I '*
I I •

vor-crivooo(Tiv£)'a' >>ocNo^^oc^l rom v^incMvocNCN n
I I • • I II •

r-in'^CNr~rH'^fMlrH(TirHrHCOIVDS>tiH(rirHCN'!j'ml \ KO

CN rHIiHiH iH I rHim rHiH II

+ I+I + I+I+I+I+I+I I +I+I+I+I+II +1+11 +I+I+I + I+I + I I I +1

iSiccNiS'^CMCNoo iS'i'SCMro tsnisisinroisi^in is

• ••••••• I I • • II .

ooix>ror-rHr^rHr~-icN<T»iHror-~ir~rHisinQOCMcr\r^si icri
(s<siscrirHrHiscriicNrHSCJ>criicriis<s<ricyiis<ricri(si icn

rHrHiH iHiHrH IrHrHrH I iHiH iH iHII

•*0 coco

mcnro ocj-h OQCjXrHXiW'ovjoa(DX!C-rH

us:2N;coxcj[L4CiiZEHia cu<cn<c)uuucMDL.s:z

68

0) CP c

CO ri: N

>

OJ

J

0)

u
c

(U
T!

■H
IW

c
o
o

tfp

IS

en
+)

0)

•iH

a
w

>i

o

m

o

(0

c
o

(1)
m
m

w

4J

r-l

3
w
(U
>-<

>i
U
(0

D
U
U
<

CM
I

vo

X!
(0

Eh

m
m

c

4J
(U

rH

e
o
u

c
o

•H

en 4->
(0 u

CQ

'U

o

tfl

dJ

u

1—1

(U

Ch

X!

a

e

(0

3

w

Z

■D

O

w

-H

(0

-P

t3

(0

C

•H

rO

>

A->

(P

W

Q

>i

M-l 4J
O C

-H

(U (0

cn-p
c u
m Q)

DS o

rH C

ID

u
<u
>

o

c

rO

in in in LD LD

^y ^sf '^ ^sf ^^

in s s s
»* s (s is

d) <U CD cTi (U

c c c • c

o o o ca o

Z 2 Z X Z

(1) <U SI <u

c c • c

O O rH O

Z Z X 2

1

minininin i inr-irHrH

I f-H rH rH

rO <T> CM •^ CTi

00 r}' r~ cs f^

CTv SI r^ m
in r~ ■^ Ko

^^^•^•^(M I in^inro

vD CO in (Ti CO

CO yp Si S!

■<*'a''*ncN incsmcN
+I+I+I+I+I +I+I+I+I

c

I I
I I

SiSiSlSlSlSlSlSlSlSiCSSlSlSl
BlSlSlSlSlSlSlSlBlSlSlSlSl®

CM 'a' rH ■^ n

0) (U0<l)O>(U(U(USlSi(U(TirH0
CCCC .CCC • -C • 'C

O OO OBlOOOrHrHOSlrHO

2ZZZ xzzzx XZ xxz

III
I I I

I I li— IVOinrH'^inVOrHr^rHVO'^VD
I I I rH >-{ rH rH

Si ■^

oor^rH»;rror^'^roo^fM
incTivD'^oo-^Eir-csioo

V£>U500fOSlCMCMrO'«a'rH<MrH00r~-

rorOLnr^r~-oocMcT^vDoor~siooin

mmoOrHSlCMCMrHrOSlrHCMVO-^

o

z

CCC
3 3 3

Di oi a

-p -p -p
o o o

z z z

VO CM ■>!}< VO 00

ro CO CM cTi r~

Si Si Si Si (Ti

rH rH iH rH

■^ V£> C^^ 0^1

o^ rH VD cyi
o^ Si cr> (Ti

rH^-(Mo^Ln•«3*s^SlSlooooLnlnrH

rO00SlSl0O0OCMS!IX>CO(TiVD0OSl
SlO^SlSlS!O^(Sl?SO^<T^c^lSlC0S

■>*o
(0 cn (0 O U rH
OSZfcicoXUfa&Z

row

O Q U ac rH X!
E-< W OisC CO

WTDiHOSCUXiC-H
kCUOUUPmOjSZ

Q) cn C
CO <C N

>

J

o
c

(U

•H

MH
C
O

o

oV
Si
<Ti

69

6.3.2 Precision

Precision calculations were based on blind field replicates
and laboratory duplicates. The maximum range of uncertainty for
laboratory duplicates (at the 90% confidence level) was 7.69%
for HCO3 (Table 6-3) . Precision calculations based on blind
field replicates generally indicated good precision (at the 90%
confidence level) except for sodium, fluoride, manganese and
total suspended solids (TSS) (Table 6-4) . The high imprecision
for sodium is the result of values at or near the instrument
detection limit where a 1 to 2 ppm difference in analyses
produced large relative standard deviations. High uncertainty
values for fluoride and manganese were due to the small number
of valid samples (2 each) and the resulting high t values. The
precision of total suspended solids (TSS) analyses based on
blind field replicates was strongly influence by one sample
(from a total of 2) . The TSS sample and replicate sampled on
April 29, 1987, exhibited values of 16 and 8 ppm, respectively,
for this parameter which resulted in a 67 percent relative
standard deviation and a precision value of +298 percent.

6.3.3 Cross Contamination

Evaluation of the field water blanks indicated minimal
cross contamination (Table 6-5) . No chemical parameters were
above the instrument detection limit except bicarbonate for the
April 1987 and January 1988 sampling trips. Some bicarbonate is
usually present in all blanks due to carbon dioxide present in
the atmosphere, a portion of which dissolves in water forming
bicarbonate. One out of the two blanks for the July 1987 trip
exhibited a sulfate value of 40 mg/L, likely as the result of
sulfuric acid (H2SO4) used for preservation. The bottles may
have been mislabeled or the laboratory may have run sulfate on
the wrong aliquot. It is not felt that an actual contamination
by sulfate occurred as all other parameters associated with this
blank, with the exception of TDS, were below the detection
limits. Sulfate values associated with normal water samples for
the July 1987 sample trip do not show any associated jump in
concentration .

6.4 SURFACE WATER QUALITY

6.4.1 Swamp Gulch Above the Carbonate Mine

The water quality of Swamp Gulch above the mine is good
and meets most criteria for drinking water (USPHS 1962; EPA
1978, 1979), Livestock (NAS 1974), aquatic life (EPA 1983; NAS
1974) and irrigation (NAS 1974) . The only exception was an
enormously high (0.21 mg/L) manganese value obtained during the
January 1988 sample collection. The April and July, 1987 values
for manganese were both below the 0.02 mg/L instrument detection

70

>
0)

U
C
0)

-r-l

c
o
o

oV>
(S

o^
0)

4J

4J
(0

w

0)

-p

(0

o

en
c

I
o
u

(U

o\o

rH

ro

a

•

4J e

CN

c ro

rH

Q) w

+ 1

e

II

D M

C

U O

o

4J H-l

-rH

CD

W

C J

-iH

M \

en

a>

w E

M

03

a

(P cs

c

«k

'V

,-^

-P c

>.

(U (0

rH

w

(U

rH

>

0) --

•rH

M '—

+)

a J

O

S D

(U

(0 l-H

Q,

m ^

CO

(U

(U -P

u

C -H

o g

Q)

•r^

4J

CPJ

(0

C

U

•■H C

■r-t

-P O

rH

(0 -rH

Q.

C -P

D

-H U

T!

e <u

■r^ 4J

T5

rH 0)

C

W Q

m

Q)

JJ

(0

O

•H

rH

a

3

■u

6V0

o\o oV

O J

J

S

S CS

C Q

Q

•

• J •

J M

l-H

cs

J

IS Q S

■'D

11

Q

II M II

J l-H

Q

n

Q Q

Q rH

i-H

&

D4 a.

l-l rH

rH

a:

a; rH OS

s: fo

(0

rH

rH u

<U -P

•.

rH

^< ^Q

> o ^

*k

J

to

J J KH

O JQ W

to

Q

Q ^Q

JQ <U

Q)

M

IW

M CO l-H

(0 «.rH

rH

CO

Q)

to Du

a

OJ

rH CO

<U <U g

g

<U

rH

(u a 0) 0)

rH rH rO

(0

rH

a

rH g rH rH

Qj &. CD

tn

a

g

a (C5 Dj Dj

g g

g

(t5

g to g g

(0 n3 c

c

(C5

CO

(0 (0 (C5

to CO <u

<u

to

to 4J CO CO

>

>

<u

x:

0) o <u

0)

Q)

>

(U en 0) CD

C S rH

rH

•C

•H

C -H c c

O Eh W

W

O

Co

o w o o

0\P

SI

•

11 u3

Q a

PU M

a;

(U (U

a

Qi

g

g

(0

(0

to

to

(U

<u

c

c

o o

a

>i

M
O
4J
(0

o
x>

(0

rH

CO

to
0)

c
<u
-p

r-H

a
g
o
o

dip <AP o\P oVP <A0 oNO cAO o\P I I I oV>

viJinLD r^ini^inrHi i lis

I I I rH

oV> I o\o oVO o\o

in I in n 00

in I in rH ro
I

I <*P o\P o\o cfP oV>

I CN o ts in ts

I 00 in B> CN CN

I rH

I 6\P I oW) <W> o\P c«P cW> o\o I

I ro I tsi in in 00 in is I

I rH I CT\ ■^ '3' rH 'i' ® I

II rH I

C
O

•a

0)

to

(0
X!

c
o

■H
CO

•rH

o

0)

c
o

•rH

CO
-H

o

OJ
M

o>P oV>

o\P o\o AO cW> oVO

<N !S

s n vD is r~

• • o\P

ro •^ vo

is is r- is rH

CN

+I+I+I

+1+1+1+1+1

I o\o o\o I o\P I CAP I cAO o\P cW> o\P I

iro CN lin lis l'*iS'<3'iSI

I • •l«l»l»«»»l

ICM iHI-^liSliSiSiSiSI

I III I

I +1 +1 I +i I +1 I +1 + 1+1 + 1 I

I CAP af> OP <if> OP

\ (Tt iS 00 tS CM

I • • • • •

I is is is in n

I

I +1 +I+I+I+I

I

Xi

m

Eh

M
0)
4J
0)

g

m

M
(0

04

(0

u

S 2

■•S-O
O U -H
!^ CO X U Cm

M Ou

ro

CO CO O

O

rHMTDODMtUXiC-HCUCnXlMC

a: o Q to o 2

Z

<<OUUUCMaiS2co<cocots3

D4W Eh Eh Q U

71

Table 6-4.

Precision, based on field replicates, at the 90%
confidence level.

Parameter

Precision

Completeness

Comments

Ca

±0.0%

100%

Mg

± 4.4%

100%

—

Na

±46.0%

100%

—

K

+ 0.0%

50%

—

SO4

± 1.2%

75%

—

HCO3

±6.0%

100%

-

CI

±0.0%

50%

-

F

±81.0%

50%

—

P

NC

0%

All values <IDL

NO3

±13.6%

75%

-

Al

NC

0%

All values <IDL

Cd

NC

0%

All values <IDL

Fe

5.6%

75%

—

Pb

NC

0%

All values <IDL

Mn

27%

50%

-

PH

2.2%

100%

..

EC

2.5%

100%

—

TDS

8.0%

100%

—

TSS

298.0%

67%

NC

Not calculated

limit. The high sample value may have been due to the amount of
sediment disturbed during sampling (ice removal) . Swamp Gulch
water above the mine has exhibited little variation during the
period of study and is characterized by a calcium bicarbonate
water type. Calcium, magnesium, bicarbonate and sulfate percen-
tages in eguivalents/L are 54.6, 37.9, 80.4 and 19.6, respec-
tively. Slight increases were noted for both sodium (2 mg/L
increase) and sulfate (7 mg/L increase) during the January 1988
sample period and are of little significance.

6.4.2 Swamp Gulch Below the Carbonate Mine

Swamp Gulch water below the mine consistently exceeded
drinking water criteria (USPHS 1962, EPA 1978, EPA 1983) for
aluminum, cadmium, iron and manganese. The water was charac-
terized by an iron sulfate water type in which iron and sulfate
comprised 31.1 and 99+ percent of cations and anions,
respectively, assuming there was an even balance between ferrous
and ferric iron forms. Other major cations were calcium (27.6
percent), magnesium (18.7 percent), aluminum (16.9 percent) ,

72

M
C
(0

1—1

TD

rH

0)

fa

in
I

V£>

0)

r-l
X!
(0

E-«

r- ■*

iH

in

CX3 r-4

ca iH cri

® ro rH CM

r-H m rH tS

r-i

\ 1

<S rH O rH

rH C9 ts es csi

C9 <Si (S> C9

<S fH

rH IH

f • • •

• » • • •

• • • •

• 1

ro M

rHrHiHiH(Slr-(r>JrO

cs tn ca ca «s

CS <Si S CSl ts

O tS <Sl s

es es

\ D

V V V V •si' 1—1

V V

V V V V V

V V V V

V V

r~ CO

00

CN

rH

CM

r-i
I

IW
M
D
<yi CO

CN

in

kO

00

c
o

•H
4J
(0

c

Cn|

•H
Q) (Q
-P 0)
(0 D

a
<u

rH

a
e

(0
CO

s rH in

(S rH is I— I

C

rH D in in in

iSCOOSCNlSliHrHfOr-HSCNSlrH

V V V V V V

r^J rH

V

■«* f-i

in O tS Si

V V

is (SI CS)

V V V

O cs <s

(S«S6)C9<S><SlCS<SI

vvvvvvvv

S iH rH

rH «a cs

is CO rH CN

rH is is iS iS

<-\ r^ <-{ r-{ r^ C>i CO
V V V V V _H

CN in is (s is

V V

is is is tS iS
V V V V V

in

rH CO rH iS

is is is is

• • • •

is IS is is

V V V V

CO
iS

is is

V V

r-t t-{ r-t r-{ r-t rH" CN rH I

is rH m

® rH is iS

insi^is

!S CO rH CN

is ts is cs

??^^^

CO
■^ O CO CO CO

(omcnOrHOQcQCja: o

ti<iZUSC0UXE-<E-"Waix<CU2

rHT>Q>4aCtQ0 3■r^^D^M(UCW

f«coix)ii2'«:ouzto<duwNto

73

manganese (3.3 percent) and sodium (2.3 percent). This water
also exhibited an increase in sulfate and iron in the January
1988 sampling. The increase in iron was especially notable
(from 18.8 to 48.4 mg/L) and likely represents the combined
result of several factors: 1) low flow which has less dilution;
2) the presence of an ice cover, which may have restricted
oxygen availability (and subsequent precipitation of iron as
iron-hydroxide) and, 3) somewhat lower pH which would allow more
iron in solution. This observation suggests physio-chemical
precipitation of iron may be important in this system and that
this process starts before the AMD is discharged into the
wetland.

6.4.3 Blackfoot River

The general water type of the Blackfoot River at Pass
Creek, located upstream of the wetland study area, was consis-
tent throughout all three sample periods (Figure 3-1) . The
dominant water type, expressed in milliequivalent per liter, was
a calcium bicarbonate in which calcium and bicarbonate composed
53 percent of both cations and anions respectively. Magnesium
and sulfate represented 44 and 47 percent of cations and anions.
Little change occurred in these percentages during high or low
flow. The water type for the Blackfoot River at the Meadow
Creek road bridge, located downstream of the wetland study area,
exhibited a marked change during low flow. The April and July
1987 water samples exhibited a similar cation-anion balance to
that found at the Pass Creek site (54, 43, 52 and 48 percent for
calcium, magnesium, bicarbonate and sulfate, respectively) .
However, during the January 1988 sample period, a slight
decrease in calcium (to 51 percent), an increase in sodium (from
2.1 to 4.6 percent) and a gross increase in sulfate (69.7
percent) at the expense of bicarbonate (30.3 percent) was
evident. The ramifications of this will be discussed later.

6.4.4 Background Statistical Comparisons

Two general comparisons have been evaluated: 1) the Swamp
Gulch water quality above and below the Carbonate Mine and 2)
the Blackfoot River above and below the wetland discharge area.
The Swamp Gulch drainage exhibited significant to very highly
significant increases in the concentrations of SO4, Al, Mn, Zn,
Cd, Fe, Pb and Cu when compared using a paired "t" test. No
significant increases were observed for As, Cu, Co, Sb, Se and
Ag at the given detection limits (Appendix B-3) . The "t" test
comparison of the Blackfoot River at Pass Creek and at the
Meadow Creek Road Bridge indicated no significant differences.
However, the limited number of samples (3) for each site did
suggest the Meadow Creek Road Bridge sample site was slightly
higher in iron, manganese and sulfate. The slight increases in
parameters in Blackfoot River samples below the Swamp Gulch
wetland discharge point as opposed to those above the wetland

74

are great enough to account for approximately one half of the
iron and all of the manganese contributed by the Swamp Gulch
acid mine drainage. Zinc is the only other trace metal present
in the Blackfoot River at notable levels and was apparently
diluted in the river reach adjacent to the wetland. Both the
April and July sample sets show the highest zinc levels in the
upper two river sample sites with each downstream sample site
showing progressively lower zinc concentrations. This effect
could not be determined during January 1988 due to the reduced
number of samples taken at that time. These data strongly
suggest the source of the zinc is either Paymaster Gulch, the
main fork of the Blackfoot River immediately above the study
area or some discharge directly into the river or wetland . The
latter may be some flowing abandoned drill holes, one of which
was measured for EC and pH (352 umbos/ cm and 3.97, respectively)
on April 29, 1987.

6.4.5 Wetland Surface Water

In addition to the Blackfoot River and Swamp Gulch samples,
surface water samples were obtained from six surface sites
within the wetland during the July sample trip. These were
obtained at sites AAA-3, BC-1, C-1.5, C-3, CD-I and F-1. Water
quality parameters at site AAA-3 were very similar to the
Blackfoot River at Pass Creek samples and exhibited a calcium
bicarbonate water type and low levels of all metals (iron at
0.07 mg/L) . This would be expected from the observed drainage
system. Parameters at site C-3 were also similar to the river
samples but iron and manganese levels were elevated to 6.94 and
0.24 mg/L, respectively. This suggests some wetland drainage is
mixing with the diverted river water at this point. The pond
sample at BC-1 was intermediate in composition between values
observed at C-3 and the Swamp Gulch drainage which suggests this
pond is being partially recharged from the upgradient, non-
polluted wetland area. The water quality at sites C-1.5, CD-I
and F-1 was similar to the Swamp Gulch AMD discharge. The water
type at C-1.5 was an iron sulfate and at sites CD-I and F-1 it
was a calcium sulfate. Sulfate was the only major anion in the
latter samples. High levels of sulfate, iron and manganese
would be expected at sites C-1.5 and CD-I, located 37.5m (123
ft) and 32.9m (108 ft), respectively, from the Swamp Gulch
discharge point. However, the high levels observed at site F-1
raise serious questions as to the actual effectiveness of this
wetland in ameliorating AMD water. The process may be one of
partial dilution which cannot be accurately accessed with the
present instrumentation. It should also be noted that these
conclusions are based on one set of surface samples which could
be subject to analytical or sampling errors which are not
apparent .

75

6.5 GROUND WATER QUALITY

The ground water of the wetland research area has been
characterized by samples from two zones, shallow and deep. The
former is generally from 0.6 to 1.2 m (2 to 4 ft) and the latter
varied up to a maximum of 4.9 m (16 ft). The shallow zone
screened interval generally spanned portions of both the
surf icial organic layer (acrotelm) and the underlying muck layer
(catotelm) . The deep wells bottomed in the gravel underlying
the muck layer. However, extreme difficulty was encountered
trying to keep the deep bore holes open long enough to place the
screen within the gravel zone and as a consequence, most deep
wells have a significant portion of their screened interval
within the muck zone. All wells were sandpacked and, on the
deep wells, the sand pack extended into the partially caved
gravel portion of the bores. The water quality data strongly
suggest most water withdrawn from the deep wells is from .the
gravel aquifer. The following sections describe water quality
parameters of the shallow and deep zones. An evaluation of the
two wells in the acid drainage seep have been evaluated
separately in section 6.5.3.

6.5.1 Shallow Ground-water System

The shallow ground-water system (0-1. 2m, 0-4ft) is
dominated by iron-calcium sulfate and calcium-magnesium
bicarbonate water types. The former is present throughout the
main Swamp Gulch AMD impact area, the latter is generally
confined to the southwestern portions of the study area (Figure
6""1) • A calcium-magnesium sulfate water type is conspicuously
present at two sites near the Blackfoot River (C-5 and F-3) .
The only other site at which the calcium-magnesium sulfate water
type was dominant is the site AA-2. It is noteworthy that
background site AAA-3 exhibits the same water type found in the
AMD impact area although overall concentrations are lower at
site AAA-3. The zone dominated by the bicarbonate anion is
confined to the area which apparently receives water from the
Blackfoot River via channels flowing past sites c-3 and E-3.
The single sample from site F-1 was the only analyses in which
bicarbonate was the dominant anion and which was not affected by
Blackfoot River water.

A plot of mean dissolved iron values for all well sites
clearly indicates that the shallow ground-water system is
enriched in iron as the result of AMD (Figure 6-2). Background
means for iron ranged from <0.03 to 27 mg/L with the very low
values (<0.03 and 0.25 mg/L) associated with sites E-3 and F-3,
respectively. Sites E-3 and F-3 are apparently recharged
directly from the Blackfoot River. Samples from shallow well
sites near AMD discharge point are 2 to 3 times above both the
background shallow ground-water concentrations and the inflow
AMD concentrations. The highest mean (73 mg/L) was found for

76

CARBONATE MINE/,

A WELL SITE

O SURFACE WATER SAMPLE

SITE
D SURVEY CONTROL STATION
:■ FLUME
O PRECIPITATION GAGE

DOMINANT CATION
^S CALCIUM
^ IRON

DOMINANT ANION
^M BICARBONATE
Em SULFATE

50
METERS

Figure 6-1. Shallow ground-water types for the wetland system.

site C-1, the closest site to the AMD discharge i oint.
Dissolved iron concentrations fall abruptly e.. ard towards
site B-2, at least partially the result of lowei ground-water
gradients in this direction. Concentrations fall more gently
towards the southwest, which suggests more AMD Hows in this
direction which is consistent with topography and piezometric
maps. Iron concentrations are apparently reduced 0.65 to 0.98
percent per meter (0.2 to 0.3 percent per foot) away from the
maximum concentration observed at site C~l in this ^^etland. It
is apparent that much of the iron in the AMD hp.s been deposited
in the 3.9 ha (9.6 ac) area surrounding the Ai€D discharge point.

Isolines of dissolved manganese concentrations in the
shallow ground-water system show the highest concentrations are
associated with sites CD-I and D-1 (Figure 6-3). The zone of
elevated manganese concentrations is generally confined to the
area between the F-2 to C-2 line and Highway 200 or roughly 1 ha

77

CARBONATE MINEX.' '/

'/k/ II

A WELL SITES

O SURFACE WATER SAMPLE

SITE
O SURVEY CONTROL STATION
•n FLUME

,V.'' O PRECIPITATION GAGE
y ISOLINES OF IRON (ra«/ L)

SEC 20, TI5N,R6W

LEWIS AND CLARK CO .MONTANA

100

FEET
200

Figure 6-2. Iron concentrations for the wetland shallow ground-
water system.

(2.5 ac) . Concentrations in this zone are 3 to 11 times
background concentrations. High manganese concentrations have
also been observed at sites C-5 and E-3 but it is doubtful that
these sites are impacted from the Swamp Gulch AMD.

Manganese is apparently attenuated to a much lesser degree
than is iron in this wetland. The manganese concentration at
site F-1, near the wetland outflow, are still about 4 times
above background levels. These results are consistent with
other studies which have shown Sphagnum wetlands are more
effective in removing iron than they are for removing manganese
from AMD (Wieder and Lang 1986, Burris et al. 1984).

78

CARBONATE MINEX-
> ^ If

A WELL SITES

O SURFACE WATER SAMPLE

SITE
a SURVEY CONTROL STATION
H FLUME

n PRECIPITATION GAGE
-— SOLINES OF MANGANESE (mg/L)

SEC 20, TI5N,R6W
LEWIS AND CLARK CO .MONTANA

FEET
200

Figure 6-3.

Manganese concentrations for the ^';pt and shallow
ground-water system.

Isolines of sulfate concentrations follow a pattern very
similar to that found for the bromide tracer (Figure 6-4) . From
a high concentration zone in the D-1 to C-1 area, levels
decrease gently towards sites E-1 and B-2, apparently along the
preferential flow paths. Sharp concentration reductions over
short distances occur at the margins of the flow paths.
Concentrations are typically 3 times above background levels in
the 0.7 ha (1.8 ac) AMD impacted zone. The sulfate isolines
suggest a low level of AMD impact does occur in the area defined
by site B-2, A-3, C-3 and C-2. This is supported by the
increase in iron and manganese concentrations observed in the

79

CARBONATE MINEX-

A WELL SITES

O SURFACE WATER SAMPLE

SITE
D SURVEY CONTROL STATION
H FLUME

O PRECIPITATION GAGE
— ISOLINES OF SO4 CONCENTRATIONS
ftvg/L)

SEC20, TI5N,R6W
LEWIS AND CLARK CO. .MONTANA

FEET
200

Figure 6-4.

Sulfate concentrations for the wetland shallow
ground-water system.

surface water sample taken at site C-3 (section 6.4.5). This
wetland is apparently effective in attenuating shallow ground-
water sulfate concentrations to levels approaching background
values.

Considerable variation was apparent in TDS concentrations
within the wetland. Mean values ranged from a low of 119 mg/L
at site D-2 to a high of 403 mg/L at site D-1. Higher TDS con-
centrations were generally confined to a 0.6 ha (1.6 ac) area in
the same general vicinity where elevated manganese
concentrations were found (Figure 6-5) . Total dissolved solids
concentrations are reduced from values about 2 times above

80

CARBONATE MINE/>

' ^' 1 1 I,

I -Nv 'I

A WELL SITE

O SURFACE WATER SAMPLE

SITE
D SURVEY CONTROL STATION
yt FLUME
O PRECIPITATION GAGE

ISOLINES OF TOTAL DISSOLVED

SOLIDS (m«/L)

SECZO, TI5N,R6W
LEWIS AND CLARK CO .MONTANA

FEET

200 400

Figure 6-5.

Total dissolved solids concentratio.
wetland shallow ground-water systr"

for the

background values to levels similar to background concentrations
within the wetland area.

Concentrations of aluminum, cadmium, lead and zinc were
generally low throughout the wetland. The maximum aluminum
value was 2.8 mg/L in well BC-l-S for April, 1987 saiaple trip.
Aluminum was detected in 18 other samples from 7 other wells
which all but one sample exhibited values lass than 1 mg/L.
Only two samples contained cadmium in excess of 0.01 mg/L, the
recoxamended criteria for potable and irrigation water supplies.
The two samples were obtained from sites BC-1-2 (0.032 mg/L) and
BC-l-S (0.044 mg/L) both of which are in close proximity to the
Swamp Gulch AMD discharge point. These two wells were also the

81

source of the only two samples in which detectable levels of
lead were found (both 0.03 mg/L) . All zinc concentrations were
well below applicable criteria for all common uses. The maximum
zinc concentration in the impacted area was 0.49 mg/L from well
D-l-S for the April, 1987 sample trip. The maximum zinc value
observed in the wetland was 1.18 for site C-5-S, but this site
is not likely affected by the Swamp Gulch AMD.

Numerous other trace metals were analyzed in the initial
April 1987 samples. These included arsenic, chromium, cobalt,
copper, nickel, antimony, tin, selenium and silver. No shallow
wells had values above detection limits for arsenic, chromium,
copper, nickel, selenium and silver. Only site C-5-S had
detectable cobalt at 0.09 mg/L. Antimony concentrations at or
above the 0.01 detection limit were found at 14 sites. The
highest antimony levels found were 0.03 mg/L at BC-1-2, D-l-S
and C-5-S. These data suggest there is some enrichment of
antimony due to the Swamp Gulch AMD, but the effect is minor.

In summary, the levels of aluminum, cadmium, lead, zinc,
sulfate and TDS observed in the wetland shallow ground water
system should have little overall adverse impact on vegetation
or wildlife. However, the high iron and manganese
concentrations observed exceed all applicable water use criteria
and can be expected to adversely impact all species not adapted
to high levels of these metals. The elevated cadmium
concentrations observed near the AMD discharge point may also be
of concern and may be detrimental to wildlife through
bioaccumulation (Section 9) .

6.5.2 Deep Ground-water Svstem

The overall water type for the deep ground-water system
exhibited markedly less variation than the shallow ground-water
system. A calcium-magnesiiom sulfate water type was dominant in
all sites except AAA-3, B-2 and C-2, which were characterized by
an iron-calcium sulfate water type, and sites D-1 and F-1 in
which calcium-iron bicarbonate and calcium-magnesium bicarbonate
dominated, respectively. Site C-3-D exhibited a sodium-iron
sulfate water type which was likely the result of problems with
bentonite contamination during construction.

Isoline map evaluations for concentrations of iron, man-
ganese, sulfate and TDS suggest the AMD has a much more limited
impact on this aquifer system. Only iron and manganese show
increases in concentrations associated with AMD and the effect
is generally subdued as compared to the shallow aquifer system
(Figures 6-6 and 6-7). The areas in which iron and manganese
levels are clearly elevated above background concentrations
roughly coincide with the impact area observed for elevated TDS
levels in the shallow ground-water system and cover approxi-

82

CARBONATE MINE/>

/I f/ "

'tW ' II

A WELL SITES

O SURFACE WATER SAMPLE

SITE
D SURVEY CONTROL STATION
)■ FLUME

O PRECIPITATION GAGE
— ISOLINES OF IRON (mg/L)

SEC20, TI5N,R6W

LEWIS AND CLARK CO, MONTANA

FEET
200

Figure 6-6.

Iron concentrations for the wetland deep ground-
water system.

mately 1.5 ha (3.6 ac) and 1.7 ha (4.3 ac) , respectively.
Concentrations and attenuation gradients in the deep ground-
water system for both iron and manganese are considerably less
than in the shallow ground-water system which suggests a very
limited AMD input to the deep system. This conclusion is
supported by sulfate and TDS concentrations which were clearly
elevated above background levels only at wf^ll C-l-D.

No deep wells exhibited lead values above the 0.01 mg/L
detection limit. The highest wetland aluminum values from deep
wells were found at sites AA-2 and B-2 with maximums of 1.7 and

83

CARBONATE MINE/^

A WELL SITE

O SURFACE WATER SAMPLE

SITE
D SURVEY CONTROL STATION

•m FLUME

PRECIPITATION GAGE
iSOLINES OF MAN6ANESE(mg/L)

SEC 20, T15N.R6W
LEWIS AND CLARK CO .MONTANA

FEET
200

Figure 6-7.

Manganese concentrations for the wetland deep
ground-water system.

3.0 mg/L, respectively. No other sites were found with aluminum
values greater than 1.0 mg/L and many were below the 0.1 mg/L
detection limit. Detectable cadmium concentrations were found
at sites AAA-3, AA-2, B-2, BC-1, C-1 and C-2 . Maximum values
were consistently found for site AA-2 with a mean and maximum of
0.007 and 0.012 mg/L, respectively. This was the only site
exhibiting levels above 0.01 mg/L, a recommended criteria level
for potable and irrigation water. As with shallow ground -water
samples, arsenic, chromium, cobalt, copper, nickel, antimony,
selenium and silver were determined on the April 1987 sample
set. Arsenic was detected at only two sites, AAA-3 at 0.012
mg/L and AA-2 at 0.005 mg/L. Cobalt was detected only at site
C-5 which contained 0.02 mg/L. Copper was present in detectable
quantities only at sites AA-2 (0.02 mg/L) and B-2 (0.07 mg/L).
The maximum antimony concentration found (0.03 mg/L) was for

84

site C-1. Sites C-5 and D-1 both exhibited antiaony values of
0.02 mg/L and all other sites were at or below the 0.01 mg/L
detection limit for antimony. No detectPble amounts were found
for chromium, nickel, selenium and silver in the deep wells.
With the exception of antimony, the secondary trace metals have
no apparent relationship with the AMD.

It is possible that some of the observed effect of elevated
concentrations of iron and manganese in deep wells is due to
subsurface seepage from old mine workings that migrates below
the highway in fractured bedrock rather than vertical perco-
lation through the wetland muck layer. There is insufficient
data at present to evaluate this factor.

6.5.3 Acid Mine Drainage Seepage Wells

The water quality of the shallow and deep "seep" wells is
important as it represents the actual water quality that is
likely emanating from old mine workings prior to dilution by
surface water (Figure 5-9 and 5-10) . The general water type for
this site is a calcium sulfate with only small to moderate
differences noted for most major parameters between deep and
shallow samples.

This water was characterized by low pH (3.0) and moderate
to high levels of most trace metals. Mean values for TDS, EC
and sulfate are 1838 mg/L, 1924 umhos/cm and 1228 mg/L,
respectively. Parameters which did exhibit notable differences
between shallow and deep samples were aluminum, iron, manganese,
zinc and several minor metals. Aluminum and zinc were
consistently higher in shallow samples, 70.4 mg/L versus 3.6
mg/L and 31.9 mg/L versus 8.5 mg/L, respectively. This trend
was also noted for arsenic, cadmium, copper, nickel and
antimony. Both iron and manganese were higher in the deep well
than in the shallow well (57.5 versus 5.0 mg/L a i 65.1 versus
40 mg/L, respectively) . Silver was also apparently present at
slightly higher concentrations in the deeper V'^" " at this site.
It is readily apparent that many of the trace el-ment
concentrations present in water at these two wells are quickly
diluted by high quality upper Swamp Gulch water before entering
the wetland. This process likely eliminates most of the
toxicity problems that might otherwise occur. The high levels
of many metals, sulfate and fluoride would make this water
unsuitable for any common use including potable supplies,
livestock, irrigation and aquatic life usage.

6.6 SUMMARY

Of the parameters analyzed (Appendix B) , only aluminum,
cadmium, copper, iron, lead, manganese, zinc and sulfate were
significantly elevated in the Swamp Gulch watershed drainage due

85

to AMD. Concentrations of these parameters in the wetland
shallow ground-water system are elevated to levels above the
concentrations found in the AMD near the AMD inflow point and
decrease both laterally and vertically from this point. The
only catotelm area notably affected by AMD is in the immediate
vicinity (30 m - 100 ft) of the AMD discharge point and, in
general, this stratigraphic unit is only minimally impacted by
the AMD.

Minor changes in the water quality of the Blackfoot River
above and below the wetland discharge point suggest that
manganese and possibly other several metals may not be
effectively ameliorated within the wetland. However, the Swamp
Gulch wetland area is at least partially effective in removing
iron from the AMD. These data suggest about half of the 21 mg/L
weighted mean iron concentration in the mean 45m3 daily AMD
inflow may be precipitated within the 3.9 ha (9.6 acre) iron
impacted acrotelm area (Section 8). These data suggest some
means of physio-chemical precipitation of iron should be
incorporated in artificial wetland construction. This system
should provide aeration of the AMD to enhance iron
precipitation, possibly through the use of waterfalls. Since
optimum treatment of AMD pH and sulfate by microbial action
(Section 10) requires reduced conditions, the aeration stage
should be constructed near the discharge end of the artificial
wetland, under the assumption that it will be much easier to add
oxygen to the AMD than it would be to remove it.

86

7.0 DISSOLVED METAL LOADING OF A NATORAT. ..iTLAND

An estimation of the AMD metal load supplied to the Swamp
Gulch wetland through the April 1987 - March 1988 year has been
made based on water sample concentrations and estimated flow
volumes. A comparison of this loading with observed metal
masses in the acrotelm (Section 8) permit an estimation of past
wetland efficiency in amelioration of AMD.

7 . 1 METHODS

The calculation of metal loading into the wetland was
hampered by the short duration of runoff records necessitating
estimations be made for a number of factors. The assumptions
that have been made are:

o the Swamp Gulch input represents all metal loading to
the study area;

o the three sample dates, April 29, July 29/ 1987 and
January 11, 1988, are representative of all flows for
the periods April 1 through June 30, 1987, July 1
through November 16, 1987 and November 17, 1987 through
March 31, 1988, respectively;

o the observed flow on April 29, 1987 is representative of
the entire month; and

o that the mean daily flow rate for the period May 1
through June 15, 1987, was equal to the value for the
period June 15 through June 30, 1987.

The latter assumption is based on the Lincoln Ra: jer Station
precipitation which was similar for the two montl.s at 4.1 cm and
3.4 cm (1.62 and 1.33 inches) for May and June, t aspectively.
The Lincoln Ranger Station records also indicate 3.99 cm (0.39
in) of precipitation fell during April, 1987, and that
precipitation for December, 1986, January and February 1987 were
far below normal. Casual field observations during April and
May indicated little snow melt runoff occurred and thus the
assumption was made that the April 1987 flow estimate was
reasonable for the entire month. The time intervals were
determined on the general flow regime and the time of sampling.
These are, in general, spring flow consisting of some snow melt
and precipitation; summer flow consisting of base flow and
precipitation runoff; and, winter low flow (base flow) when no
runoff and little precipitation input occur and during which
freezing conditions persist.

87

The wetland metal loading figures are thus basou on a
limited record for a period in which abnormally low precipi-
tation conditions prevailed. Extrapolation of these data to
cover the annual period of April 1987 through March 1988 may
involve appreciable error and further extrapolation to other
years is not generally recommended.

7.2 RESULTS AND DISCUSSION

The approximate weights (mass) of dissolved metals dis-
charged via Swamp Gulch into the wetland from April 1987 through
March 1988 varied from 338 kg (744 lbs) for iron to values less
than 100 grams (3.5 oz) for arsenic and silver (Table 7-1).
High values for aluminum (107 kg) , manganese (63 kg) , zinc (44
kg) and copper (12 kg) are also evident. All other metal
loading values were less than 1 kg and of these, only cadmium
with its ability to be readily bioaccumulated, may be of
additional concern. The implications and fate of these metals
is discussed further in section 8.

Table 7-1.

Total dissolved metal loading (kg) on the Swamp Gulch
wetland, April 1987 through March 1988.

Apr. 1- June 30
Flow Volume fm-^) 7.940

PERIOD ■

July l-Nov.16
7.360

Nov.l7-March Annual
1.110

Parameter

Al

As

Cd

Cr

Co

Cu

Fe

Pb

Mn

Ni

Sb

Se

Ag

Zn

48

52

6.1

107

<0.04

<0.04

<0.01

<0.09

0.302

0.213

0.027

0.542

<0.159

- —

<0.079

<0.074

<0.011

<0.164

6.274

4.931

410

11.6

46

138

5'

338

.278*

0.221

0.044

.543*

30.8

28.8

3.75

63.4

0.32

0.29

0.03

0.64

0.08

<.07

<0.06

0.08

<0.04

.

<0.04

<0.04

<.01

<0.09

22.6

17.9

2.9 est.

43.6

*Estimate

Annual Total 565 kg

88

8.0 WETLAND GEOCHEMISTRY AND THE CONTROL OF ACID HINE DRAINAGE

The Swamp Gulch wetland area core samples have been
analyzed for numerous chemical and physical parameters (Appendix
C) . These analyses were made to determine 1) the extent of past
metal loading; 2) the distribution of various parameters with
depth and stratigraphic changes; and 3) the basis for comparison
of elemental concentrations observed in vegetation growing on
this site.

8.1 METHODS

Samples for geochemical evaluation were obtained during
construction of well and auger hole installations. Core samples
were obtained by methods outlined in section 5.2.1. These
samples were placed in plastic sample bags and shipped to Energy
Laboratories in Billings, Montana. Samples for metal analyses
were digested using hydrogen peroxide, EPA Method 3050 (EPA
1982) . The very high iron content of most wetland soil samples
necessitated special corrections to eliminate interference in
samples analyzed by inductively coupled argon plasma spectro-
scopy (ICP) . These corrections were calculated at the 25
percent iron content and, as a result, were insufficient for the
very high iron content found at site C-1. The apparent
concentrations reported for several metals are depressed at site
C-1 for this reason but the laboratory reported values are still
included in Appendix C. Another result of the iron interference
was the high tin and antimony values (positive interference)
reported in the initial screening samples upon which the
selection of parameters to be analyzed was based. As a
consequence, tin and antimony were run on samples from several
sites and, following interference corrections, all values were
below the instrument detection limit. Two Natior\l Bureau of
Standards (NBS 1982) samples of river sediment, t ;andard 1645,
were run with wetland samples. Recoveries were vithin 10
percent for all parameters except aluminum, arst- t Ic and iron.
The recoveries were: aluminum, 28%; arsenic, 79%; cadmium, 90%;
chromium, 97%; copper, 92%; iron, 79%; mercury, 100%; manganese,
92%; nickel, 90%; lead, 95%; and zinc 94%. The low recoveries
for aluminum make interpretation of these data difficult and the
findings should, therefore, be considered only relative. No
adjustment of these data has been made since, with the exception
of aluminum, most elements were within or very near the reported
range of estimated uncertainties for the standard. Chloride,
fluoride, bicarbonate and sulfate were analyzed from a saturated
paste extract. The cation exchange capacity was determined by
Method 8-4, American Society of Agronomy Monograph 9 (1965) for
acid soils. Organic carbon content was calculated from loss on
ignition at 550°C for 2 hours. The data base consists of single
samples for each selected interval at each site, generally based

89

on field observed stratigraphic changes. Fourteen laboratory
duplicates and the two standards were analyzed for most
parameters.

Selected metal deposition in the wetland due to the Swamp
Gulch discharge has been estimated utilizing 1) isopachs of the
acrotelm (Figure 5-7), 2) isoline maps of chemical concentra-
tions of specific metals, and 3) summation of volumes multiplied
times concentrations multiplied times bulk density. Background
concentrations were subtracted before isoline maps were con-
structed for this purpose.

8.2 RESULTS AND DISCUSSION

In general , metal and other chemical parameters are more
varied than similar data for ground-water which is likely due to
a more heterogeneous nature of the soils. Occasional
anomalously high or low values for several elements have been
noted and may be due to sampling or laboratory ICP interference.
Samples obviously in error have not been used in the following
interpretation. Total masses of metals deposited in the
acrotelm due to AMD have been calculated based on the 3.9 ha
(9.6 ac) area found to be impacted by iron deposition. The
total volume of the impacted acrotelm area is approximately
23,000 cubic meters (30,000 yd^) .

Most trace metal concentrations in the acrotelm show
similar areal distribution. Aluminum, cadmium, copper, lead,
manganese and zinc all exhibit the highest concentrations in the
area immediately south of Montana Highway 200 and generally west
of site C-1. Iron distribution more closely follows the
preferential ground-water flow paths (Section 5) . The following
paragraphs describe the distributions of specific metals in the
wetland soil materials.

8.2.1 Areal Distribution of Selected Elements
8.2.1.1 Iron

Iron concentrations in the wetland acrotelm vary from a
maximum of 415,000 ppm at site C-1 to 36,400 ppm at site AA-2
(Figure 8-1). The distribution pattern is similar to the
ground-water flow paths found during the tracer study which
strongly suggests the Swamp Gulch drainage is the source of most
of the iron found in the acrotelm. Concentrations of iron fall
from the maximum (about 8 times background levels) to values
about 2 times above background levels at site F-1. This
indicates the wetland has been effective in past removal of iron
from the AMD.

90

^ARBONATE MINE//' '/

A WELL SITE

O SURFACE WATER SAMPLE

SITE
a SURVEY CONTROL STATION
im FLUME

O PRECIPITATION GAGE
— lO-ISOLINES OF IRON (%)

SEC 20, TI5N,R6W
LEWIS AND CLARK CO, MONTANA

FEET
200

50 100

METERS

Figure 8-1. Iron concentrations in the wetland acrotelm.

An estimation of acrotelm iron mass due to the Swamp Gulch
AMD discharge indicated approximately 550 metric tons (600 U.S.
tons) of iron were deposited from this source. This amount
suggests the present annual estimated iron loading [(340 kg/year
(750 lbs/year - Section 7) has been greatly exceeded in the
past. It is likely that iron discharges to the wetland were
notably higher during the active mining/milling period and that
the annual estimated amount may have been underestimated due to
abnormal climatic conditions during the study period.

91

8.2.1.2 Manganese

The distribution of manganese in shallow wetland soils is
marked by high concentrations in the site D-1 to site F-1 area
(Figure 8-2) . Values at site D-1 are approximately an order of
magnitude above background sites AA-2 and AAA-3. Concentrations
of soil manganese decrease with increasing distance from site D-
1 suggesting some manganese is being removed from the ground-
water and sequestered in the acrotelm. However, the manganese
concentration at site F-1 is still three times greater than
background levels which indicates a notable amount of manganese
may likely be discharged from the wetland. Approximately one
metric ton (2,400 lbs) of manganese has been deposited in the
wetland acrotelm from the Swamp Gulch AMD. The boundary for the
elevated manganese impacted area was less well defined than that

CARBONATE MINE/^

f% // //

A WELL SITE

O SURFACE WATER SAMPLE

SITE
D SURVEY CONTROL STATION

r (V- * FLUME

j_r O PRECIPITATION GAGE

4^ 300-ISOLINES OF MANGANESE Cmg/ka)

SEC20, T15N,R6W
LEWIS AND CLARK CO .MONTANA

FEET
200

Figure 8-2. Manganese concentrations in the wetland acrotelm.

92

for iron and, therefore, the one ton estimate may be conserv-
ative. However, it is clear that the efficiency of this wetland
in the removal of manganese is at least an order of magnitude
less than for iron.

8.2.1.3 Aluminum

The relationship between the Swamp Gulch discharge and
elevated aluminum levels in the acrotelm are the least distinct
of the metals evaluated, possibly due to ICP iron interference
as suggested by the low recoveries found in the NBS standards.
Relative concentrations at several sites (BC-1, C-2, D-1, E-1,
E-2 and F-1) are apparently two to three times higher than
background levels at AAA-3, AA-2 and A-3 (Figure 8-3). However,
the reported aluminum values at sites CD-I, CD-1-2, and C-1.5,
all close to the Swamp Gulch discharge point, were similar to

CARBONATE MINE^

/ MJJ 'I

A WELL SITE

O SURFACE WATER SAMPLE

SITE
a SURVEY CONTROL STATION
H FLUME

O PRECIPITATION GAGE
2.0- ISOUNES OF ALUMINUM (%)

SEC20, TI5N,R6W
LEWIS AND CLARK CO .MONTANA

Figure 8-3. Aluminum concentrations in the wetland acrotelm.

93

background concnetrations . The relative aluminiim levels in the
acrotelm are consistently higher to the west of the C-1 to C-5
line than they are upgradient to the east. It is possible that
iron interference may have produced the low aluminum values
reported at sites CD-I, CD-1-2 and C-1. 5 and may have produced
lesser uncorrected interference with all samples. Due to the
apparent problems evident with aluminum analyses, no estimation
of the aluminum mass present in the acrotelm has been attempted.

8.2.1.4 Copper

Copper levels in the wetland acrotelm exhibit a marked
increase in concentration near the Swamp Gulch discharge point.
Background concentrations as represented by sites AAA-3, AA-2
and A-3 range from 210 mg/kg to 44 mg/kg. Sites B-2 and B-3
also exhibit low copper concentrations at 19 and 40 mg/kg,
respectively. Acrotelm copper values rise abruptly west of a
line running due south from site BC-1, which roughly corresponds
to the 100 mg/kg isoline (Figure 8-4) . The highest concen-
trations are found in the C-1. 5 to F-1 zone where copper levels
are generally 20 to 30 times the background mean. The remaining
wetland area to the south of this zone is characterized by
values 6 to 8 times the mean background concentration. The
distribution pattern suggests some amelioration of copper has
occurred in this system, but the high levels still evident at
site F-1 suggest much copper may pass through the system and
discharge to the Blackfoot River. The acrotelm copper mass is
about four metric tons (4.4 U.S. tons) which represents 325
years of deposition at the 1987-88 annual rate of 11.6 kg (25.5
lbs). While this period would be well in excess of the period
from mine initiation to the present, past loading rates were
likely higher (as observed with iron and lead deposition) and
hence, removal of copper from the AMD is likely considerably
less than 100 percent. The 40 to 50 mg/kg range, typical of the
wetland background areas, is common in numerous agricultural
soils (EPA 1987b) . Other studies have found soil total copper
levels of 500 to 700 ppm to produce severe yield reductions in
rye grass and bush beans (Wallace et al. 1977, McGrath et al.
1982) . Numerous other crop species are significantly impacted
at soil total copper levels as low as 100 ppm (EPA 1987b) .

8.2.1.5 Lead

Lead concentrations in the wetland acrotelm vary from less
than 1 mg/kg to values greater than 1000 mg/kg. Background
values upgradient from the Swamp Gulch discharge point are
generally less than 100 mg/kg but are higher than the 11.6 to
32.2 mg/kg range that Chattopadhyay and Jervis (1974) found for
the Holland Marsh muck soil. The anomalously high level (450
rog/kg) found at background site AAA-3 is likely the result of
other influences and not the result of the Swamp Gulch AMD
discharge. The wetland system is apparently quite efficient in

94

ARBONATE MINE

,.'■ •; '/

A WELL SITE

O SURFACE WATER SAMPLE

SITE
D iURVEY CONTROL STATION
M FLUME

O PRECIPITATION GAGE
— lOO-ISOUNES OF COPPER (m9/l'«)

SEC 20, TI5N,R6W

LEWIS AND CLARK CO .MONTANA

FEET
100 200 too

Figure 8-4. Copper concentrations in the wetland acrotelm.

the removal of lead, likely due in part to the low mobility of
lead in soils. The Swamp Gulch acrotelm contains a AMD derived
lead mass of 1.3 metric tons (1.4 U.S. tons) which represents
2,350 years of deposition at the calculated 0.54 kg (1.2 lbs)
annual loading rate. It is clear that past lead loading rates
were higher than those observed during this study. While some
of the observed lead may have been derived from natural
mineralization in the area, it is doubtful that much lead from
this source would have been deposited in the acrotelm due to the
semi-isolation of the acrotelm provided by the more limited

95

hydraulic conductivity of the underlying catotelm and the
limited mobility of lead. Elevated levels found in the zone
roughly described by sites C-1, C-1.5, D-3 and E-1 are an order
of magnitude above background sites. The concentrations at
sites E-2 and F-l are approaching background levels. Of
particular interest are the high lead concentrations found in
near surface materials at sites F-2 and F-3 (Figure 8-5). These
are likely the result of mining activities on the upper
Blackfoot River above this site. A literature review on the
phytotoxicity of selected metals (EPA 1987a) suggests lead
concentrations observed in the elevated zone would be phytotoxic
to most crops and may inhibit maximum vegetation production in
the wetland.

"ARBONATE MINE

A WELL SITE

U SURFACE WATER SAMPLE
SITE

o jUrvey control station

"m Flume

O precipitation gage

— lOO-ISOLINeS OFLEAO(m,/k,)

SEC 20, TI5N,R6W
lewis and CLARK CO ..MONTANA

Figure 8-5. Lead concentrations in the wetland acrotelm.

96

8.2.1.6 Zinc

Zinc concentrations in the wetland acrotelm exhibit several
anomalies, including high levels at background sites AAA-3 (1780
mg/kg) and A-3 (2092 mg/kg) and a low concentration at site C-1.
The cause of these variations is not apparent but may be due in
part to analytical interference. It is evident that zinc
concentrations in the acrotelm are consistently elevated south
of Montana Highway 200 from sites C-1. 5 to F-1 (Figure 8-6).
Zinc concentrations in this area are generally an order of
magnitude higher than most background levels. It is also
evident that zinc concentrations are not very effectively
reduced down-gradient from the Swamp Gulch discharge point.

"ARBONATE MINE,

A WELL SITE

O SURFACE WATER SAMPLE

SITE ■
D SURVEY CONTROL STATION
^ FLUME

O PRECIPITATION GAGE
500HS0LINES OF ZINC (mg/kg)

SEC 20, TI5N,R6W
LEWIS AND CLARK CQ .MONTANA

FEET
100 200 100

50 100

METERS

Figure 8-6. Zinc concentrations in the wetland acrotelm.

97

Concentrations of zinc in the elevated zone are consistently
above 1500 mg/kg, a level that would be expected to result in
nearly a 100 percent yield reduction for most agricultural crops
(EPA 1987a) .

The Swamp Gulch wetland acrotelm zinc mass due to AMD is
approximately 6.0 metric tons (6.6 U.S. tons) which would
represent 137 years of deposition at the observed 43.6 (95.9
lbs) annual loading rate. Although this time interval is in
excess of historic mining and AMD discharges, past discharges of
zinc were undoubtedly higher and hence, the actual time interval
represented is likely considerably less.

8.2.1.7 Arsenic

Arsenic concentrations at the four wetland sites analyzed
for this element (AA-2, C-1, C-4 and F-1) exhibit similar
distribution patterns with depth. All sites are characterized
by highest concentrations located at the top of the muck zone.
The maximum arsenic value is 86 mg/kg for the 0.46 to 0.91 m
interval at site C-4. Values at AA-2 and C-1 are similar and
suggest the wetland acrotelm has little effect on AMD arsenic
content at this study site. Total arsenic levels as low as 10
ppm have been shown to be detrimental to production for some
crop species (EPA 1987a) . Surf icial acrotelm values range from
18.5 mg/kg at site AA-2 to 29.5 mg/kg at site C-4, both of which
sites are not likely impacted by the Swamp Gulch AMD. Some
impact to wetland vegetation and wildlife may occur from the
observed arsenic concentrations.

8.2.1.8 Barium and Bervllium

Barium and beryllium distributions are generally similar to
arsenic. These two elements are more generally distributed
within the muck zone with the highest concentrations tending
more towards the central area of this stratigraphic zone. A
notable exception to this observation occurred at site C-1 where
the maximum beryllium concentration (240 mg/kg) occurred in the
acrotelm, strongly suggesting this element is enriched due to
the Swamp Gulch discharge. The maximum barium concentration
(2140 mg/kg) was found in the 1.77 to 1.89 m depth interval at
site F-1.

8.2.1.9 Cadmium

Cadmium concentrations in the wetland soils range from <0.5
to 45 mg/kg. This metal is apparently enriched 10 to 20 times
in wetland soils due to the Swamp Gulch AMD. The highest
concentrations are associated with the acrotelm layer at site F-
1 (44 and 45 mg/kg) . At site F-1, the muck zone maximum is 4
mg/kg. This is in contrast to site C-1 where there is
apparently little difference between most muck values and the

98

acrotelm (1.2 to 21 mg/kg and 22 mg/kg, respectively) . The AMD
derived acrotelm cadmium content is only 4 kg (8.8 lbs) which
represents 6.7 years of deposition at present loading rates.
These data suggest some Swamp Gulch discharge is percolating
through the muck zone at site C-1 and that the wetland is
generally less effective in ameliorating cadmium in AMD. At
background site AA-2, the cadmium concentration in the acrotelm
is 2 mg/kg and a maximum concentration of 8.6 mg/kg is
associated with the gravels beneath the muck zone. All muck
zone values at this site are less than 2 mg/kg. The 22 to 45
mg/kg values found in the acrotelm at sites C-1 and F-1 can be
expected to severely impact production of many crops (EPA 1987a)
and can be expected to add significant cadmium to the food
chain, possibly impacting wetland wildlife (Section 9) .

8.2.1.10 Cation exchange capacity

Cation exchange capacity (CEC) of the acrotelm determined
on sites AA-2, C-4 and F-1 (Appendix C) , had a weighted mean of
35.6 raeq/lOOg. The range of CEC in the wetland, including site
C-1, was 4.9 to 45.0 meq/lOOg. These CEC findings are lower
than those typically found in wetlands. Mathur and Farnham
(1985) found wetland peat materials with a mean CEC of 148.5
meq/lOOg. These low CEC findings in the Swamp Gulch wetland may
be the effect of the method of CEC analysis. Presuming the CEC
found is true, the low CEC may be a reason for the limited level
of wetland effectiveness in ameliorating the AMD metals.

All samples from sites AA-2, C-1, C-4 and F-1 that have
been analyzed for mercury, antimony, and thallium are below the
1 mg/kg instrument detection limit for these elements.

8.2.2 Distribution of Elements with Depth

An evaluation of selected trace metals distributions with

depth indicates that, in general, the anaerobic muck zone below

the acrotelm contains lower concentrations of most elements
(Table 8-1) .

Aluminum is a notable exception, and is often found at
higher concentrations in the muck zone. A comparison between
mean background levels and concentrations within the AMD
impacted area suggests the muck zone is impacted by AMD (Table
8-1) . All metals evaluated in the impacted muck zone were
higher than corresponding background values. However, due to
the lower hydraulic conductivity of the muck zone, it likely has
limited effective remediation potential for AMD amelioration. A
comparison of muck zone concentrations with the underlying
gravel materials reveals a less clear situation in which,
compared to the catotelm, some metals are higher and some lower
in the gravels, zinc levels were often higher in the gravels.

99

Table 8-1. Mean selected elemental concentration (mg/kg)

changes with depth for background and AMD impacted

areas,

Site

Al

Cu

Fe

Mn Pb

Zn

Zone - Acrotelm

Background

(AA-2,A-3) 9,930 46 64,100 225 125 1380

Impacted Zone

(C-1,D-1,F-1) 27,150 1287 236,000 898 689 2580

Zone - Catotelm

Background

{AA-2,A-3) 22,300 92 42,300 123 117 188

Impacted Zone

(C-1,D-1,F-1) 37,800 597 63,800 468 291 704

Zone - Gravels

Background

(AA-2,A-3) 23,200 285 26,250 95 77 1190

Impacted Zone

(C-1,D-1,F-1) 17,830 563 32,870 413 170 693

especially near the top of this unit (Appendix C) . Relative
sulfate levels, as determined from the saturated paste extracts,
suggest the distribution of sulfate is similar to zinc. The
highest sulfate values for background sites AAA-3, AA-2 and A-3
are found in the underlying gravels. This distribution is also
evident at sites D-1 and D-2. Sites B-2, C-5 and D-3 have the
highest sulfate concentrations in the acrotelm. The only site
where high sulfate values are associated with the catotelm is C-
1. Low sulfate values in the catotelm would be expected due to
reducing conditions which would tend to reduce sulfate to
insoluble sulfides which would not be recovered in a saturated
paste extract. It is possible that the total sulfur
concentration in this zone would exceed that in other strati-
graphic units. This observation of sulfate removal is
substantiated by the studies on microbiological sulfate removal
presented in Section 10, Phase II.

100

8 . 3 SUMMARY

Annual deposition of metals into the acrotelm have
probably decreased from levels present during the period of
active mining and milling. Acrotelm masses of lead, iron,
copper and zinc would, at present depositional rates, represent
accumulation periods ranging from 2,350 to 137 years (Table 8-
2) . If the assumption is made that the apparent depositional
interval (2,350 years) found for lead represents 100 percent
removal of this metal, and that this interval can serve as a
relative basis of comparison with other elements, the wetland
efficiency in removal of metals is: iron, 70%; copper, 14%;
zinc, 5.8%; manganese, 0.7%; and cadmium, 0.3% (table 8.2).
Assuming past loading rates were higher than those observed at
present, it is apparent that cadmium, manganese and likely zinc
are all passed through the wetland without much amelioration.
Copper may also be passed through the system, but it would
appear that it is at least partially ameliorated. Both lead and
iron may be effectively removed within the wetland area. Much
of the apparent decrease in wetland water sample concentrations
of cadmium, manganese, zinc and possibly copper with increasing
distance from the AMD discharge point, is probably due to
dilution processes rather than immobilization. The higher
levels of metals observed in the gravels may be derived from two
sources: 1) from the original gravel materials which contain
pebbles and grains of mineralized materials derived from
upstream outcrops; and 2) from mineralized ground-water flowing
laterally in these materials. It is also

possible, especially at site D-1, that there could be an upward
ground-water gradient beneath the gravels in bedrock materials.
This ground -water could be derived in part from seepage
emanating from old underground mine workings.

Table 8-2. Acrotelm metal masses and the relative efficiency of
wetland metal removal.

Annual

Relative

Acrotelm

Input from

Years represented efficiency

metal mass

AMD

by deposition

compared

f metric tons)

(Kq)

rate

to leadf%)

Cadmium

0.004

0.54

6.7

0.3

Copper

3.8

11.6

325.

14.

Iron

553.

338.

1,640.

70.

Lead

1.3

0.54

2,350.

100.

Manganese

1.1

63.4

16.9

0.7

Zinc

6.0

43.6

137.

5.8

101

It is clear that most observed sequestering of AMD metals
occurs in the acrotelm, with the catotelm playing a minor role.
Efforts should be concentrated on constructing effective acrotelm
zones in artificial wetlands, and over time, a natural catotelm may
form from material derived from the acrotelm.

102

9.0 WETLAND VEGETATION AND THE CONTROL OF ACID MINE DRAINAGE

The objectives of this portion of the Swamp Gulch wetland
study were to:

1) quantitatively describe plant associations and determine
production by dominant species, and

2) assess metal loading of dominant vegetation and
potential impacts to wildlife in the natural wetland.

9.1 SAMPLE COLLECTION, PREPARATION AND ANALYSIS

Vegetation observations were recorded during three trips to
the Swamp Gulch wetland from April through August, 1987. Plant
reference specimens and vegetation samples were collected from
the Swamp Gulch study area and from the Hardscrabble Creek
background wetland during August 1987. Specimens have been
deposited in the Reclamation Research Herbarium.

The composition of the vegetation at the Swamp Gulch site
was determined by estimating plant canopy coverage. Eight
transects were located across the study area, along which
quadrats (20 X 50 cm) were placed at regular, preselected
intervals. A total of 104 quadrats were used. Species specific
canopy coverage was estimated at each sampling station to the
nearest percent (Sensu Daubenmire 1959) . Depth to the water
table was also estimated at each sampling station. At the
background site on Hardscrabble Creek, canopy coverage values
for the dominant species were visually estimated for the site as
a whole.

Aboveground production for the dominant herbaceous species
(Carex rostrata ) was estimated at the Swamp Gulch and the
Hardscrabble background site using 0.25 m^ (0.5 x 0.5 m)
quadrats. Plants were clipped at ground level from seven
quadrats at the Swamp Gulch site and in two quadrats at the
background site. These samples were placed in paper bags, oven-
dried at 50° C to a constant weight and then weighed to the
nearest 0.1 gram. Values were converted to kg/ha and
comparisons were made between the Swamp Gulch site and the
Hardscrabble site at the 0.05 level using the Students t-test.

Vegetation samples from throughout the Swamp Gulch area and
background sites were collected for chemical analysis. To
expedite this process, the Swamp Gulch study area was divided
into seven collection zones. From each zone, whole plants were
collected for Carex rostrata and the mosses, Isopterygium
pulchellum and Sphagnum tenellum . Current growing season shoots
and leaves were collected from dominant shrub species (Salix

103

boothii and Betula alandulosa ) , as were roots up to approxi-
mately 3 cm in diameter. Aboveground materials were placed in
paper bags and below ground materials were placed in plastic
bags. Initial sample preparation consisted of washing each
sample with tap and distilled water and then subjecting it to
ultrasound until no residues were observed in the water. Each
sample was then oven-dried (50° C) , placed in paper bags and,
together with National Bureau of Standards (NBS) quality control
samples, were shipped to Energy Laboratories in Billings,
Montana on September 28, 1987. A total of 67 samples (65
natural and 3 NBS citrus leaf samples) were analyzed for Al, As,
Cd, Ca, Cu, Fe, Pb, Mn, Ni and Zn. At the laboratory, each
sample was digested with nitric acid and hydrogen peroxide
according to method 3050 of the EPA Manual SW-846 (EPA 1982).

Concurrent with vegetation sampling, a map was drawn
delimiting the plant associations, areas of open water, transect
locations, and man-made features such as roads.

Canopy coverage and frequency data were summarized for the
Swamp Gulch study area as a whole. Nomenclature follows Dorn
(1984) for vascular plants and Crum et al. (1973) for mosses.
The final plant association map was constructed with the aid of
the field map, field notes, canopy cover data, and after a
review of wetland classification schemes made by other authors.

9.2 DATA QUALITY ASSURANCE/ QUALITY CONTROL

The quality of vegetation chemical data were determined for
both field and laboratory procedures. Field sampling included
the collection of duplicate samples and the insertion of NBS
Blind Field Standard (BFS) reference samples into the sample
train. Laboratory QA/QC procedures consisted of splits of
natural and field duplicate samples. Analytical results of
these samples were used to calculate precision and accuracy
statements following standard procedures developed for the EPA
Contract Laboratory Program. Appendix D-1 presents the results
of the data accuracy and precision statements.

Accuracy is interpreted through the following example: for
Al, we are 90% confident that the reported value is 44.5 ± 6.3%
of the true value. Precision is interpreted as follows: for
Al, we are 90% confident that the true value is within ±4.5% of
the reported values.

104

9.3. RESULTS AND DISCUSSION

9.3.1. Vegetation Patterns

The Swamp Gulch wetland was a mosaic of plant
associations, the presence of which was directly related to
flooding frequency and duration. Land surface varied from areas
of continually open water (about 4.6% of the site) to areas
where the soil was periodically saturated (sometimes nearly year
round) , to areas where the water table was rarely high enough to
saturate the entire soil profile. The wettest areas were
characterized by pure stands of Carex rostrata (Figure 9-1).
This species was nearly ubiquitously present, being associated
with Betula qlandulosa and Salix boothii, and with Pinus
contorta. Carex rostrata was found growing as an emergent (in
as much as 20 cm of water) and where the soil surface was only
damp. A relatively dry site was occupied by Pinus contorta,
Chimaphila umbellata and S phagnum tenellum . The driest areas
were dominated by species indicative of the surrounding forest;
these areas were characterized by Pinus contorta, Picea
englemannii . and Arctostaphylos uva-ursi . A list of vascular
plants and bryophytes identified within the Swamp Gulch wetland
study area can be found in Appendix D-2. Beaver dams, extending
across the study area near transect 6 (Figure 9-1) , contributed
to the patterning of plant associations. The eastern two-thirds
of the site had a higher water table than the western one-third,
and was therefore dominated by Carex rostrata and associated
communities. The western portion was dominated by Pinus
contorta and forest grasses and forbs.

9.3.2. Canopy Coverage and Production

Carex rostrata was the dominant species throughout the
study area. It was found at over 90 percent of the sampling
stations and covered more than 55 percent of the study site
(Table 9-1) . Betula glandulosa was the dominant shrub and Pinus
contorta was the dominant tree. Salix boothii was also
important; it covered 5.5 percent of the study area and was
found at almost 20 percent of the sampling stations. Two
mosses. Sphagnum tenellum and Isopterygium pulchellum were
commonly encountered, especially in the wetter, eastern two-
thirds of the site. Chimaphila umbellata was associated almost
exclusively with Pinus contorta in a small area adjacent to
Highway 200 (Figure 9-1) . Sphagnum tenellum grew profusely
within this association.

Many dead Pinus contorta can be observed, especially in
the eastern portion of the wetland. These were from 20-30 feet
tall and probably from 40-60+ years old. The most likely cause
of death was a rise in the water table, which may have been the
result of fluctuating climatic conditions.

105

I
-J

■J

m

M
Q)

>

o
o

>1
a
o
c

(0

u

•o
c

03
U)

c
o

•l-l

m

•rH
O

o

CO

to

(0

c •
o m

-i-l 4J

-P u
m 0)

4J

CD U

> -p

I

0)
M
D

-10 6

Table 9-1.

Mean percent canopy coverage and percent frequency
for plant species.

Life

% Canopy

%

Species

Forml

Coveraae
55.5

Freauencv

Carex rostrata

GL

90.4

Betula glandulosa var. hallii

S

11.3

31.7

Pinus contorta

T

6.3

8.7

Salix boothii

S

5.5

19.2

Sphagnum tenellum

M

4.7

12.4

Chimaphila umbellata

S

3.0

8.7

Potomogeton sp.

F

2.8

1.9

Isopterygium pulchellum

M

2.8

18.1

Pyrola asari folia

F

1.2

7.6

Arctostaphylos uva-ursi

HS

1.1

5.7

Abies lasiocarpa

T

0.9

1.0

Juniperus communis

S

0.9

1.9

Picea engelmannii

T

0.9

4.8

Scirpus acutus

GL

0.9

1.2

Pseudotsuga menziesii

T

0.6

1.0

Equisetum sp.

F

0.5

11.5

Vaccinium scoparium

HS

0.3

1.0

Achillea millefolium

F

0.1

1.9

Elymus glaucus

G

<0.l2

1.0

Fragelia virginiana

F

<0.1

4.8

Festuca idahoensis

G

<0.1

1.0

Geum macrophyllum

F

<0.1

1.0

Phleum pratense

G

<0.1

1.0

Unknowns

2.2

1.0

TOTAL

101.8

LF

Life Form: GL = Grasslike, S = Shrub, T = Tree, M = Moss,

F = Forb, HS = Half -Shrub and G = Grass
2 Values of <0.1 were tabulated as 0.05.

Herbaceous production on the Swamp Gulch wetland averaged
3820 kg/ha (Table 9-2) . For comparison, vegetation composition
and production were estimated at a pristine background wetland
on Hardscrabble Creek (off the Alice Creek road) . The
composition of the vegetation was very similar to the Swamp
Gulch wetland. Carex rostrata dominated with about 60 percent
canopy coverage and Salix boothii and Betula glandulosa were the
dominant shrubs. Water temperature, pH and EC were 11° C, 7.4
and 400 umbos/ cm, respectively. Herbaceous production at the
background site was 3750 hg/ha, which was not significantly

107

Table 9-2. Peak standing crop (kg/ha) at the Swamp Gulch
and Hardscrabble Creek (background) wetlands.

Hardscrabble Creek Swamp Gulch Wetland

(background)
Mean^ STD Range of Values Mean STD Range of Values

3750a ±532 3374-4127 3820a ±1550 1983 - 6776

^ Means were not significantly different at the p<.05 level.

different than production on the Swamp Gulch site (Table 9-2) .

9.3.3 Element Concentrations in Plant Material

This section presents data on the concentration of elements
in the dominant plant species at the Swamp Gulch study site and
at the Hardscrabble Creek (background) wetlands. These data are
compared, contrasted and discussed with respect to water quality
data.

Statistical analysis was conducted to determine if
elemental differences in plant tissues could be detected between
the Swamp Gulch and the background sites. Large variations in
the analytical data (and low sample numbers) resulted in large
standard deviations. This situation restricted the usefulness of
the statistical analyses by limiting the number of possible
significant determinations. As a consequence, element data are
discussed in a general manner without the use of statistics.

A literature search revealed no data on the metal levels of
the plant species occupying the study site. Considerable data
exists however, for species of cattails (T ypha ) . Hutchinson
(1975) reviewed the literature on metal levels in aquatic
vegetation. EPA (1987a and b) conducted a literature review on
metal concentrations in terrestrial macrophytes for the purpose
of establishing plant and animal hazard criteria. These data
sets will be compared, in a general way, with element data in
the tissues of the macrophytes at the Swamp Gulch and
Hardscrabble Creek wetlands. Metal levels in a variety of
bryophytes, particularly Sphagnum spp., growing in metal-
enriched areas have been reported. These data will also be
compared with the bryophytes metal data from the Swamp Gulch and
Hardscrabble sites.

Water quality data are presented for three locations at the
Swamp Gulch site. This is done to show the degree of metal
enrichment in water as it flows past the mine site, and to help
explain the level of metal attenuation by the wetland.

108

Background water chemistry levels were established from samples
collected in Swamp Gulch above the Carbonate Mine. References
will also be made to the water quality at station F-1, located
directly down-gradient from the Swamp Gulch wetland. These data,
therefore, should provide information as to the quantity of
metals being attenuated by the wetland.

Table 9-3. Concentration (ug/g) of elements in Carex rostrata .

Background Wetland
Element Matrix^ Mean(n=4) Range

Swamp Gulch Wetland
Mean(n=5V Range

Al

As

Cd

Ca

Cu

Fe

Pb

Mn

Ni

Zn

AG

61

11-170

232

41-400

BG

408

72-1085

1720

850-2670

AG

0.9

0.352-2.1

0.9

0.6-1.4

BG

16

1.8-35

19

1.5-41

AG

0.8

0.35-1.6

0.9

0.5-3.5

BG

6.2

0.6-21

14

6.5-17

AG

3285

360-4530

2456

1110-3450

BG

6420

2340-15000

3500

670-5150

AG

3.8

0.352-8.7

62

0.35-160

BG

30

5.1-78

984

88-2240

AG

1327

350-3840

6134

2440-15300

BG

20800

5070-58500

37140

24800-54000

AG

6.5

1.0-22

29

5.2-70

BG

128

2.0-490

354

190-800

AG

248

81-460

250

160-430

BG

177

88-380

131

43-190

AG

0.72

0.7-0.7

0.8

0.7-1.2

BG

1.9

0.7-5.5

9.3

0.7-17.0

AG

96

14-320

134

69-290

BG

1096

42-4210

1296

310-2700

Ipiant material: AG=above-ground, BG=below-ground
2 Values below the instrument detection limit were estimated by
multiplying the instrument detection limit value by 0.7.

109

9.3.3.1 Effects of Carex rostrata in Remediating AMD

Below-ground tissues had higher concentrations than above-
ground tissues (Table 9-3). This was true for the background
and the Swamp Gulch wetlands and for all the elements with the
exception of Mn, which appeared to be more concentrated in the
leaves than in the roots/rhizomes. These findings were consis-
tent with previous research on metal uptake by species of Typ ha
(Bayly and O'Neill 1972, Heil and Kerins 1988, Taylor and

Crowder 1983a, 1983b) . In the present study, above-ground
tissues concentration were different from below-ground
concentrations by an order of magnitude for Al, As, Cd, Cu, Fe,
Pb and Zn.

Water quality samples collected down-gradient from the
Swamp Gulch wetland indicated that some of the metals were being
attenuated by Carex rostrata . Water at site F-1 had lower
concentrations of Al, Cd, Cu, Fe, Pb, Ni and Zn than the
influent water (Table 9-4) . Sulfate was also lower and pH was
unchanged.

Some elements (Al, Cu, Fe and Pb) were substantially
elevated in plant tissues from the Swamp Gulch site, compared to
background levels. These elevated levels were apparently not

Table 9-4. Water quality^ in Swamp Gulch above and below the
Carbonate mine site, and below the Swamp Gulch
wetland at station F-1.

Above Carbonate

Below Carbonate

Swamp Gulch

Mine

Mine

Wetland

Parameter

(background)

rAMD)

rstation F-1)

Al

<0.1

6.2

0.4

As

<0.005

<0.005

<0.005

Ca

12.7

22.0

26.0

Cd

0.001

0.038

<0.001

Cu

<0.01

0.61

0.04

Fe

0.7

25.8

15.1

Pb

<0.01

0.035

<0.01

Mn

0.08

3.73

4.30

Ni

<0.03

0.37

<0.03

Zn

<0.01

2.64

0.12

SO4

10.7

202.3

142.0

pH

6.8

3.6

3.3

^Values are means from three sampling dates: 4/29/87, 7/28/87
and 1/11/88.

110

phytotoxic . Carex rostrata did not appear to be under abnormal
stress at the Swamp Gulch site; production (3820 kg/ha) and
cover (56 percent) were high, and individual plants appeared to
be healthy. As previously discussed, production was
statistically similar between the two sites and no difference in
canopy cover was visually apparent. Phytotoxicity for Carex
rostrata is discussed by element in the remainder of this
section.

Although Al was elevated in plant tissue (approximately four
times background) at the Swamp Gulch site, the above-ground mean
of 232 ug/g fell within the range for aquatic plants reported by
Hutchinson 1975 (Table 9-5) . The below-ground tissue concentra-
tion of 1720 ug/g , however , was substantially higher than data
from non-enriched areas reported by other researchers (Table 9-
5) . Heil and Kerins (1988) reported similar results for T ypha
latifolia in a wetland constructed in Montana to control AMD;
tissue concentrations of Al were higher from the area receiving
AMD than from a control site. The uptake of Al by Carex
rostrata may have contributed to the low quantity (0.4 mg/L) of
Al measured in water flowing from the Swamp Gulch wetland (Table
9-4) at station F-1. The most reasonable explanation for the
higher levels of tissue Al in the Swamp Gulch wetland was that
Carex rostrata has the ability to absorb this element if it is
available. Water quality data showed that Al was substantially
higher in the AMD entering the

Table 9-5.

Range and mean element concentrations (ug/g) in
aquatic forbs and grasses (from Hutchinson 1975)'^

Metal

Range

Mean

Al
As
Cd
Cu
Fe
Pb
Mn
Ni
Zn

250-785
2.8-202
2.6-28
2.5-243

70-31, 500^
2.0-53
100-23,000
1.1-44
26.5-1000

366

8.0
48
3170
11
2380

143

-"■Except where noted, these values are from non-enriched environ-
ments. However, some of the plant species investigated may be
metal lophytes .

^From As-enriched environment.

•^This range may be biased high due to analyses of vegetation that
were not completely cleaned of Fe precipitates.

Ill

wetland compared to water above the Carbonate Mine site (Table 9-
4) . This indicates, at least to some degree, that Al will be
absorbed by Carex rostrata .

The average concentration of Fe in the influent AMD (25.8
mg/L) was 37 times higher than the water above the Carbonate Mine
site (0.7 mg/L). Although these data are limited, the concentra-
tion of Fe in water at station F-1 was 15.1 mg/L, which was less
than the concentration in the influent water (25.8 mg/L). Tissue
concentrations for Fe in the Swamp Gulch wetland were 4.6 (above-
ground) and 1.8 (below-ground) times higher than the background.
This suggests that a portion of the dissolved Fe was being
absorbed by Carex rostrata . Iron precipitates adhering to
organic matter also contribute to a reduction in dissolved Fe
content as the water passes through the wetland. Leaf tissue
levels of Fe at the Swamp Gulch wetland were within the range
reported for aquatic plants (Table 9-5). Bayly and O'Neill
(1972) reported that Tvpha latifolia had leaf tissue Fe levels
from 200 to 1800 ug/g and rhizome Fe levels from 700 to 3600
ug/g, depending on the season in which samples were collected.
In a metal smeltering area, Fe concentrations in T ypha latifolia
leaves and roots reached 905 and 57,138 ug/g, respectively
(Taylor and Crowder 1983b) . These data were comparable to those
measured in the Swamp Gulch wetland, as were values reported by
Heil and Kerins (1988) . These researchers reported Fe concentra-
tions in leaf tissue of 286 ug/g and root tissue levels of 40,884
for Typha latifolia growing in an artificial wetland receiving
AMD.

Like Al and Fe, the concentration of Zn in water increased
as it flowed past the Carbonate Mine site (Table 9-4) and was
likewise attenuated as the water passed through the wetland. The
influent concentration of Zn was 2.64 mg/L, while the concentra-
tion at station F-1 was only 0.12 mg/L. This large decrease was
not due to absorption by Carex rostrata since tissue concentra-
tions were only slightly greater at the Swamp Gulch wetland
compared to the background wetland (Table 9-3) . Discussion in
the following sections will show that the shrubs accumulated
substantial amounts of Zn. Tissue levels in Carex rostrata from
both sites were similar to aquatic plant data from non-enriched
areas (Table 9-5) . In a metal-enriched environment, root and
leaf Zn concentrations in T ypha latifolia ranged from 13 to 101
ug/g and from 24 to 572 ug/g, respectively (Taylor and Crowder
1983b) . These researchers also reported that tissue levels of Zn
were not correlated with soil/sediment levels. Phytotoxicity of
Zn in terrestrial plants varies from 60 to more 800 ug/g
depending on the plant species, the variety and the part of the
plant analyzed (EPA 1987a) . These authors suggested a 50 ug/g
tolerable Zn level and a 500 ug/g phytotoxic level for above-
ground tissue.

112

Dissolved Cu increased in Swamp Gulch water as it flowed
past the Carbonate Mine site. Copper concentrations in water
increased from <0.01 mg/L above the mine to 0.61 ug/L below the
mine. Apparently, Cu was attenuated in the wetland since the
concentration in water at station F-1 was 0.04 ug/L. Vegetation
data indicated substantially higher (16 and 33 times greater than
background in above-ground and below-ground tissues, respective-
ly) Cu concentrations in Carex rostrata tissue from the Swamp
Gulch wetland compared to the background site, suggesting that
this species was absorbing dissolved Cu from AMD. Leaf tissue
concentrations in the Swamp Gulch wetland (0,35 -160 ug/g) were
within the range reported for aquatic plants (Table 9-5) .
However, the range of Cu values in the below-ground material (88-
2240 ug/g) often exceeded the range for other aquatic plants.
From a natural but metal-enriched wetland, Typha latifolia leaf
concentrations reached 24 ug/g and root levels reached 265 ug/g
(Taylor and Crowder 1983a) . The global diagnostic background
level for Cu in above-ground tissue of terrestrial plants ranges
up to 20 ug/g (EPA 1987b) . Phytotoxicity in terrestrial plants
generally begins at levels >20 ug/g. Some Carex rostrata plants
in the Swamp Gulch wetland had leaf tissue levels up to 160 ug/g,
suggesting that some phytotoxicity may be occurring due to Cu.
However, Carex rostrata may have an internal exclusionary
mechanism that allows it to tolerate relatively high levels of Cu
in its tissues.

The concentration of Pb in the wetland influent AMD was
0.035 mg/L, which was slightly greater than <0.010 mg/L in the
water at station F-1 (Table 9-4) . These data suggest that some
Pb (albeit, very little) was being attenuated within the wetland,
a portion of which may have been absorbed by the sedge. Carex
rostrata Pb tissue levels were 4.5 (above-ground tissue) and 2.8
(below-ground tissue) times higher in samples from the Swamp
Gulch wetland compared to the background wetland (Table 9-3).
Above-ground tissue concentrations were within the range for
aquatic plants reported by Hutchinson (1975) (Table 9-5). It has
been difficult to establish phytotoxic levels for Pb in plants
because of the great variability between species. Background
levels in lettuce leaves, for instance, have been reported as
high as 50 ug/g, while phytotoxic levels in alfalfa have ranged
down to 10.8 ug/g (EPA 1987a). These authors felt that a leaf Pb
level of 25 ug/g could, in general, be tolerated by grain crops
in Montana. Since leaf tissue levels ranged from 5.2 to 70 ug/g
in the Swamp Gulch wetland, some plants may be under stress due
to high levels of Pb. However, it is also possible that Carex
rostrata has a mechanism which allows it to tolerate higher Pb
levels than crop species.

The lower concentrations of the aforementioned metals (Al,
Cu, Fe, Pb and Zn) in water at station F-1 compared to the
influent AMD indicates that a portion of these elements were
being attenuated within the wetland. Vegetation data suggests

113

r

that Carex rostrata absorbed a portion of these metals.
Manganese, by contrast, was not accumulated by Carex rostrata
even though the concentration of Mn was notably elevated (47
times) in the influent AMD. Concentration in the influent AMD
was 3.73 mg/L compared to the background water level of 0.08
mg/L. Both above- and below-ground tissue concentrations were
similar between the background and Swamp Gulch wetlands.
Apparently, Carex rostrata has an external exclusionary mechanism
that prevents the absorption of Mn. Not only was Mn not being
absorbed by Carex rostrata, it was not being attenuated in any
other manner. The concentration of Mn in water down-gradient
from the wetland (at station F-1) was 4.30 mg/L, which was
slightly greater than the influent concentration of 3.73 mg/L.
In contrast, Typha latifolia can absorb substantial amounts of Mn
from water, Heil and Kerins (1988). These investigators reported
mean leaf tissue levels of 1053 and 1760 ug/g of Mn in T ypha
latifolia from two artificial wetlands receiving AMD having an
average of 1.35 mg/L of Mn in the influent water. Manganese
tissue levels have ranged up to 23,000 ug/g in aquatic plants
(Table 9-5). However, these data were from a study conducted in
the Ukraine and it is not clear from Hutchinson's discussion
whether the samples were from a metal-enriched area or from a
metallophytic species.

The concentration of As in the tissues of Carex rostrata
were similar between the background and the Swamp Gulch wetlands
(Table 9-3) . The concentration of As in the Swamp Gulch water
did not increase as it flowed past the Carbonate Mine site (Table
9-4) . The tissue concentrations observed in the Swamp Gulch
wetland were within the range reported for aquatic plants (Table
9-5) . Depending on the plant species and the particular tissue
analyzed, phytotoxic symptoms in terrestrial plants have been
reported from 5 to 20 ug/g As (EPA 1987a) . Most of these data
were from crop species, which may be more sensitive to As than
Carex rostrata . Based on the currently available aquatic and
terrestrial plant data, as well as data from the Hardscrabble
Creek (background site) , there is no reason to suspect that As is
toxic to Carex rostrata at the levels observed.

The concentration of Cd in water above the Carbonate Mine
was 0.001 ug/g compared to 0.038 ug/g below the mine site (Table
9-4) . Mean leaf tissue concentrations were similar between sites
but the range of values suggests that Carex rostrata was
absorbing and accumulating more Cd at the Swamp Gulch wetland
than the background wetland. Leaf tissue levels at the back-
ground site ranged from 0.35 to 1.60 ug/g, compared to 0.5 to 3.5
ug/g at the Swamp Gulch wetland. Tissue levels from this study
were similar to those of other aquatic plants (Table 9-5). A
large body of data exists on Cd toxicity in crops. Background
levels for Cd have been reported up to 3.1 ug/g in alfalfa tops
and yield increases have been reported when Cd was present at
much higher concentrations (EPA 1987a) . In a study of hazardous

114

levels in crops from a metal enriched area, 10 ug/g Cd in plants
was considered tolerable while a tissue level of 50 ug/g would
definitely cause a yield reduction. In light of these hazard
levels and the data from the Swamp Gulch site, the levels of Cd
in Carex rostrata appear to be normal and well below the
phytotoxic level. It should be noted that while these levels may
not be phytotoxic, they are at least an order of magnitude higher
than concentrations found in the AMD water and represent notable
bioaccumulation .

Like most of the other elements, Ca was more concentrated in
the roots/rhizomes than in the leaves (Table 9-3). As previously
mentioned, Ca levels were higher in tissues from the background
wetland compared to the Swamp Gulch wetland. This was most
likely due to a greater availability of Ca at the background
site, although the water and soil at that site were not analyzed.
Tissue Ca levels from the Swamp Gulch wetland were within the
ranges reported by Taylor and Crowder (1983b) for Typha . Bayly
and O'Neill (1972) reported similar levels for rhizomes (2000-
9000 ug/g) but higher values for leaves (5000-11,000 ug/g),
compared to Swamp Gulch wetland values.

The concentration of Ni in leaf and below ground material in
the Swamp Gulch wetland were probably not elevated above
background (Table 9-3). These levels fell within Hutchinson's
(1975) range for aquatic plants (Table 9-5). Taylor and Crowder
(1983b) reported leaf tip concentrations for Tvpha latifolia that
ranged up to 91 ug/g and root levels ranging up to 388 ug/g in
plants from a metal enriched area. Nickel leaf concentrations of
467 ug/g have been obtained without toxic symptoms, indicating an
internal exclusionary mechanism (Taylor and Crowder 1983a) .
Below-ground tissue concentrations have been correlated with
soil/sediment concentrations, but no correlation between leaf and
soil levels have been found (Taylor and Crowder 1983b) .

9.3.3.2 Effects of Salix boothii in Rem ediating AMD

Like Carex rostrata the metal concentrations in Salix
boothii were generally higher in below-ground compared to above-
ground plant tissue (Table 9-6) . This pattern occurred at both
the background and Swamp Gulch wetlands and for all elements,
with the possible exception of Zn. Zinc appeared to be more
concentrated in the leaves/stems than the roots/rhizomes,
especially in the Swamp Gulch wetland.

Aluminum was absorbed by Salix boothii and concentrated in
the below-ground tissue as opposed to the leaves and stems. The
level of Al in the below-ground tissue from the Swamp Gulch
wetland was substantially higher (3.4 times) than the tissue
concentration from the background site (Table 9-6) . However,
above-ground tissue concentrations were similar between wetlands.
The ability of Salix boothii to absorb and store Al may have

115

Table 9-6. Concentration (ug/g) of elements in Salix boothii .

Background Wetland
(n=2)
Element Matrix^ Mean Range

Al

As

Cd

Ca

Cu

Fe

Pb

Mn

Ni

Zn

Swamp Gulch Wetland
(n=6)
Mean Range

AG
BG

AG
BG

AG
BG

AG
BG

AG
BG

AG
BG

AG
BG

AG
BG

AG
BG

AG
BG

51
190

0.352
2.9

0.6
0.6

6680
8580

0.352
11

1535
3000

0.72
3.9

110
125

0.72
0.7

130
105

34-68

50

110-270

655

0.35-0.35

0.35

2.3-3.4

3.6

0.5-0.6

5.6

0.5-0.7

9.1

6560-6810

6817

7590-9570

4535

0.35-0.35

1.3

8.7-13

169

890-2180

132

1830-4290

18418

0.7-0.7

1.7

1.7-6.0

112

99-120

257

120-130

456

0.7-0.7

0.9

0.7-0.7

1.6

110-150

388

100-110

278

35-74
170-1640

0.35-0.35
0.35-10.0

1.0-9.9
2.3-17

5040-7750
3240-4860

0.35-6.0
0.35-390

59-420
4470-33000

0.7-3.8
9.2-350

150-500
63-1380

0.7-1.6
0.7-2.7

210-500
160-420

^Plant material: AG=above-ground, BG=below-ground

2values below the instrument detection limit were estimated by multiplying
the instrument detection limit value by 0.7.

contributed to the decrease in Al measured in water leaving the
Swamp Gulch wetland (Table 9-4) .

Arsenic tissue concentrations were similar between the
background and the Swamp Gulch wetland. This was expected since
As was not elevated in the influent AMD. In the above-ground
tissue, As was well below the phytotoxic range (5-20 ug/g)
reported for terrestrial plants (EPA 1987a) . Arsenic in the
below-ground tissues ranged from 2.9 to 3.6 ug/g, indicating
that As accumulates in these materials over time.

116

Salix booth ii tended to accumulate Cd, despite the fact
that it was present at a low concentration in the influent water
(Table 9-4) . The level of Cd in the AMD entering the wetland
was 0.030 mg/L compared to 0.001 mg/L in the non-enriched water
above the Carbonate Mine site. This increase in dissolved Cd
was enough to increase tissue levels more than nine times
background (Table 9-6). The absorbance of Cd by Salix boothii
may have contributed to the reduced level of Cd measured in
water leaving the Swamp gulch wetland at station F-1. Based on
the tolerable level (10 ug/g) for terrestrial plants (EPA
1987a) , it is reasonable to assume that Salix boothii is not
experiencing toxic affects from Cd.

Calcium behaved differently in Salix boothii than in Carex
rostrata . For Carex rostrata . Ca levels were lower in plant
tissues from the Swamp Gulch site compared to the background
site. For Salix boothii . this pattern only occurred for the
below-ground tissues. The concentration of Ca was 8580 ug/g in
the roots at the background site compared with 4535 ug/g at the
Swamp Gulch wetland. Calcium levels in the above-ground tissues
were similar at the background (6680 ug/g) and the Swamp Gulch
(6817 ug/g) wetlands, respectively.

Copper tissue concentrations were higher in above- than
below-ground material and were also higher at the Swamp Gulch
wetland compared to the background wetland (Table 9-6) . The
concentration of Cu was 3.6 (above-ground tissue) and 15.3
(below-ground) times greater in tissues from the Swamp Gulch
wetland than the background wetland. Salix boothii probably
contributed to the removal of dissolved Cu from the AMD as it
flowed through the wetland. Above-ground tissue levels for
Salix boothii were well below the phytotoxic level of 20 ug/g
for terrestrial plants (EPA 1987b) .

Iron concentrations were generally lower in the tissues of
Salix boothii than in Carex rostrata . In the above-ground
tissues for instance, the Fe level in the Swamp Gulch wetland
was 132 ug/g, which was notably lower than the 1532 ug/g in
tissue from the background wetland on Hardscrabble Creek. Iron
concentrations in below-ground tissues were higher in the Swamp
Gulch wetland compared to background tissue levels.

Lead tissue levels were lower in Salix boothii than in
Carex rostrata. The Pb was more concentrated in below- than
above-ground tissue and more concentrated in tissues from the
Swamp Gulch wetland than the background site. Compared to
phytotoxic values for terrestrial plants (>10 ug/g) , it is
unlikely that Pb caused toxic effects on Salix boothii (EPA
1987a) .

117

Manganese was also distinctly more concentrated in the
below- than the above-ground tissue, and was more concentrated
in tissue from the Swamp Gulch wetland compared to the back-
ground site. These patterns differ from those of Carex rostrata
where Mn levels were highest in the leaves and similar between
sites. The data suggests that Salix booth ii may have a greater
affinity of Mn than Carex rostrata . Despite the absorption of
Mn by Salix boothii . the concentration of Mn in the water was
not reduced. The influent water value was 3.74 mg/L and the
value at station F-1 was 4.30 mg/L (Table 9-4).

Salix boothii did not appear to accumulate Ni to a notable
degree. Tissue levels at the background site were below the
instrument detection limit of 1.0 ug/g, so values have been
estimated (Table 9-6) . Tissue concentrations at the Swamp Gulch
site were not phytotoxic.

Salix boothii responded to higher Zn levels in AMD at the
Swamp Gulch wetland by absorbing more. Tissue levels of Zn were
higher at the Swamp Gulch wetland than the background site.
Salix boothii tended to store this element in leaves and stems
rather than in roots. At the background wetland, Zn levels were
130 ug/g in the above-ground tissue and 105 ug/g in the below-
ground tissue. Likewise, Zn levels at the Swamp Gulch wetland
were 388 and 278 ug/g in the above- compared to below-ground
tissues, respectively. The Zn levels in these wetlands were
well below the 500 ug/g phytotoxic level for terrestrial plants
recommended for the Helena Valley (EPA 1987a) .

9'3.3.3 Effects of Betula alandulosa in Remediating AMD

Betula qlandulosa tended to absorb and accumulate metals in
much the same way as Salix boothii. In general, tissue con-
centrations for the analyzed metals were higher in the below-
ground than in the above-ground plant material. This was true
for both wetlands studied and for all elements, with the
exception of Mn and Zn (Table 9-7). These patterns were similar
for the accumulation of Zn in Salix boothii and the accumulation
of Mn in Carex rostrata .

With the exception of Al and Ca, plant tissue metal
concentrations were higher in the Swamp Gulch wetland than at
the background site. Aluminum levels were similar between the
wetlands, while Ca values were lowest in samples from the Swamp
Gulch wetland. The relatively low tissue metal levels make
phytotoxicity of Betula alandulosa unlikely.

9.3.3.4 Effect s of Brvophvtes in Remediating AMD

Metal levels were determined for two mosses, Isopteryqium
pulchellum and Sphagnum tenellum . Isoptervaium pulchellum was

118

Table 9-7. Concentration (ug/g) of elements in Betula glandulosa.

Element Matrix^

Al

As

Cd

Ca

Cu

Fe

Pb

Mn

Ni

Zn

Background
Mean fn=2)

AG
BG

AG
BG

AG
BG

AG
BG

AG
BG

AG
BG

AG
BG

AG
BG

AG
BG

AG
BG

35
141

0.352
0.35

0.352
0.35

6835
2150

0.352
7.1

73
205

0.7
5.5

81
39

0.72
0.7

185
56

Wetland
Range

Swamp Gulch
Mean fn=6)

Wetland
Range

29-41

28

12-270

151

0.35-0.35

0.6

0.35-0.35

0.7

0.35-0.35

0.5

0.35-0.35

1.4

5670-8000

3470

1600-2700

1808

0.35-0.35

2.8

0.35-14

15

57-89

158

120-290

1417

0.7-0.7

1.4

5.1-26

19

76-86

452

11-67

186

0.7-0.7

1.1

0.7-0.7

0.7

160-210

229

30-82

147

23-31
67-250

0.35-1.6
0.35-1.6

0.35-0.7
0.5-2.0

2730-4320
1070-3060

0.35-8.8
0.35-48

38-630
3510-26200

0.7-2.4
4.2-49

120-850
110-350

0.7-1.8
0.7-0.7

71-580
84-320

Ipiant material: AG=above-ground, BG=below-ground

2values below the instrument detection limit were estimated by multiplying
the instrument detection limit value by 0.7.

present at both wetlands, while Sphagnum tenellum was found only
at the Swamp Gulch site.

Table 9-8 presents tissue concentration data for various
bryophyte species in metal-enriched and non-enriched environ-
ments. The range of values for a particular element varies
widely due to the different species analyzed and because of the
different environments in which they were found. For some of
the elements (Fe, Pb, Mn, Ni and Zn) , the range of values for
the non-enriched areas overlaps the range for the enriched Table

119

US

o
a

00 ■»> f-i

o

IS

r-i

o
o

c

IS

o.

I 01

I |j4

I o>

IS

c

IS
4J
C
O

ID
C

ID IS
■D »

a
u

o

IS
IS

a
o

•p
c
v

e

SI

rH II)

a e
o c

e

S) -r^

J3 !>
O C

a

■o

01

o
•a -n

OJ M

j= c

U 01

•rt I

u c
c o
u z

0)

J3
O

01

o

o

B
O

e

3
>■

M

ja
o

■u
a

0)

0)

0)
0)

u

01

m

at M

>i 0)

U M lO

u £ 01

a> 01 ■<->

a » s

■D
01

o

I

C 10

o -u

c c

o>

w e

c

HJ O
M

01 ,-1
D> >
C B
IS 0)

IS

c 0)

• IB c

01 B

•p -

r^ . w ^

0) IS 01 cu
e 01 o

n u B >o

IS ••-t rH

-I > o •

Z B O CO

IB M e .C

IB ^ Oi O 4J

M U C

E a 01 M-i o

o g B

W E -rf B

4-t IB 4J IS ca

-ri 0) rH

^ C M U

3 -H IS 4J U

£ U O
r-fO C <W

r^ B 0) 3

IB IS i: o Q

M-l »H 4J lO S

>i 0) B n

a 3 -H -O O
E JJ

■O rH >, IB

0) IB 0> <H lo

,E M > 4J 0)

U 3 U 01 U

•r. 4J 3 3 O

M IB (0 O4

E B rH X

01 rH IB OJ
IS IS U

m B 3 -O

■O 6 O iJ B

E O -rH (B IS

IS U 0> B rH

rH >M 0) -P

4J M B 0)

01 u 03

• 3 U) IS M

IQ o U-i — 4

OrH e g

IS o

■O M M U 01

B 3 01 «H ID

IS u x: -n

O

-rl
M (0

E -P

0) B

0)

" e

O B
>H O
M
OJ -n
D> >
B B
IS 0)

n p e

n XI E IB 3

0) fX 0) u

•rt e >W -rl

■a - o B

3 E U -a IB

•P M >M E rH

U IS rH

^ 4J a)

U) -P IS •01

0) -H 0) 04 IS
■n 3 o, m E
o

ai-a e
a 01 3

CO J3 B _

P -rl IS

3 u x:

O H

IS u

a IS

o B Ql Q D x:

(n 0) CO CO CO tK „

rH r4 m -«i< m \D

D<0 I

E

- a

120

areas. Despite this apparent anomaly, these data indicate the
general range of element concentrations for mosses.

With the exception of Mn, the element levels in
Isoptervaium pulchellum (Table 9-9) were within, or close to,
the range for other mosses from enriched environments (Table 9-
8) . Data from previous studies indicate a high level for Mn of
412 ug/g in Sphagnum spp. from an artificial wetland that
received AMD (Wieder et al. 1985). In contrast, Isoptervaium
pulchellum accumulated 1423 ug/g of Mn at the Swamp Gulch
wetland. This difference is likely due to the inherent
abilities of Isoptervaium pulchellum to accumulate more Mn than
the Sphaanum species studied by Wieder et al. (1985). However,
this difference may also be due, at least in part, to the length
of time that the two mosses were exposed to AMD. The Sphaanum
moss was exposed over a 10 month period, whereas Isoptervaium
pulchellum had been growing at the Swamp Gulch site for years.
In any case, Isopteryaium pulchellum apparently can absorb and
store relatively large amounts of Mn.

With the exception of As and Ca, all elements were elevated
in Isopterygium pulchellum at the Swamp Gulch wetland compared
to background levels (Table 9-9) . This result was expected for
As since it was not elevated in the influent water (Table 9-4) .
The pattern for calcium was similar to that for the other plant
species since the concentration was lower on the Swamp Gulch
compared to the background wetland.

Tissue concentrations of all the elements except Ca were
lower in Sphaanum tenellum than in Isoptervaium pulchellum at
the Swamp Gulch wetland. The concentration of elements in
Sphagnum tenellum were within the range or were lower than the
range for other mosses from enriched environments (Table 9-8) .

9.3.4 Summary of Element Enrichment in Plants

Table 9-10 lists those plant species analyzed at both
wetlands and the elements that were enriched above background.
Iron, Cu, Pb and Zn appeared to be enriched in all species. The
moss, Isopteryaium pulchellum . accumulated all elements except
Ca. Significant phytotoxicity of any of these elements is
probably not occurring at the study site. Plant enrichment of
Cd is of particular concern because the threshold level for safe
intake of the element by animals is relatively low. This
potential problem is discussed in subsequent sections.

121

to

4J

a
o
>^

M

X5

W
-P
C
QJ

e

o

en

\

cn

D

C
O
-1-1
4J
(0

4J
C

u
c
o
u

I

0)

I-l

(0

0)
C
<U
4J

e

a
c
en

(0
CO

0)
"D en
c c

(0 (0

-p

0)

U

H

II

a, c
e ^

(0

S c

w (0

QJ

u

rH

D

CM

g

D

•H

>i

M

-P

a
o

CO

"U en
c c

(0 (0

5:

s:
o

i-H

D ->
O ^
II
Cb c

e —

(0

s c

Q)

c

(0

-p

0)

s —

73 II

c c

D ^
O

W C
en (B
>: (u
u s

4J

c

CD

e

1— I
W

S >^ r-l

{S • •

■^ rH CN

III

tSt Si Si Si Si
CM • • CM

CM

<S Si

00 CM rH S> V£> S

CO s ^ • r- '00

r~ iH rH CM CM r~ <N

I I I I I I I

Si cm SirH C<~)

CO • rH r- "sf

Lf) CO \0 •>a< rH

ro s

SlCMlDSlSlSlSSlCMS!

SlrHrH'<!t<<TiS!r--(S'J'CM
r-t in '* rH CTl VO

f^ -^ rH rH

Si s
S S) S) Si S

IT) CM VO S IT) ■^ CM
■^ •rHVOCMrHinCMrHSl

I I I I I I I I I I

SI in SI Si in S CM CNrH •^
CM • •CTirOSlrHrHr^'^
LD rH rO CT> • CM •

rH Si VjD S

Sicj^sisisisissiirts

ooror-cTicNinsirovD"*
<-\ VO ro rvj in CM r^

in s rH Si '5P s

r-i rH <Tl rH rH

CM

svor^sssiSisicMsi

snssiinst^S!

VD Sl rH 00 r~

r^ in Si rH

r^ CM

rHCO'OlODCUXlC-H

122

A-)

c
<u

e
a

M

-p

CO

c

<u

-p

en
c

-ri

>1

rH

a

-H
-P

rH

e

X!

■o

0)
■V
(0

e

•r)

4J
CO
<U

OJ
M
Q)

c
o

•H

4-)

D

<U
4J

0)
T)

4-1
C

d)

e

a •
-p •

CO Sl

c

-r1 >,

x: 4J

JJ -rl

e

O rH

<u c
^ o

CO 4J

Q) O

D <U

rH 4->

(0 <U

> 13

Table 9-10,

Plant species demonstrating element enrichment
(i.e. greater than background levels) at the
Swamp Gulch wetland site.

Plant Species

Enriched Elements

Carex rostrata
Salix booth ii
Betula glandulosa
Isopteryqium pulchellum

Al, Cd, Cu, Fe, Pb, Zn

Al, Cd, Cu, Fe, Pb, Mn, Zn

Cd, Cu, Fe, Mn, Pb, Zn

Al, Cd, Cu, Fe, Pb, Mn, Ni, Zn

9.3.5 Vegetation Element Levels and Wildlife Hazards

Concern has been expressed that the construction of
artificial wetlands for AMD abatement may result in high levels
of metals in the vegetation and that this would contaminate or
poison the wildlife resource. The following sections address
this concern by comparing animal hazard criteria levels with
plant tissue data demonstrating elevated levels of the target
elements.

9.3.5.1 Dietary Hazard Levels for Animals

There is a general lack of data on hazardous metal intake
levels for wildlife species. However, there is considerable
data for domestic animals. Table 9-11 lists the maximum
tolerable levels of minerals in diets of cattle, horses and
domestic rabbits. When these levels are fed for a limited
period, they will not impair animal performance and should not
produce unsafe levels in human food derived from these animals.
These threshold values were generated from a variety of studies
in which graded levels of the elements were ingested and the
specific affects examined. Highly soluble and purified forms of
these elements were generally used in the studies. This
practice tended to bias the threshold levels low. In other
words, a particular concentration of a metal in the diet may
exceed the threshold level but not be dangerous. In addition,
adverse physiological effects may not occur at a particular
threshold level if the element has been complexed into plant
matter during growth and is not digested by the animal. Animal
tolerance to chemical elements will vary with age and
physiological condition. For the purpose of discussion, these
data (Table 9-11) will be used as a general guide for potential
contamination of wildlife species utilizing the Swamp Gulch
wetland.

123

Table 9-11. Maximum tolerable levels^ of dietary minerals for

domestic animals (National Academy of Sciences 1980)

Species

Element

Cattle

Horse
(200)

Rabbit

Al

1000

(200)

As

50

(150)

50

Cd

0.5

(0.5)

(0.5)

Ca%

2

2

2

Cu

100

800

200

Fe

1000

(500)

(500)

Pb

30

30

(30)

Mn

1000

(400)

(400)

Ni

50

(50)

(50)

Zn

500

(500)

(500)

^ All values are in

ug/g unless

Otherwise spe

cified. Values in

parentheses were derived by extrapolation.

9.3.5.2 Wildlife Observed and at Risk

Wildlife observations were recorded during each field trip.
Voles (Microtus spp. and possibly Clethrionomvs aapperi ^ were
the animals most often observed. They were generally seen among
Carex rostrata where the soil was saturated or nearly so. One
cottontail rabbit f Svlvilacrus nuttalli) was seen in the center
of the site. The only ungulates actually observed were mule
deer (Odocoileus hemionus) . A total of six adult deer were seen
during the study period while a fawn and doe were observed
several times. Tracks of moose (Alces) and possibly elk f Cervus
canadensis) were seen twice. Their use of the area was minimal.
Evidence of beaver ( Castor canadensis ^ use of the site was low.
Beaver tracks were observed only once and freshly chewed willow
stems were seen on several trips. Beaver were probably moving
through the area but they were definitely not residents of the
site. Brook trout ( Salvelinus fontinalis ^ were often in the
larger channels in the southeast portion of the wetland study
area. An approximately eight inch brook trout was seen in a
channel near station AAA-3. Only very small trout were observed
m the Blackfoot River. Numerous frogs were observed throughout
the Swamp Gulch wetland. Coyote ( Canis latrans) tracks were
seen on only one occasion. Very few passerine birds were
observed. A redtail hawk ( Buteo iamaicensis ^ and presumably a
great horned owl (Bubo virainianus ) were observed once.

Animal species most likely to be poisoned by heavy metals
are those whose home ranges encompass a metalliferous environ-
ment. In the case of the Swamp Gulch wetland, these would be

124

primarily herbivores such as voles, muskrats ( Ondatra zibethica )
and beaver, and carnivores such as weasel (Mustela frenata) and
mink ( Mustela vison ) . Other possible residents of the study
site that could be at risk are shrews ( Sorex spp.) and the skunk
(Me phitis ) . Mammals and birds that feed over a large area are
not likely to be poisoned by occasionally consuming vegetation
or contaminated animals from the Swamp Gulch wetland. These
include ungulates such as elk, moose and mule deer, and
predators/ scavengers such as coyote, fox and various raptors.
This discussion will focus on those species with small home
ranges, meaning those more likely to consume large quantities of
vegetation or other animals from the Swamp Gulch wetland.

9.3.5.3 Hazards to Wildlife at the Swamp Gulch Wetland

Cadmium appears to be the only element present in the
macrophytes (Tables 9-3, 9-6 and 9-7) at levels that could be
toxic to animals. The bryophytes contain metal levels that
exceed maximum dietary intakes, however, these plants are not
consumed in quantities that could cause metal poisoning. Of the
macrophytes, Salix boothii had elevated Cd levels on the Swamp
Gulch wetland that greatly exceeded the maximum dietary intake
level. The Cd level in the above-ground tissues was 5.6 ug/g
which was significantly greater than the dietary tolerable level
of 0.5 ug/g (Table 9-11). Beavers consume large quantities of
Salix boothii . so could be at risk for Cd poisoning. If beaver
have been poisoned by consuming Salix boothii . it would help
explain the absence of resident beavers at the site. Beaver are
currently active in areas upstream and downstream of the Swamp
Gulch wetland and were once very active at the site.

In summary, Salix boothii on the Swamp Gulch site had a
level of Cd that exceeded the maximum tolerable dietary intake
for domestic animals. Beaver appear to be at particular risk
from Cd poisoning because of their heavy utilization this shrub.
Other animals, such as voles, may also be at risk. If these
small rodents are accumulating high levels of Cd then the
animals that prey on them (weasel, mink, owl) may also be
experiencing toxic effects. It is less likely that infrequent
visitors to the Swamp Gulch wetland such as elk, deer, moose,
coyotes, foxes and birds would be poisoned.

At this point, it is unknown whether true metallophytes
exist at the Swamp Gulch study site. If these types of plant
species are present, certain animals may be at special risk if
they consume these plants. Future work at this site should
include an examination of metal levels in tissues of the common
herbivore (s) . These data would help determine the extent to
which certain metals are present in the food chain.

125

10.0 MICROORGANISMS AND THE IMMOBILIZATION OF HEAVY METALS
IN WETLANDS

10.1 BACKGROUND

This portion of the study characterized microbial com-
munities and conditions associated with AMD in constructed and
natural wetlands using a new method for measuring the popu-
lations of microorganisms based on their unique biochemical
components. MSI Detoxification, Inc. (MDI) uses this technology
to characterize hazardous waste sites and proposed its
implementation in this study. This new method allows the
analysis of mixed microbial populations found in natural
environments and does not require cultivation of organisms in
the laboratory. Thus, it is a powerful tool which avoids
isolation of microorganisms by traditional cultivation
techniques, and therefore, may be more representative of the
microorganisms present in situ. As a part of the study of the
efficacy of natural wetlands in the remediation of acid mine
drainage, it was determined that the role of microorganisms,
specifically sulfate reducing bacteria, be investigated in the
unique circumstances presented by the Swamp Gulch natural
wetland in Lincoln, Montana. A reference for this wetland was
provided by the Sand Coulee wetland, a constructed wetland near
Tracy , Montana .

The hypothesis for the perceived success of natural
wetlands in remediating AMD considers that microorganisms have
the ability to carry out the reduction of in-flowing sulfate.
This, in turn, raises the pH of the system and produces sulfide.
The sulfide may be involved in reductions of ferric iron (Fe-^"*")
to ferrous iron (Fe2+) and subsequent immobilization, possibly
as inorganic or organic sulfides. If the process (or the
process initiator) is microbial rather than chemical, then
sterilized soil samples should not cause the pH of AMD to rise
after incubation under anaerobic conditions.

To test this hypothesis in the Swamp Gulch environment,
experiments were performed under which pH, sulfate, iron, and
microbial biomass would be measured over time and related to AMD
remediation. The experimental approach was divided into three
Phases. Under Phase I, the microbial component of the Swamp
Gulch AMD remediation activity was investigated through a
comparison of remediation capacities in sterilized versus
untreated core materials (soils, organic detritus, inorganic
materials) . Phase II sought to verify the presumed microbial
activities observed under Phase I by following pH, sulfate, and
iron in wetland core materials in a controlled greenhouse
environment. Under Phase II, materials from the Swamp Gulch
natural wetland were compared with sediment core materials taken
from the Sand Coulee constructed wetland. Phase III focused on
the microbial component of the wetlands and compared biomass and

126

(inferred) microbial diversity in field samples extracted from a
natural, constructed and control wetland (Hardscrabble Creek
near Lincoln, Montana) .

The characterization of the major types of microorganisms
in artificial and natural wetlands and their correlation with
effectiveness of AMD remediation are intended to allow us to
develop rapid and effective methods to diagnose and renovate
malfunctioning wetlands.

10.2 PHASE I: SWAMP GULCH WETLAND BUFFERING CAPACITY AND THE
EFFECT OF SOIL STERILIZATION ON WETLAND PH

10.2.1 Introduction

A preliminary experiment to compare the abilities of Swamp
Gulch wetland soil samples to remediate AMD under various
conditions was performed. The purpose of this experiment was
two-fold: 1) to gain insight regarding the buffering capacity
of Swamp Gulch soil on exposure to AMD, and 2) to determine if
there was evidence that microorganisms beneficially raise the pH
in Swamp Gulch wetland soil.

10.2.2 Technical Approach

Soil samples from the Swamp Gulch wetland were incubated
with AMD under aerobic and anaerobic conditions. Initial pH
values of the soil-AMD mixture and those of the same suspensions
after incubation for nine days were compared.

10.2.3 Materials and Methods

10.2.3.1 Soil Sampling

Soil samples (0.5 and 2.0 g) were obtained from cores taken
from the Swamp Gulch wetland (Figure 10-1) . Each sample was
placed in a plastic "Whirl-pak" bag and transported to the
laboratory where it was transferred to a Waring Blender and
homogenized for a total of 60 seconds. Soil samples (0.5 g and
2.0 g) were weighed into tared 30 ml beakers (titration experi-
ment) or Hungate roll-tubes (microbiological experiment) ,
respectively.

10.2.3.2 Titration Experiment

To determine soil buffering capacity, duplicate 0.5 g soil
samples suspended in 10 ml of distilled water were titrated with
AMD. Readings of pH were made for six minutes or until a stable
value was obtained (< ten minutes) . Similarly, AMD was titrated
into distilled water for comparative purposes. AMD titrated
into distilled water was not replicated. All samples were

127

Kalispell

Swamp

Quich ,

Natural

Wetland

Havre

MONTANA

Hardscrabble
Creek
Control \
Wetland \

Lincoln
Missoula
I Deer Lodge •Helena

Glasgow

.Great falls
. Tracy

•Lewistown

Sand

Coulee

Constructed

Wetland

. Anaconda

butte

Bozeman . .Livingston .Billings
Laurel

Sidney .
Glendive.

.Miles City

Figure 10-1. Location of wetland sites.

continuously mixed using a magnetic stirring device during the
titration.

10.2.3.3 Microbiological Experiment

In the microbiological study all sampling was carried out
in triplicate and three treatments were considered: a sterilized
control, an aerobic and an anaerobic treatment. All samples
were incubated under ambient laboratory temperature conditions
for a period of nine days. For the sterilized control, soil
samples (2.0 g) suspended in 14 ml water or 14 ml AMD were
sterilized on three consecutive days by autoclaving for 15
minutes at 15 psi. Anaerobic treatment samples were gassed with
Oo-free N2 and stoppered with Hungate stoppers. Aerobic
treatment samples were incubated in 125 ml Erlenmeyer flasKs
with loose fitting caps and incubated with vigorous shaking.
Readings of pH were made initially and after the incubation
period.

After incubation and pH measurement, triplicate samples of
each treatment were pooled to provide sufficient sample for iron
analysis. The pooled samples were centrifuged at 10,000 x G for
ten minutes and the supernatant fraction decanted. This

128

fraction was acidified with concentrated HCl and total dissolved
iron determined by the phenanthroline color imetric method (APHA
1985, Iron Determination, pp. 215-220).

10.2.4 Results and Discussion

10.2.4. 1 Titration Experiment

Figure 10-2 depicts the titration of wetland soil
homogenate and distilled water. After a volume of AMD was added
to the stirred soil suspension, the pH fell to a minimum of 4.5
during the six minute measurement period. This indicates that
the soil has a capacity to adsorb acidity independent of
microbial involvement, since the experiment was conducted on
time scales too short for significant microbial activity.

J-0-2-4.2 — Microbio logical Experiment

The incubation under anaerobic conditions of AMD soil from
the Swamp Gulch wetland raised the pH of the AMD by 0.72 units
(Table 10-1) . When the soil and AMD were sterilized, no such

Sample Sources: Swamp Gulch Soil & AMD Water

PL,

12345G70 9

Volu me AMD Added, niL

Figure 10-2. Titration of Swamp Gulch soil with AMD.

129

10 11

Table 10-1. pH and iron values^ for AMD incubated under various
treatment conditions after a period of nine days.

Pre Post
Treatment Incubation pH Incubation pH

Total
Net Dissolved
Difference Iron. mq/L

Aerobic
Anaerobic

Sterile
Control

7. 08+0. 03a
6. 28+0. 03b

6. 35+0. 05b

5. 17+0. 13c
7. 00+0. 05a

-1.92±0.10a
+0.72+0. 06b

162.9+2.5ab
20. 7+0. O'^

6. 37+0. 10b +0.02+0.06c3 172. 8+5. la

^ Means + standard deviation; n = 3 for pH data; n = 2 for total

dissolved iron data.
2 Treatment means (within columns) are significantly different

(P=0.01) if followed by a different letter.
^ The sterile control pH means from preincubation to postincu-

bation were not significantly different from one another.

increase in pH took place. Anaerobic conditions were necessary for
this process. When AMD was incubated aerobically, the pH fell by
1.92 units and the soil became bright orange, presumably due to the
production of iron oxides. In all anaerobic incubations the soil
maintained its dark brown/black color which would be expected if
metal sulfides were being produced. Comparison of iron
concentrations in the presence and absence of oxygen indicate that
there was much less iron in the anaerobically treated sample.
Presumably this anaerobic response is the consequence of microbial
sulfate reduction to sulfide followed by the precipitation of iron
sulfide, which is black in color. To verify that sulfides were
present the soil was treated with 2N HCl and an H2S odor was
detected.

The Swamp Gulch wetland soil was able to raise the pH of the
applied AMD from 2.5 to about 5.0 in the absence of microbiological
contribution (titration experiment) . As more AMD was applied to
the soil, this dropped to 4.5 (Figure 10-2). Because the pH scale
is logarithmic, this represents a 50 to 100-fold increase in the
ability of the soil to neutralize acidity (as hydrogen ions).
Aerobic incubation resulted in a final pH (near 5) which was
similar to that of the soil titration experiment. However,
anaerobic incubation raised the pH to 7.0 (4.5 pH units) indicating
that wetland soil combined with its indigenous anaerobic
microorganisms was able to neutralize AMD acidity by a factor of
greater than 10,000-fold.

130

These findings support the hypothesis that sul fate-reducing
bacteria play a notable role in the remediation of AMD. Further
evidence in support of the involvement of sul fate-reducing bacteria
comes from the phospholipid fatty acid analyses of core samples
taken from the same area of Swamp Gulch wetland (Phase III) .

10.3 PHASE II: COMPARISON OF NATURAL AND CONSTRUCTED WETLAND
SOILS IN THE REMOVAL OF ACIDITY, SULFATE AND IRON

10.3.1 Introduction

Biological oxidation of reduced sulfur due to the activities
of sulfur oxidizing bacterial action on pyritic ores is the primary
cause in the formation of AMD waste water problems. The resulting
products in the waste are acidity, sulfate ion and iron. The
sul fate-reducing bacteria remove sulfate by reducing it to sulfide,
which can cause the co-precipitation of heavy metals as sulfides.
The objective of these experiments was to compare acidity, sulfate
and iron removal in a natural and an artificial wetland system. In
addition these studies were carried using samples of sufficient
size to minimize inhomogeneities which would be present in small
samples.

Iron, sulfate and pH measurements were included in the
experimental protocol, as microbial sulfate reduction is the
presumed basis for AMD remediation through immobilization of heavy
metals. Iron was chosen as an indicator metal due to its ease of
determination and its ubiquitous presence in the wetlands under
study .

10.3.2 Technical Approach

The study was carried out in a greenhouse at the Montana State
University Plant Growth Center. Three of the cores were obtained
from the Swamp Gulch natural wetland near Lincoln, Montana (Figure
10-1). These were designated SG-1, SG-2, and SG-3 for this study.
The other three cores were obtained from the Sand Coulee
constructed wetland large cell near Tracy, Montana (Figure 10-1) .
These cores were designated SC-V, SC-VN, and SC-CH.

The cores received AMD that had been collected at the Swamp
Gulch site. AMD was eluted through all of the cores for the
duration of the experimental study. The effluent (eluate) as well
as the influent AMD (eluent) was analyzed for temperature and pH
several times a week. Sulfate and iron concentrations were
determined periodically over the course of the study.

Direct microbiological parameter measurements were not made
under this Phase of the study. Microbial diversity and biomass
analyses comprised phase III of the investigation.

131

10.3.3 Materials and Methods

10.3.3.1 Field sampling

A site was selected at the Swamp Gulch wetland and was sampled
to provide sediment cores for the greenhouse study. The selected
site was between sites C3 and C2 (see Figure 4-1). It was assumed
that the remediation capacity of the Swamp Gulch wetland was not
saturated at this field location. Site vegetation was dominated by
Carex with subsidiary Sphagnum moss.

Three sediment samples (designated SG-1, SG-2 and SG-3) were
collected on October 22, 1987. Each core vegetation mat was sawn
along the sides and bottom with a keyhole saw, and removed from the
wetland using a flat-bladed shovel. Each core was laid on its side
on the ground surface and trimmed to fit tightly into the sediment
container. The containers, five-gallon plastic buckets modified as
shown in Figure 10-3, had been acid-washed and disinfected with
chlorine bleach solution before they were taken to the field.
After pressing the sediment firmly to remove air spaces along the
container walls, native water from the sampling trench was poured
over each sample so that standing water was visible just below the
core top. A five-gallon container of the in situ water was
collected so that it would be possible to replenish sample water in
case of losses during transport.

Sediment cores were transported to Bozeman the following
afternoon. The next morning, samples were sprayed with MAVRIK
Aquaflow, a broad-spectrum insecticide (active ingredient =
(aRS, 2R) -f luvalinate [ (RS) -a-cyano-3-phenoxybenzyl- (R) -2-{2-
chloro-4-trifluoromethyl) anilino}-3-methylbutanoate] ) .
Fumigation, required of users of the MSU greenhouse facility, was
carried out to prevent the spread of disease among experimental
plants. It was believed that this treatment did not affect the
microbial or chemical integrity of the samples due to its focused
surface application.

Samples from the Sand Coulee site near Tracy, Montana were
collected on October 30, 1987, and placed into five-gallon
containers in an identical manner as were those collected at the
Swamp Gulch wetland (Figure 10-3). The three samples were
collected about one meter from the middle baffle of the first cell
of the wetland, in an area with sporadic Tvpha plants and pools of
standing water on an irregular sediment surface. Such vegetation
cover was typical for that cell (Figure 10-4). The absence of a

132

23.5 cm

Sampling
ports

False >
bottom

O

•> O

O
21.3 cm

]

33 cm

All joints sealed
with silicone.

Three sampling ports
on side.

Glass wool covers
the inlet of the
effluent port.

2.5 cm sand on bot-
tom, below cone,
above false bottom.

Clamp —
on Tygon
tubing

Effluent
port

PH

XI

m /\

Inorganic
Analyses

Figure 10-3.

Diagram depicting soil core containers used ih
Phase II greenhouse study.

133

Outiel Weir

Aerotion Slruclure

Jnlel Weir

•Limestone Chonnel
PLAN VIEW

Bertn

I'-CottoilSod Mat

23 m. Peot

45m. Limestone Mix

Bentonite Liner

Figure 10-4.

Sand Coulee constructed wetland sampling site
(Kiel and Kerins 1988) .

vegetative mat and the high water level in the artificial wetland
precluded removal of a coherent core. Instead, the saturated
sediment was shoveled into the container, preserving the depth
sequence of the peat. This procedure was deemed acceptable for the
Sand Coulee wetland samples because the sediment color and texture
were homogeneous throughout the peat profile,- and because the
sediment was fluid enough to fill the container without forming
pockets of air.

134

Cores were extricated from the Sand Coulee constructed wetland
large cell area (designated "SC" cores) for a variety of treat-
ments. The core designated SC-V contained cattails which were
cropped for this study. The core designated SC-VN represented a
sample void of vegetation during sampling. The core designated
SC-CH contained cattails and was treated with chloroform for a
period of about two months prior to beginning this study at which
time the chloroform addition was halted. The intent in adding
chloroform was to chemically sterilize the core. The Sand Coulee
constructed wetland cores were treated as replicates and studied in
the same manner as were the Swamp Gulch wetland cores.

10.3.3.2 Soil core flow characteristics

The effluent flow rate from each of the cores varied,
presumably due to flow characteristics unique to each wetland
sample. Flow rates were easily controlled for most cores.
However, the effluent flow rate of the SC-V core decreased after a
period of time so that less than 200 ml per day would freely flow
from the core unless extracted using a vacuum device. This
impediment of flow in the SC-V core occurred 26 days into the study
and the average overall flow rate dropped to 125 ml/day.
Consequently, its behavior during the first 26 days was similar to
the five other cores and subsequently was extremely slow (57
ml/day) . The flow rate for all other cores ranged from 253 to 291
ml/day.

10.3.3.3 Sampling and preparation of AMD

AMD was collected from the Swamp Gulch natural wetland site by
siphoning AMD from a receiving pond, located adjacent to the
wetland and on the north side of Montana Highway 200 (see Figure 3-
1) . The collected AMD was stored in plastic 55-gallon drums that
had previously been acid-washed with dilute HCl and rinsed several
times with distilled water. The 55-gallon drums were stored closed
in the greenhouse.

As AMD water was needed for the experiment, it was siphoned
from the 55-gallon drums and filtered through glass wool. The
filtered AMD was stored in 5-gallon carboys under cover to avoid
photooxidation resulting from exposure to sunlight in the
greenhouse. This filtered AMD was used as the eluent for both the
Swamp Gulch and Sand Coulee wetland cores . Temperature and pH
determinations were made on the AMD eluent (influent) and eluate
(effluent) collected from each core, every day effluent was
sampled.

10.3.3.4 Sulfate and iron determinations

Sulfate and iron concentrations were determined several times
during the course of the study period. Dissolved sulfate
determinations were made on days 61, 82 and 88 of the study.

135

Dissolved ferrous and total iron measurements were made on days 42,
61, 82 and 88. The schedule for these analyses was based on an
estimate of the percent of one pore volume of water which had been
eluted from each of the soil cores. The percent water contained in
the soil was determined to be approximately 85% based on percent
moisture determinations. Thus, the water contained in each of the
core was estimated to be 16.9 L. The core effluent was sampled
when approximately 13.5 and 20.3 L had passed through each core and
again at the end of the study. Due to the innate low flow rate of
the SC-V core it was not possible to obtain iron and sulfate data
beyond 12.2 L of flow- through volume.

Sulfate concentrations were determined by ion chromatography
(subcontracted to Camas Laboratories, Inc., Missoula, Montana)
using standard methods of analysis (APHA 1985, Sulfate Determina-
tion, pp. 483-488) . Two-hundred mL aliquots of eluate for sulfate
analysis were collected from each core and filtered through
cellulose acetate membrane filters (nominal exclusion 0.45 um) .
All of these samples were refrigerated (4°C) overnight, packed on
ice in a cooler and transported to Missoula, Montana, for chemical
analysis. All sulfate analyses were performed within one to three
days of receipt of the samples.

Iron analyses were performed using the phenanthroline
colorimetric method (APHA 1985, Iron Determination, pp. 215-220).
Eluate samples were collected for iron analysis at the same time
and in a similar manner as were the eluate samples for sulfate
analysis. For dissolved ferrous iron determinations the core
eluates and AMD eluent were collected directly into HCl to acidify
them immediately and minimize oxidation of the ferrous iron. All
dissolved iron samples analyzed were determined on the same day as
they were collected; the dissolved ferrous iron samples being
determined within nine minutes of collection, and the total
dissolved iron samples deteirmined within eight hours of collection.
It should be noted that the collection of the samples from the SC-V
core, which did not flow freely, was performed under a vacuum
pressure .

10.3.4 Results and Discussion

10.3.4.1 Acidity determinations

The temperatures of the influent AMD and the effluents from
the cores were generally in the 18-2 1°C range. The influent AMD pH
ranged from 2.17 to 2.70, with a mean and standard deviation
of 2.44 + 0.09 (N=94). This indicated that AMD applied to the
various cores remained stable with respect to pH over the study
period (Figure 10-5) .

The Swamp Gulch wetland core eluates exhibited pH values with
nearly identical trends indicating that they were replicates
(Figure 10-6) . The effluent pH values for the Swamp Gulch wetland

136

cores ranged from 5.67 to 7.58, with an overall mean and standard
deviation of 6.43 + 0.37.

The Sand Coulee wetland core eluates, while not true
replicates, demonstrated some important trends. Eluate pH values
from cores SC-V and SC-VN were similar (Figure 10-7). These pH
values dropped at the start of the study, flattening out by 30 days
into the experiment and ranged from 2.39 to 4.25, with an overall
mean and standard deviation of 3.24 + 0.45. The Sand Coulee
wetland core which had been treated with chloroform, SC-CH, showed
an ascent in its pH for the initial 30 days of the study and a
leveling of this pattern subsequently. The SC-CH eluate pH ranged
from 4.25 to 5.73 with a mean and standard deviation of 5.12 ±
0.39.

3.00

ffi

Ph

2.80 -

;-4

CD

•^

cd

2.60-

Q

<

2.40-

Oh

2.20-

1-H

2.00-

100

Time in Days

Figure 10-5.

Influent (eluent) AMD pH measured over study
period.

137

B.OO

7.00--

6.00--

^ 5.00 +
4.00--
3.00--
2.00

B.OO

- — Replicalion 1
• ■ RepUcaiion 2
— ReplicBiion 3

25 50 75

Time in Days

100

2.00

50 100 150 200

Estimated % Pore Volumes Eluted

Figure 10-6.

Swamp Gulch soil core's effluent (eluate) pH
measured over study period.

13 8

8.00
7.00
6.00

4.00

3.00--

2.00

Chloroformed Treatment

- Vegetation Treatment
No Vegetation Treatment

B.OO

25 50 75

Time in Days

100

50 100 150 200

Estimated % Pore Volumes Eluted

Figure 10-7.

Sand Coulee soil core's effluent (eluate) pH
measured over study period.

139

For the SC-V and SC-VN cores for most of the study period,
Sand Coulee constructed wetland soil raised pH of applied AMD from
2.5 to about 3.0. It is questionable whether there was any acid
neutralizing microbial component in these cores. Consequently, the
0.5 increase in pH may reflect the innate acidity neutralizing
capability of the Sand Coulee wetland sediment.

It can be surmised that the application of chloroform to the
SC-CH core influenced the rise in pH. Although the eluate pH of
the SC-CH treatment never achieved the near-neutral pH (6.43 +
0.37) typified by the Swamp Gulch wetland cores, it demonstrated
remediation of acidity not evident in either of the other Sand
Coulee wetland cores.

Two hypotheses can be advanced which would explain this
observation. Both of these are based on the principal of microbial
selection. In the first, a population of microorganisms arises
because the chloroform is lethal to most of the indigenous
organisms. A population is selected due to exclusive effects
imposed by the chloroform on most of the organisms. As this group
of microorganism proliferates, its metabolic activities result in
increased pH and amelioration of the acidity. In the second
hypothesis, selection occurs through nutritional enrichment of a
microbial population able to use the chloroform as a source of
carbon and energy. Because they are able to use the chloroform,
they come into dominance. Presumably they had the ability to
remediate the AMD pH prior to the addition of the chloroform, but
were unable to because they were starved for a carbon source.

Because there was no apparent lethal effect of the chloroform
in the SC-CH core, and because the cattails in this core did not
die, the second hypothesis is more probable. Also, the latter
hypothesis is supported by remediation of AMD pH in the Swamp Gulch
wetland cores where a high level of detritus from decaying sedges
and other plant material was present. Conversely, the Sand Coulee
wetland cores lacked conspicuous amounts of detritus which would
provide carbon and energy and allow indigenous microbes to
proliferate. These findings indicate that the constructed wetland
is ineffective in removing acidity, presumably because it lacks
sufficient carbon to support indigenous microorganisms. Therefore,
it should not be automatically assumed that poor performance in a
constructed wetland is due solely to wetland size. This scenario
is supported by observations on the phospholipid analysis of the
Sand Coulee wetland in situ site (Phase III) .

10.3.4.2 Sulfate determinations

The influent AMD applied to the cores exhibited a sulfate
concentration of 412 ± 56.1 mg/L, ranging from 336 to 482 mg/L
(Figure 10-8) . The Swamp Gulch wetland cores are remediating the

140

C30

tl

>
O

m
m

45

35"

d 25
o

u

15-

w 5-

-5

■Total Iron
Ferrous Iron

•A- A

25

50

75

100

3000

25 50 75

Time in Days

100

Figure 10-8.

Influent (eluent) AMD iron and sulfate
concentrations measured over study period

141

added sulfate (Figure 10-9) . The eluate concentrations for the
Swamp Gulch wetland core vary from to 158 mg/L with a mean and
standard deviation of 59.5 ± 56.2 mg/L. In general, the ^^ „,
remediation was better as the flow rates were slower (Table 10-2).

Sulfate concentrations in the effluent increased as the flow
rates were increased. This could be associated with the rate at
which indigenous sul fate-reducing bacteria are able to facilitate
sulfate reduction to sulfide, where greater retention times provide
for greater sulfate reduction. This phenomenon would be
particularly important where sul fate-reducing bacteria were either
not present, present at low concentrations, or present, but not
active (Phase III) . For example, in a constructed wetland if flow
rates are high and/or channeling is evident, one would expect^poor
sulfate-reduction performance to occur. Therefore channelization
and high flow rates in constructed wetlands should be avoided by
implementation of improved engineering design.

Instead of remediating the sulfate concentrations the Sand
Coulee wetland cores are releasing sulfate into the effluent (Table
10-3 Figure 10-9) . The effluent concentrations for the Sand
Coulee wetland cores range from 1385 to 2660 mg/L. Presumably this
is because the in situ addition of sulfate at the Sand Coulee
wetland site is much higher then that in the Swamp Gulch AMD input
water, and it is beginning to be removed as the lower
sul fate-containing Swamp Gulch AMD is being applied to the cores
(Figure 10-9) .

The flow rates for these cores (except SC-V) increased during
the later part of the study, as they did for the Swamp Gulch
wetland cores, and there was a subsequent decline m sulfate
concentrations in the effluent, suggesting that leaching of excess
sulfate had occurred in these cores. These high levels of sulfate
suggest that sulfate reduction was not taking place in the Sand
Coulee wetland soils. This could be due to the absence of
sufficient carbon to allow the sul fate-reducing bacteria to
proliferate (Lovley 1987). It could also be due to the acidity of
the cores as sul fate-reducing bacteria do not grow well at pH
levels <six (Widdel and Pfennig 1984).

In the SC-CH core, chloroform may have killed the sulfate-
reducing bacterial population that was present. Alternatively,
chloroform treatment may have enriched other microorganisms,
effectively displacing sul fate-reducing bacteria. Since the other
Sand Coulee wetland cores did not reflect lower sulfate levels, it
may have been that sulfate-reduction was not occurring as a
consequence of unknown factors.

142

o

o

Q

O

o

o

o

o

O

o

o

o

o

1(3

O

in

o

in

n

M

M

•H

o

o

o

o

o

o

o

o

o

o

o

o

o

in

o

in

o

in

n

N

01

«N

^4

B

CO

El]

l/§m '_tos psApssiQ

q/§ni '^tos psAiossiQ

o

o

o

o

o

o

o

o

o

o

o

o

o

in

o

in

o

in

CO

CVl

Ctf

o

o

o

o

o

o

o

o

o

o

o

o

o

in

o

in

o

in

n

CM

N

«*4

•^

Ed

l/Sui '„^0S psApssiQ

q/§ui '_tos psApssTQ

>i

•o

d

+j

U)

■p

c

M

o

QJ

3

>

1— 1

o

iw

M-l

Ti

(1)

Q)

u

Q)

3

U

to

o

ro

o

(U

g

rH

-f-H

m

o

c

CO

o

•r-H

<D

+J

QJ

m

rH

S-l

D

4J

O

c

U

(U

o

■o

c

C

o

fO

o

w

d)

■n

jj

c

(0

(0 M-(

rH

^

D

o

cn

1-1

D

^-v

O

0) •

A-> T)

CU (0 O

g

3 -H

(0

rH U

^

0) d)

w

— a

»

<Tt

cs

rH

0)

M

D

cn

bj

143

Table 10-2. Swamp Gulch sulfate concentrations vs. flow rate.

Core Flow rate^ Mean sulfate^

Number fmL/dav) (jm/Id

SG-1 219.3 35.3 ± 3.0

401.7 12.7 + 3.6

472.7 33.9 ± 2.5

SG-2 213.5 0.0 ± 0.0

255.7 49.0 + 6.2

596.4 130.5 + 9.2

SG-3 198.7 6.2 ± 5.7

424.3 152.5 ± 7.8

441.4 115.5 + 2.1

1 n = 1

2 n = 2

Table 10-3. Sand Coulee sulfate concentrations vs. flow rate,

Core Flow rate^ Mean sulfate2

Number fmL/dav) (mg/L) _

SC-V 144.2 1399.0 ± 19.8

82.4 2425.0 ± 173.9

91.4 2540.0 ± 169.7

SC-VN 243.9 1450.5 ± 26.2

330.4 2353.5 ± 72.8

584.3 1585.0 ± 35.4

SC-CH 172.0 1588.0 + 25.5

468.3 2245.0 ± 59.4

340.0 2050.0 ± 169.7

1

n = 1

2 n = 2

144

These findings indicate that sulfate remediation at the
Sand Coulee constructed wetland was not occurring. In order to
optimize wetland performance, construction specifications should
favor the proliferation and activities of sul fate-reducing
bacteria. This means that:

° Anaerobic conditions must prevail within the sediment.
Therefore slow flow rates and microbial population
diversity enhancement is important. Varying wetland
depth would also be beneficial.

° Hetabolizcible carbon and other nutrients should be
enhanced. This would include providing living plants
and fertilizer. This would also increase the con-
centrations of aerobic microorganisms (increased species
diversity) whose activities would lead to more favorable
conditions for the anaerobic sul fate-reducing bacteria
(by removing oxygen and providing alternate carbon
siibstrates) .

° For the sul fate-reducing bacteria to proliferate, the pH
of the constructed wetland should be near six or
greater. Therefore it would be prudent to select wet-
land construction materials that have an initial pH
near six or above. The use of lime in a constructed
wetland can help attain this goal.

10.3.4.3 Iron determinations

Iron analyses show a consistent total dissolved iron
concentration in the influent AMD of 10.2 to 16.9 mg/L,
averaging 13.3 mg/L (Figure 10-8). This iron was present in the
ferric form, which was not surprising, since the source of the
acid mine drainage was a shallow pond.

The presence of ferrous iron in the effluent of the soil
cores was not surprising, since the soil cores provide a
reducing environment. It was unexpected that no apparent iron
remediation occurred in any of the cores since all six of the
wetland cores eluted more iron than was added to them by the
eluent AMD. Conceivably the organic material acts as a cation
exchange column. In the case of the Swamp Gulch wetland soils
there appear to be hydrogen ions displacing metal ions at the
exchange sites. These three cores remained fairly stable in
their effluent iron concentrations over time. The eluate iron
concentrations ranged from 77 to 196 mg/L for the Swamp Gulch
wetland cores. These concentrations did not decrease over time
or over % pore volume eluted (Figure 10-10) , nor did flow rate
affect them.

In the Sand Coulee wetland cores the iron concentrations
averaged 203 + 136.6 mg/L. The iron concentrations of the SC-VN

145

o
■ o

e
o

o

I-

•o

>

n
n

« ca n

s S g

o o o

? ? -s

5 " ii

a S- £•

a v ■

£ BS BS

r-

CO

>%

ca

O

o

C

to

.«-rt

Q}

E

»^m

m

6-

N

o
o
o

o

o

o
a
CO

o
o

o
o

01

o

o
o

o
o
CO

o
o
CO

o
o

q/Sui 'Si

q/Sm 'ai

r-

M ca

a 6

«

n

n

0.

»
as

..O

m

O

u

« 1'

S •-'

ca

s s

3

"o

>

a O

— u

O

a.

u
09

o
o
o

o

o
CO

— (—

o
o

CO

o
o

— t—
o
o

CM

O

o
o

o

o

CO

o
o

CO

o
o

o

o

-fo — ;

o ca

Ed

q/Sui '3i

q/Sin '3i

>-l

(U

Tl >

0) O

>

^ TD

O <U

M W

to D

-rH W

"O <0

0)

'- g

(U

4J 03

ro c

3 o

i-H -H

<U -P

— m

M

4J 4J

c c

(U 0)

D o

-H C

>w O

IH O

0)

c

(U o

W >-l

O ■<-^

O

m

,-1 o

■rA O

O M

m u

»

Q) t3

jn ^

o

u

■l-l

--4 -a

U

a c

<D

o m

Q<

CbrH

>i

g <c -o

its 4->

3

5 O

4J

CO A->

CO

Si

I

IS

(P
u
3

en

•rH

146

(no vegetation) core effluent constantly declined (Figure 10-
11). On day 42, the concentration of iron in this core effluent
was 476 mg/L. By day 88, it had decreased to 35.5 mg/L. The
SC-V (vegetation) and SC-CH (chloroform treated) cores were
similar except that their iron concentrations were > 100 mg/L at
the end of the study. The SC-CH core had high iron concentr-
ations initially (450 mg/L) and began to decrease at a later
date than the other cores (day 62) . By the final iron analysis,
the Sand Coulee wetland iron levels had dropped by half.

Thus all of the Swamp Gulch and Sand Coulee soil core
eluate iron concentrations were higher then those of the
influent (eluent) AMD. However, cores from each of the sites
exhibit different trends. The Swamp Gulch wetland cores
released about the same concentration of iron over time,
irrespective of: 1) flow rate or, 2) percent pore volumes
eluted. The Sand Coulee wetland cores eluted extremely high
concentrations of iron early in the study, with a tendency
toward decreasing iron concentrations with time. This may be
related to flow rate or it may be a washing out of iron present
prior to the beginning of the study.

We suggest different hypotheses for these phenomena. In
the Swamp Gulch wetland cores no apparent iron remediation took
place. The iron seems to be released from the soil at a fairly
steady rate suggesting that cation exchange with acidity
(hydrogen ions) is occurring.

In the case of the Sand Coulee wetland cores, the iron
concentrations decreased over time. This could be indicative of
either, 1) leaching of iron by the Swamp Gulch AMD, 2) or these
cores may be remediating iron. Of these suppositions, it seems
most likely that iron is being leached. When the total effluent
volumes for the SC-V and SC-VN cores are compared with one
another (Table 10-4) , the iron concentration in the effluent
decreases as the flow through volume increases. In the SC-CH
core, the chloroform may have altered the iron remediation
properties .

These findings indicate that iron removal at the natural
wetland and at the constructed wetland sites was not occurring.
In order to promote iron removal in constructed wetlands,
materials should be selected that:

° Provide as much cation exchange capacity as possible to
remove acidity and metals.

° Provide favoredsle conditions for the proliferation of

sul fate-reducing bacteria. If these bacteria are able to
reduce sulfate to sulfide, metals will be immobilized
through to formation of metal sulfide precipitates.

147

q/§ni 'ai

q/Sm 'Si

e
o

u

01

S

o

u

b.

u
_>

"o
n
in

a

S i

».^

a

§ 3

o • >

2 ? 5

u > z

o
o

o

— t-
o
o

CO

to

c-

m

5^

d

a

o

G

Ifl

•0m

o

£

in

E-

CM

o

o

CO

o

o

o
o

CM

q/Stii 'ai

q/Sxu 'ai

u

0)

HD >

<U O

>

i-i "O

O Q)

m u

CO D

•H tn

t3 (0

(P

— e

0)

4J TO

(0 c

D O

r-l —1

CU +J

>-' (0

M

-P 4J

c c

0)

3 U

rH C

y-i o

iw o

0)

c

Q) O

w w

O -H

u

CO

r-l D

-rH O

O M

(A U

•

<U -O

0) iw

o

0)

■rt

r^ T3

M

D C

0)

O nj

a

U

iH

>i

"O "0 "O

C -tJ

p

(0 o

4J

CO JJ

CO

I

SI

cn

•i-H

Cm

148

Table 10-4. Sand Coulee iron concentrations vs. total flow
volume eluted at end of study.

Core Volume AMD^ Average Iron^

Number Eluted. mL ma/L

SC-V 12,453 104.2 ± 1.1

SC-VN 27,180 35.5 ± 0.0

SC-CH 23,768 199.2 +3.5

1 n = 1

2 n = 2

10.4 Phase III: MICROBIAL BIOMASS, COMMUNITY STRUCTURE AND

PHYSIOLOGICAL STATUS ASSESSMENT OF NATURAL AND CONSTRUCTED
WETLANDS

10.4.1 Introduction

Quantitative extraction of cellular components of sedimentary
microbiota permits an assessment of the viable microbial community
biomass, composition and, to a degree, physiological status,
without the problems associated with direct enumeration or cultural
methods (White 1986) . In particular, phospholipid ester-linked
fatty acids (PLFA) have proven to be reproducible, capable
indicators of a wide suite of microbiota encountered in the
environment (e.g., Guckert et al. 1985). The phosphate of the
microbial phospholipids has been shown to have a rapid turnover in
both living and killed cells (White et al. 1979), indicating that
an accurate analysis selective for this lipid class would permit a
quantitative estimation of the viable microbiota. The objective of
these studies was to use PLFA analysis to assist in the assessment
of the role of microorganisms in the amelioration of AMD in natural
and constructed wetlands by demonstrating the range of biomasses
and the differences in the structure and physiological status of
the viable microbial communities within these wetlands.

10.4.2 Technical Approach

A total of twenty wetland sediment samples were collected in
the field and subjected to PLFA analysis in conjunction with the
wetlands study.

149

10.4.3 Materials and Methods

10.4.3.1 Field sampling

Sediment samples were collected from several wetland sites:
Hardscrabble Creek, Swamp Gulch and Sand Coulee (Figure 10-1) .
Hardscrabble Creek is a natural wetland site. The Swamp Gulch
wetland site receives inputs of AMD (pH =2.5) from the
Carbonate mine. Hardscrabble Creek was chosen as a "control
site" that did not receive AMD, but was in close proximity to
Swamp Gulch. Sediment samples were collected from the 20-21 cm
depth horizon for both sites in individual replicate cores. The
Sand Coulee wetlands are constructed wetland cells which began
operation within the past two years. In both Sand Coulee
wetland cells studied, the surface sediment (approximately 1-2
cm) and 20-21 cm depth horizon were sampled in independent cores
collected along a transect across the wetlands.

The principal differences in the wetland sediments observed
included the dense vegetation mat and complex root systems of
the natural wetlands which held the wetland sediment together
into a cohesive unit, whereas the man-made wetland sediment
lacked this vegetation and cohesiveness. All wetland sediment
samples were frozen and then lyophilized (freeze-dried) . After
drying, the samples were inventoried, randomly relabeled, and
total dry weights were determined.

10.4.3.2 Statistical analyses

All statistical analyses utilized SPSSX statistics programs
available on the Montana State University Honeywell 66/DPS
mainframe computer with CP-6 operating system. Data transforma-
tions were used to meet the homogeneity of error variance
assumptions for the analysis of variance model used (Winer
1971). In general, a log^g transformation was performed, except
when the variable was a percentage, in which case an arcsine
square root transformation was made (Winer 1971) . Following the
analysis of variance (ANOVA) , Tukey's Honestly Significant
Difference test (HSD, SPSSX programs) was calculated for a
multiple comparison of means. Significant difference maps were
produced keeping the within-experiment, family-wise error rate
set at a = . 05 .

10.4.3.3 Lipid nomenclature

Fatty acids are named as the total number of carbon atoms,
followed by the number of double bonds closest to the aliphatic
(w) end with the two numbers being delimited with a colon (i.e.,
carbon atom number: double bond number) . Furthermore, the
geometry ('c' for cis and 't' for trans ) of the double bond is
also denoted with the double bond number (e.g., 18:lw7c). The
prefixes 'i', 'a', and 'br' refer to iso, anteiso and methyl-

150

branching of the unconfirmed position, respectively.
Cyclopropyl fatty acids are designated as 'cy' with the ring
position in parenthesis relative to the aliphatic end (e.g., cy
19:0).

10.4.4 Results

10.4.4.1 Total Biomass

The total PLFA quantified can be used as an estimate of the
biomass of the viable microbiota in the wetland sediments. The
total nmoles of PLFA per gram dry weight of wetland sediment is
shown for each site and depth horizon in Figure 10-12. His-
tograms which are labeled with symbols for the same group (e.g.,
Bl and Bl+2, Figure 10-12) are included in the same homogeneous
subset as defined by the Tukey HSD test. Histograms with no
common symbols (e.g., Nl and N2, Figure 10-13) are significantly
different by the Tukey test. The letters for these symbols
indicate the parameter measured (B=biomass, N=nmoles/gdw, M=mole
percent, T= trans/cis . C=cy c 1 opr opy 1 / c i s ) . The numbers differen-
tiate the groups, with 1 being the group with the highest mean
value of the parameter measured.

At the 20 cm depth horizon, therefore, the Hardscrabble
Creek and Swamp Gulch sites have a significantly greater micro-
bial biomass than the Sand Coulee sites (Figure 10-12) . The 20
cm horizons for the Lincoln sites are as dense in microbial
biomass as the surface sediments of the Sand Coulee sites.

10.4.4.2 Community structure

In addition to estimating total microbial biomass, the PLFA
profiles can be used to compare the microbial community
structure for the sampled wetlands. Table 10-5 includes the
PLFA profiles for all sites and depth horizons expressed as the
mole percent of the entire PLFA profile. The fatty acids are
listed in their elution order off the chromatography column.
The data are expressed as average ± one sample standard
deviation. Statistically significant differences within
individual PLFA are marked with an '*• (ANOVA, a = 0.05).

While the presentation of a complete data set such as shown
in Table 10-5 is important, microbial community structure
interpretations are difficult using individual PLFA. Table 10-6
describes some PLFA groups assembled to aid interpretation. The
eukaryotes are characterized by polyunsaturated fatty acids
(Shaw 1966). The bacterial group is a combination of the
terminally-branched fatty acids common to gram-positive-like
organisms (Kaneda 1977) ; the PLFA indicative of sulfate-reducing
bacteria (Bowling et al. 1986); the cyclopropyl fatty acids and
their immediate precursor, the major product of the anaerobic
desaturase fatty acid biosynthetic pathway, 18:lw7c (Fulco

151

100

^ 80

u

60

I 40

d

I— I

■g 20

H

Surface (1-2 cm) Sample

Bl+2

nd nd

Bl+2+3

100

Lincoln Lincoln Sand Sand

Hard- Swamp Coulee Coulee
scrabble Gulch Large Small

Creek Bog Bog

nd = not detennined.

Histograms represent means + 1 std. deviation.

B1+2+3 symbols used to distinguish subsets generated by Tukey's test

of iog^ Q transfonned data (see text for details).

Figure 10-12. Total microbial biomass as measured by analysis of
phospholipid ester-linked fatty acids.

152

V)

'o

a

o

o

s

1—1

a

■*->
o

Surface (1-2 cm) Sample

--EZIl nmoles/gdw
ESS mole %

4-

3--

Kl+2

Ul+2

20

16

12

O

v

1—1

o

Lincoln Lincoln Sand Sand

Hard— Swamp Coulee Coulee

scrabble Gulch Large Small
Creek Bog Bog

nd = not determined.

Histograms represent means + 1 std. deviation.

N1+2 symbols used to distinguish subsets generated by Tukey's

test of log.|Q transformed data; Ml +2 used 2*arcsin square root

transformea data (see text for details).

Figure 10-13. Sulfa te-reducing bacterial phospholipid fatty acid
biomarker expressed as density per gram of soil and
proportion of total microbial biomass.

15 3

II ,-(CM00V£)rHO<M<^CNO«J'vr>,-l^ «) fO rH CM iD CN rH CM O fN r-l rH TT CM 'J* f- rH OJ II

m

0)

•r-l

m

en
C

6

to
in

a
o

-H

u

fO

>

a;

■u

u
o

Q)

O

e

•rH
O

>l

(0

fa

in
I

1-1

X!
(0

'OOOOOOOOf-lrHOOOOCvlOO

1-5 g |{ OOOOOOOOOOOOrHOOOww^v^ — ■••• - - ij

. I+I +1+1 +I+I+I+I+I+I+I +1+1 +1+1 +1+1+1 +1+1 +1+1 +1+1+1 +l+l+l+l+l+l|j

CN II ir)-!rr-<nrHin^inoo»x)ro-<j'u-)a)oo'«'r-.Hr-nO'3'r-o<NOOf-trMts).Hro^_.H II

1^ U II
Ci] II
U o II

Or^•w•<TOOr^r-llnor^o.^r^<Nr-l^^ot-ooMO^~vor^o^ocMa^oo

II
II

II OM

vf)t^^,Hi^i-Hrgoocvi"3'v£>"ircMrH'r«3rofMPOooTrr-chrH^O"a'<Nivr>oll

B S il O^OOOOOO<NOOOr-ioOOOOf0OO(NtHrHt-)OO.HOC^rHrH jj

",2 il +i+i+i+i+i+i+i+i+i+i+l+l+l+l+l+l+i+i+l+i+l+i+i+l+i+l+l+l+l+l+l+l||

Z 3 II ,^ooo^o^<5^oo^lnovov(~'I•o>r^<x>'!r«•u^ror-coln^£>«>«a'C-or-o^c^) II

to II
v> II

OVDCM<NOiHrHOr~OOOrOi-HrH

in CM CTl <£> (y> P~ 00 «r ■«' CM

rHrHOinOr-HVOOr-COOOCMrHOCslCM II

^ ^ 11

II
II
II
II

vor-pit-uiro^oo'TCMinc^ironomrHVocsjt-mi' ji

^ II OrWrHCv)OOOOr-JOOr-l,HOOOOOOOOOOr-ICNJOrHOC5rOOO II

El § 8 +|+l+|+|+|+|+|+|+|+|+|+|+|+|+|+|+|+|+|+l+l+l+l+l+l+l+l+l+l+l+l+l}|

(D o it

(Mil oovonp^r-o^r-^or-cM^oc~'3•"9•r-oo^o^noo^-r-oocMOO^-Olnvoo^vc«3• ii

m

U

<

u

D Q) II
O O II
u m 11

<U II

Q M II

2 3 II

< W II

CO II

II

II

II

II

II

II

11

II

_ll

ft .Cll

e oil

(3 t-iii

s 311

zw oil

•-^ !
'^ B

H 0)

xi

II OOOinrOOOrHiH"!rOrH>-ltOCMfOi-ICMO^OOrHOVO'3'OrHnr-tinOO jj

jj -^ ^ "^ Ii

II
II

u >'

n II

II rHtrovDfooorH-o'oocMini-iofOtjMnincMin'srr-OM-vovDforMOCMrooooo ii

C>r<ioOC>oiod'q'OOOCMOrHOrHOOOOrH O O CMOOOOVDOn II

+1+1 +1+1 +1+1+1+1+1+) +1+1 +1+1 +1+1 +1+1 +1+1 +1+1 +1+1 +1+1 +1+1 +1+1 +1+11

CMr~^tocvivocor~inc>jpOr^'<rroe»cMC^csoovC'<rrHro«rrococMooin'«J'rHOI
c>iocMCMOrorHoaiooooT-iniHr>)OT-iOrH»a<oinc>ioorHOr--ro'3'

CM ""I

^CMOV£>CMO<r)rOOOOPOCM^tS)tnrHOCMOrHr-rHVf>rHOrHtMCvl'9<U>T-lrH

OOOOOOOOCMOOOOOOOOOrHOOOOtHCVlrHOOOVDOO

+ 1+1 +1+1 +1+1 +1+1 +1+1 +1+1 +1+1 +1+1 +1+1 +1+1 +1+1 +1+1 +1+1 +1+1 +1+1 +1+1

tr)o^^-o^JC^^^oo^ooao«•lOCM^-lcv"Tct>vocMoo'^r^t-^oooeJ^o^^ocMU^cJ^^-^o

rM.-ifOVOOOrHOVO{NT-IOC>li-lCMOrHOCMOrHmOt-OOCOOCMOOO

—1 rH 1-1

* -Sll

Sll

II
II

IIOOOCMOOVOCN'3'OrHOnrHr-OCMCMCSJOOCyiOOrHVOCNi-HCMCMOi-H

U

in
-a

OOOOOOOOOOOOOOOOOOOOrHOOOOi-IOOOrHOO

+ 1+1 +1+1 +1+1 +1+1 +1+1 +1+1 +1+1 +1+1 +1+1 +1+1 +1+1 +1+1 +1+1 +1+1 +1+1+1 +

vD r- o er> CM vD 00 t vo rH'q'OvovoinoorHCMr-r-r-iiDinoooovct-iiHCvjro'rcM

OOCMnOOOiHOOrHCMOOrHVOOCNIrHrHOi-HlOOr-mOOtNrHOOOO

rH CM

m
o

o
II
to

•K-k* -k *•)<**■•!** -K -K-K-K-X **1t*

O

iH

4J O O t~

O O 4J O 00 C-"rHO 004J0 _

o<t~-r~mtH SVC -^vDo >x>vooocnr*r~m t-iooco

O OO O S S S S S rH rH O O » '• SSSSSSS ••••SS

..O" ••rHO "rHrHrHrHrHO •• 0) •• .•rHP-OnCMCOrHrHrHr-IOCnO<"3<lO

•q. .. lO in •• •■ vo •• ■• i~ g r- r^ •• rH V* "Z* J^ J^'

rH"3<rHrHir)inrHVOVO«>VDVOVDrHOrHrHt-- >ir- OOOOOOOOOOOOOOCO iH >iO O
-HrH-H <DrHrH-HrHrHrHrHrHrH-HrH-H (OrH OrHrHrHrHrHrHrHrHrHXl OCMCvJ

154

1983) . The anaerobic desaturase group includes the w7c products
and the corresponding cyclopropyl and mid-chain branched isomers
resulting from transmethylation, as well as the trans -isomers
which are hypothesized to be formed by an isomerase of the w7c
products of anaerobic desaturase (Guckert et al. 1986). The
sul fate-reducing bacteria can also be further defined based on
the genera-specific PLFA biomarkers (Dowling et al. 1986; Edlund
et al. 1985) .

When the PLFA groups defined in Table 10-6 are used to test
for changes in microbial community structure, the significant
difference map shown in Table 10-7 is produced. The map is set
up such that treatment means for each PLFA group increase from
left to right, and those connected by a common line segment are
not significantly different by the Tukey test (a = 0.05).

The results presented in Table 10-7 will be discussed in
more detail later. Briefly, the Sand Coulee wetland surface
depth horizons (site codes = 3 and 5) were characterized as
having the highest eukaryotic and lowest bacterial proportions.
There were no significant differences in the proportions of
gram-positive- like organisms or the total anaerobic desaturase

Table 10-6. Description of phospholipid ester-linked fatty acids
(PLFA) used to define microbial groups for data
interpretation.

I
I

MICROBIAL GROUP

PLFA USED TO DEFINE GROUP

Eukaryotes

20:4w6 + 20:5w3 + 18 : 3w3 + 18 : 2w6 + 18 : 3w6 +
16:lwl3t

Bacteria

Gr am-pos it ive

Anaerobic
Desaturase

18 : lw7c

Sulfate Reducing
Bacteria

Desulfobacter spp.

Desul f ovibr io spp.

il5:0 + al5:0 + il7:lw7c + I0mel6:0 + il7:0
+ al7:0/17:lw9c + cyl7:0 + 18:lw7c + cyl9:0

il5:0 + al5:0 + il7:0 + al7 :0/17 : lw9c

16:lw7c + 16:lw7t + 10mel6:0 + cyl7:0 +
18:lw7c + 18:lw7t + cyl9:0

18 : lw7c

il7:lw7c + 10mel6:0 + cyl7:0
10mel6:0 + cyl7:0
il7:lw7c

155

products. However, the principal unmodified product of anaero-
bic desaturase, 18:lw7c, was in the greatest proportion in the
Hardscrabble Creek site (site code = 1) . All sulfate-re- ducmg
bacterial groups were in the greatest proportion of the total
microbial biomass in the Sand Coulee 20 cm depth horizons (site
codes 4, 6) with the lowest proportions being in the Swamp Gulch
20 cm (site code 2) and Sand Coulee surface (site codes 3, 5).

Since sul fate-reducing bacterial biomass and activity had
been assumed to have a possible influence on AMD amelioration, a
further illustration of sul fate-reducing bacterial distribution
between the sites was done (Figure 10-13). The mole percent
histograms graphically display the relationship shown in the
significant difference map (Table 10-7) . The total sul fate-
reducing bacterial biomass estimates in nmole per gram dry
weight wetland sediment are also shown indicating that although
the Swamp Gulch site has the lowest proportion of biomass as
sul fate-reducing bacteria (group M2, Figure 10-13), the density
of sul fate-reducing bacteria (per gram dry weight) for this site
along with Hardscarbble Creek is included in the highest group
(group Nl, Figure 10-13). In general, the Sand Coulee sites are
lower in sul fate-reducing bacterial biomass density with the
higher values occurring in the surface sediments. Greater total
biomass and species diversity was evident at the Hardscrabble
Creek and Swamp Gulch wetland sites.

Even though sul fate-reducing bacteria are present in the
Sand Coulee constructed wetland they are ineffective in removing
sulfate. Where these organisms are present (at the surface)
environmental conditions (e.g. , ©2 levels, nutrient
availability) would preclude their activity. Therefore,
specifications for wetland construction should address carbon
and other nutrient additions as well as reduced oxygen levels
necessary to enhance the densities and activities of these
organisms .

10.4.4.3 Physiological stress

In addition to microbial biomass and community structure
estimates the PLFA can be used to describe aspects of the com-
munity physiological status. Specifically, the trans/cis ratio
of monoenoic PLFA and the cyclopropyl/cis PLFA ratios have been
suggested as possible "starvation or stress" lipid indices
(Guckert et al. 1986). These ratios for each wetland site and
depth horizon are shown in Figure 10-14 . The Sand Coulee sites
can all be characterized by having higher cyclopropyl/sis ratios
(group CI, Figure 10-14), whereas the Swamp Gulch site is char-
acterized by an extremely high trans/cis ratio (group Tl, Figure
10-14) . The control Hardscrabble Creek site is low for both of
these "stress" lipid ratios (groups T2 and C2, Figure 10-14).

156

Table 10-7.

Significant difference map generated from
Tukey's HSD test^.

Site
Codes:

1
2
3
4
5
6

Lincoln, Hardscrabble Creek (20 cm)
Lincoln, Swamp Gulch (20 cm)
Sand Coulee Large Bog (surface)
Sand Coulee Large Bog (20 cm)
Sand Coulee Small Bog (surface)
Sand Coulee Small Bog (20 cm)

MICROBIAL GROUP '

Eukaryotes

Bacteria

Gram Positive

fbranched chains)
(Anaerobic

desaturase)
18:lw7c

Sulfate Reducing
Bacteria
Desulfohacter spp.

Desulfovihrio spp.

Low

6 4

High
3 5

1

3

5

4

2

1

6

3

5

4

6

2

1

2

3

5

1

4

6

2

3

5

1

4

6

2

3

5

1

6

4

Within experiment familywise error rate set at a=0.05
for 2*arcsine square root transformed mole % data of
Table 10.5 grouped according to designations in Table
10.6.

2

Treatment means for each group of fatty acids increase
from left to right, and those connected by a common
line segment are not significantly different for this
test.

157

0.80

o 0.60 J-

O

w
C
cd
u

E-

0.40

0.20

0.00

0.8

Surface (1-2 cm) Sample

nU trans/cis
ESS cy/cis

Cl+2

Cl+2

f^^

N

T2

;^kti

nd

nd

O 0.6

• rH

CO

^ 0.4 T

W

^ 0.2 +

0.0

20-21 cm Sample

Tl

T2

C2

Cl+2

TT

I

^ T2+3

I

1

T3

fl

Cl

I

T2

I

Lincoln Lincoln Sand Sand
Hard— Swamp Coulee Coulee
scrabble Gulch Large Small
Creek Bog Bog

3.0

-2.5 o
a

-2.0 w

o

1.5 -
o

-1.0 cL

o

I— t
o

0.5 o

0.0
3.0

"2.5 o
•^"^ .,-1

cd
--2.0 w

-- 1.5 -r"

Cl,

o

--1.0 A

o
o

0.5 o

0.0

nd = not determined.

Histograms represent means + 1 std. deviation.

T1+2+3 and C1+2 symbols used to distinguish subsets generated by

Tu key's test of log.| q transfonned data (see text for details).

Figure 10-14. Microbial community trans/cis and cyclopropyl/cis
phospholipid ester-linked fatty acids ratios.

158

10.4.5 Discussion

The results of the PLFA data presented here need to be
discussed in the context of results of the other projects
conducted within the wetlands study. The Swarop Gulch wetland
site has had in situ AMD inputs. The natural wetland sediments
(Phase III) , as well as sediment cores brought into the labora-
tory and continually flushed with AMD (Phase II) , are able to
effectively buffer the AMD water. The cores analyzed in the
laboratory have been shown to have a concomitant removal of
sulfate from the AMD during the buffering process. When
sediment cores from the Sand Coulee large wetland cell were
exposed to AMD, no buffering occurred except when a core was
treated with chloroform. No apparent sulfate utilization was
measured from the Sand Coulee wetland cores.

All of these wetland sediments are bacterial ly dominated
communities (Table 10-5, note proportions of bacteria PLFA as
described in Table 10-6) as opposed to a prevalence of
eukaryotic (i.e., higher life forms such as green algae, fungi,
etc.) organisms. The Sand Coulee wetland surface horizons show
a higher eukaryotic input (Table 10-7) . The natural wetlands,
which contain more vegetation and root material, support a much
higher microbial biomass density than the constructed wetlands
of the Sand Coulee cells (Figure 10-12).

The apparent correlation of sulfate loss with AMD buffering
suggests an important role for sul fate-reducing bacteria in the
AMD amelioration process. Sul fate-reducing bacteria data
displayed in Figure 10-13 indicate that, as was seen for total
microbial biomass, the natural wetlands (Hardscrabble Creek,
Swamp Gulch) support a higher sul fate-reducing bacterial biomass
density than the artificial wetlands. Interestingly, the
sul fate-reducing bacterial proportion of the total microbial
biomass shows the opposite trend with the non-buffering Sand
Coulee site having a higher proportion of the total microbial
community as sul fate-reducing bacteria. The data in Figure 10-
13 suggest several important aspects of AMD buffering by wetland
systems: 1) If total sul fate-reducing bacterial biomass density
is an important factor in the buffering process, we might
predict that the Hardscrabble Creek control site would also be
able to adequately buffer AMD; 2) The higher proportion of
sul fate-reducing bacteria in the less diverse Sand Coulee
microbial coxununities with the lower sul fate-reducing bacterial
densities suggests the importance of other microbial populations
to the sul fate-reducing bacteria. Sul fate-reducing bacteria are
obligate anaerobes which are considered to be terminal carbon
users when sufficient sulfate levels are present. Organisms at
this nutritional level rely upon other aerobic, facultatively
anaerobic and obligately anaerobic primary degraders to generate
the carbon and energy substrates required for sul fate-reducing
bacterial growth from the more complex molecular inputs into the

159

wetland (Dowling 1987) . Therefore, in the constructed wetland
the availability of additional simple carbons sources (e.g.,
viable plants) would be beneficial to the sulfate-reducmg
bacterial populations.

This discussion suggests that if the primary degraders
could be stimulated by an input of carbon and energy sources the
sul fate-reducing bacterial populations might be better "fed,"
thereby increasing their density. If the density of sulfate—
reducing bacterial populations is actually associated with AMD
buffering, this artificial wetland system might then provide
acidity and sulfate amelioration. One obvious difference in the
wetlands that might account for the differences in microbial
population proportions is the lack of vegetation in the Sand
Coulee wetlands. Root systems provide nutritional inputs.
Also, bioturbation of burrowing micro-eukaryotes associated with
this vegetation can have an effect on the wetland sediment. _^
Note that after the Sand Coulee surface horizons the Swamp Gulch
and Hardscrabble Creek sites also have higher proportions of
eukaryotic biomass than the corresponding 20 cm depth horizons
at the sand Coulee cells (Table 10-7) .

It should not be assumed that wetland size alone governs
performance. These findings indicate that the amount of
microbial biomass is deficient in constructed wetlands when
compared to natural wetlands. Enhancement of the microbial
components would be expected to markedly affect the amelioration
of sulfate and acidity due to AMD.

The PLFA group made up of the products of the anaerobic
desaturase fatty acid biosynthetic pathway shows no significant
differences in their proportion to the overall microbial
community for any of the sites (Table 10-7) . However, propor-
tions of the immediate biosynthetic products (18:lw7c and
16:lw7c) and their modification products (cyclopropyl and trans )
differ between sites. Table 10-7 indicates that 18:lw7c is
significantly higher in proportion for the Hardscrabble Creek
profiles, followed by Swamp Gulch and the 20 cm depth horizons
of Sand Coulee wetlands.

Figure 10-14 indicates that the cyclopropyl/ci^ ratios and
trans/cis ratios also differ between sites, suggesting a
difference in the microbial community physiological status. A
full discussion of the lipid biochemistry and microbial physiol-
ogy behind these proposed microbial "stress" indices is given in
Guckert et al. (1986). The interpretation of the higher
cyclopropyl/cis ratios at the Sand Coulee site must take into
consideration the input of cyclopropyl PLFA from
sul fate-reducing bacteria, where physiological response of
cyclopropyl PLFA is unknown. It is evident that the Sand Coulee
microbial communities are stressed in a way that might be
analagous to anaerobic

160

microcosms which have been shown to have dramatic loss of cis -
monoenoic PLFA with the corresponding production of cyclopropyl
PLFA (Guckert et al. 1985).

The other dramatic physiological shift shown in Figure 10-
14 is the extremely high trans/cis ratio in the Swamp Gulch
microbial community. This ratio generally ranges from 0.01 to
0.08 for most bacterial cultures and 0.01 to 0.09 for marine,
mangrove and estuarine sediment surfaces (Guckert et al. 1986).
When Vibrio cholerae was starved for 7 days, the trans/cis ratio
was 0.85 (Guckert et al. 1986). Significantly higher trans/cis
ratios (0.15) were found in a highly consolidated sediment
surrounding the burrow of a mud shrimp in estuarine sediment
(Dobbs and Guckert 1988) . Four hours after an estuarine
sedimentary microbial community was disturbed by sieving, the
trans/cis ratio increased to 0.95 (Findlay et al. 1988). It is,
therefore, unusual to find an in situ trans/cis ratio above 0.50
as was observed in the Swamp Gulch sediments. This finding
might be a response to low pH, metal toxicity, or another
unknown stress. The differences between the trans/cis and
cyclopropyl/cis stress of the Swamp Gulch and Sand Coulee
wetlands are unknown. The control site, Hardscrabble Creek, is
low for both "stress" lipid indices. It was previously
suggested that this community might buffer AMD as the Swamp
Gulch wetland has. It is unknown if the trans/cis ratio would
follow the same pattern following AMD input.

The microbial populations at both Swcimp Gulch and Sand
Coulee wetlands are physiologically stressed. These physiolog-
ical indices represent a means by which wetland performance can
be measured. For example, if amendments applied to constructed
wetlands resulted in improved AMD eumelioration and the stress
indices dropped in magnitude, this would indicate that the
overall "health" of the microbial community improved. This
would translate to enhanced acidity removal, sulfate control and
metal immobilization.

161

11 . LITERATURE CITED

Alden, W.C. 1932. Physiography and Glacial Geology of Western
Montana and Adjacent Areas. U.S. Geological Survey
Professional Paper 231. 2 00 p.

Alexander, H.C., J. A. Quick, Jr. and E.A. Bartlett. 1981.

Static acute toxicity of sodium bromide to fathead minnows.
Bull. Environ. Contam. Toxicol. 27:326-331.

Aley, T. and M.W. Fletcher. 1976. Water tracer's cookbook:

Journal of the Missouri Speleological Survey. V. 16(3): 32
p.

Allems, M.I.H. 1972. Metabollic role of sulfates and sulfides
producing bacteria in pollution of waters. Kentucky Water
Resources Inst., Lexington Res. Rpt. No. 46, 52 pp.

Alt, D.D. and D.W. Hyndman. 1972. Roadside Geology of the

Northern Rockies. Mountain Press Pub. Co. Missoula, MT.
280 p.

Alter, J.C. 1937. Shielded storage precipitation. Monthly
Weather Review. 65(7): 262-265.

American Public Health Association. 1985. Standard methods for
the examination of water and wastewater. 16th ed.
Am. Public Health Assoc, Inc. New York, NY.

American Society of Agronomy. 1965. Monograph 9, Methods of
Soil Analysis, Part 2. Black, C.A. Ed. pp 891-900.

Arts Manufacturing and Supply. 1987. 105 Harrison and Oregon
Trail, American Falls, Idaho.

Baden, W. and R. Eggelsman. 1964. Der Wasserkreislauf eines
Nordwestdeutschen Hochmoores. Schrifteur. Kurat. Kult.
Hamburg. 12:1-156. From: Chasen and Siegel (1986).

Baden, W. and R. Eggelsman. 1961. Moorhydrologische

Untersuchungen am Westrand des Steinhuder Meeres zur
Feststellung eines unterirdischen Seebflusses. Wasser
Boden 13:403-410. From: Chason and Siegel (1986).

Baden, W. and R. Eggelsman. 1963. Zur Durchlassigkeit der
Moorboden (On the permeability of marshland). Z.
Kulturtech. Flurbereinig. 4:226-254. From Chasen and
Siegel (1986) .

Barnes, H.L. and S.B. Romberger. 1968. Chemical aspects of
acid mine drainage. J. Water Pollution, Contr. Fed.
40(3) :371-384.

162

Barnes, G.D., B.S. Dhillon, and A.S. Hume. 1981. The

Identification and Treatment of Bromides and Trihalome-
thanes in Potable Water from the Pearl River at Jackson,
MS: Evaluation of Toxicity on Species Diversity. Water
Resources Research Institute, Mississippi State, MS. 70 p.

Bayly, I.L. and T.A. O'Neill. 1972. Seasonal ionic

fluctuations in a Typha alauca community. Ecology 53:714-
719.

Bear, J. 1972. Dynamics of Fluids in Porous Media. American
Elsevier, New York, NY. 764 p.

Bedish, J.W. 1976. Cattail moisture requirements and their

significance to marsh management. Am. Midi. Nat. 78:288-
300.

Beers, W.F.J, van. 1983. The Auger Hole Method. International
Institute for Land Reclamation and Improvement/ILRI,
Wageningen, The Neatherlands . 23 p.

Bell, P.R. 1959. The ability of Sphagnum to absorb cations
preferrentially from dilute solutions resembling natural
waters. J. Ecol. 47:351-355.

Berg, W.A. and G.W. Thomas. 1959. Anion elution patterns from
soils and soil clays. Soil Sci. Soc. Amer. Proc. 23:348-
350.

Biggar, J.W. and D.R. Nielson. 1962. Miscible displacement:
II. Behavior of tracers. Soil Sci. Soc. Amer. Proc.
26:125-128.

Blake, G.R. and K.H. Hartge. 1986a. Bulk Density. IN: A.
Klute (ed) . Methods of Soil Analysis, Part I, Second
Edition. Agronomy 9:363-375.

Blake, G.R. and K.H. Hartge. 1986b. Particle Density. IN: A.
Klute (ed) . Methods of Soil Analysis, Part I, Second
Edition. Agronomy 9:377-382.

Boelter, D.H. 1969. Physical properties of peats as related to
degree of decomposition. Soil Sci. Soc. Amer. Proc.
33:606-609.

Boelter, D.H. 1965. Hydraulic conductivity of peats. Soil
Sci. 100(4) :227-231.

Bohn, H.L., B.L. McNeal and G.A. O'Connor. 1985. Soil

Chemistry, Second Edition. John Wiley and Sons. New York,
NY. 340 p.

163

Bovmier, K.H. 1987. Nutrient Removal from Effluents by an

Artificial Wetland: Influence of Rhizosphere Aeration and
Preferential Flow Studies Using Bromide and Dye Tracers.
Water Resource, 21(5) :591-599.

Boyd, C.E. 1970. Production, mineral accumulation and pigment
concentrations in Tvpha latifolia and Scirpus americanus .
Ecology 51:285-290.

Brodie, G.A. , D.A. Hammer, and D.A. Tomljanovich. 1986. Man-
Made Wetlands for Acid Mine Drainage Control. Procs. 8th
Annual National Abandoned Mine Lands Conference. Billings,
Montana.

Brooks, R.P. 1984. Optimal designs for restored wetlands. IN:
Burris, J.E. (ed.). Treatment of mine drainage by wetlands.
Contribution No. 264, Dept. Biology, Penn State University,
University Park, PA. 50p.

Brown, D.H. and R.P. Beckett. 1985. Intracellular and

extracellular uptake of cadmium by the moss Rytidiadelphus
scpiarrosa . Ann. Bot. 55:179-188.

Burris, J.E., D.W. Gerber and L.E. McHerron. 1984. Removal of
iron and manganese from water by Sphagnum moss. IN:
Treatment of Mine Drainage by Wetlands, J.E. Burris (ed.),
Pennsylvania State University, University Park, PA. 50 p.

Cannon, H.L. 1960. Botanical prospecting for ore deposits.
Science 132:581-598.

Cedergren, H.R. 1977. Seepage, Drainage and Flow Nets. Second
Ed. John Wiley and Sons, N.Y.

Chaney, R.L. and P.T. Hundermann. 1979. Use of peat moss

columns to remove cadmium from wastewaters. J. WPCF 15:17-
21.

Chason, D.B. and D.I. Siegel. 1986. Hydraulic conductivity and
related physical properties of peat. Lost River Peatland,
Northern Minnesota. Soil Sci. 142(2) :91-99.

Chattopadhyay, A. and R.E. Jervis. 1974. Multielement

determination in market-garden soils by instrumental photon
activation analysis. Analytical Chemistry. V 46(12) :1630-
1639.

Coffin, D.L. and K.R. Wilke. 1971. Water resources of the

upper Blackfoot River Valley, west central Montana. Dept.
of Nat. Resources and Conservation: Technical Report Series
of Water Resources Division, No. 1.

164

Cole-Parmer Instnment Co. 1987. Cole-Parmer. Chicago, XL.
960 p.

Cowardin, L.M. , V. Carter, F.C. Golet and E.T. LaRoc. 1979.
Classification of wetlands and deepwater habitats of the
United States. US Fish and Wildlife Service, FWS/OBS-
79/31.

Crum, H.A., W.C. Steere and L. Anderson. 1973. A new list of
mosses for North America, North of Mexico. Bryologist
76:85-130.

Dai, T.S. and J.H. Sparling. 1973. Measurement of hydraulic
conductivity of peats. Canadian Jour, of Soil Sci. 53:21-
26.

Danielson, R.E. and P.L. Sutherland. 1986. Porosity. IN: A.
Klute (ed) . Methods of Soil Analysis, Part I, Second
Edition. Agronomy 9:443-461.

Dasberg, S. and S.P. Neuman. 1977. Peat hydrology in the Hula
Basin, Israel: I. Properties of Peat. Jour, of Hydrology.
32:219-239.

Daubenmire, R. 1959. A canopy cover method of vegetation
analysis. Northwest Sci. 33:43-64.

Davis, S.N., D.J. Campbell, H.W. Bentley and T.J. Flynn. 1985.
An Introduction to Ground-Water Tracers. U.S. EPA, Ada,
OK. EPA/600/2-85/022. 201 p.

Davis, S.N., G.M. Thompson, H.W. Bentley and G. Stiles. 1980.
Ground-water tracers — a short review. Ground Water, 18:14-
23.

de Saussure, T. 1804. Richercbes chemigues sur la vegetation.
US.'' Soil Conditions and Plant Growth by E.W. Russel,
Longman publ . , London .

Dinges, R. 1982. Natural Systems for Water Pollution Control .
Van Norstrand Reinhold Company, NY.

Dobbs, F. C. and J. B. Guckert, 1988. Callianassa trllobata
(Crustacea: Thalassinidea) influences abundance of
meiofauna and biomass composition and physiological
state of microbial communities within its burrow.
Marine Ecol. Prog. Ser., In Press.

Dorn, R.D. 1984. Vascular Plants of Montana . Mountain West
Publishing, Cheyenne, WY.

165

Dowling, N. J. F. , F. Widdel, and D. C. White. 1986. Phos-
pholipid ester-linked fatty acid biomarkers of acetate-
oxidizing sulphate-reducers and other sulphide- forming
bacteria. J. Gen. Microbiol. 132:1815-1825.

Dowling, N. J. E. 1987. Role of sulphate-reducing bacteria

involved in corrosion using phospholipid fatty acid bio-
markers and electrochemical analysis. Ph.D. Dissertation,
Florida State University.

Dowling, P.C. 1969. The hydraulic conductivity of peats. M.
Eng. Sci. thesis, University College Dublin (microfilm).
From: Chason and Siegel (1986) .

Driscoll, F.G. 1986. Groundwater and Wells. Second Edition.
Johnson Division. St. Paul Minnesota. 1089 p.

Edlund, A., P. D. Nichols, R. Roffey and P. C. White. 1985.

Extractable and lipopoly saccharide fatty acid and hydroxy
acid profiles from Desulfovibrio species. J. Lipid Res.
26:982-988.

Eggelsman, R. and T. Makela. 1964. Ein fluss von Entwasserung
und land wirtschaftlicher Natzung auf die Durchlassigkeit
des Moorbodens. Maataloust Aikakausk. 36:77-84. From
Chason and Siegel (1986).

EPA. 1978. National Interim Primary Drinking Water Regulations.
IN: EPA-570/9-76-003. 159 p.

EPA. 1979. National Secondary Drinking Water Regulations. IN:
Federal Register, Vol. 44(140), July 19, 1979. pp 42195-
42202

EPA. 1982. Test methods for evaluating solid waste. Physical/
chemical methods. U.S. Environmental Protection Agency,
Office of Solid Waste and emergency Response. SW-846, 2nd
Ed. Washington, D.C.

EPA. 1983. Methods for chemical analysis of water and wastes.
U.S. Environmental Protection Agency Environmental
Monitoring and Support Laboratory, Cincinnati, OH. EPA-
600/4-79-020, March 1983 Revision.

EPA. 1987a. Assessment of the toxicity of arsenic, cadmium,

lead and zinc in soil, plants and livestock in the Helena
Valley of Montana. Reclamation Research Unit, Montana
State University, Bozeman, MT. 206 p.

166

EPA. 1987b. Assessment of the toxicity of copper, mercury,
selenium, silver and thallium in soil and plants in the
Helena Valley of Montana. Reclamation Research Unit,
Montana State Universtiy, Bozeman, MT. 88 p.

Erickson, P.E., R.L.P. Kleinmann and S.J. Onysko. 1985.
Control of acid mine drainage by application of
bactericidal materials. IN: Control of Acid Mine Drainage,
Proc. Tech. Trans. Seminar. US Dept. Int, Bureau of Mines,
IC 9027.

Ernst, W.H.O. 1965. Okologische soziolofische Untersurchungen

der Schwermetall-Pflanzengesellschaften Mitteleuropas unter
Einschluss der Alpen. Abh Landesmus. Naturk. Munster 27.

Fetter, C.W. Jr. 1980. Applied Hydrogeology Merrill Pub. Co.,
Columbus, OH. 488 p.

Findlay, R. H., M. B. Trexler, J. B. Guckert and D. C. White.
1988. The response of the benthic microbial community to
small-scale predator-disturbance events. I. Laboratory
Studies. Marine Ecol. Prog. Ser. , In Press.

Freeze, R.A. and J. A. Cherry. 1979. Groundwater. Prentice-
Hall, Englewood Cliffs, NJ. 604 p.

Fried, J.J. 1975. Developments in Water Science, Groundwater
Pollution. Elsevier, New York, NY. 330 p.

Fulco, A. J. 1983. Fatty acid metabolism in bacteria. Prog.
Lipid Res. 22:133-160.

Gadgil, R.L. 1968. Tolerance of heavy metals and the

reclamation of industrial wastes. J. of Appl. Ecol. 6:247-
259.

Galvin, L. F. and E. T. Hanranahan. 1968. Steady state

drainage flow in peat. Highway Res. Rec. 203:77-90.
From: Chason and Siegel (1986) .

Galvin, L.F. 1976. Physical properties of Irish peats. J.

Agric. Res. 15:207-221. From: Chason and Siegel (1986).

Gerber, D.W. , J.E. Burns and R.W. Stone. 1985. Removal of

dissolved iron and manganese ions by a Sphagnum moss system
IM: R.B. Brooks, D.E. Samuel and J.B. Hill (eds) . Wetlands
and Water Management on Mined Lands. Penn State Univ.

Girts, M.A. and R.L.P. Kleinmann. 1986. Constructed wetlands
for treatment of acid mine drainage: a preliminary review.
Nat. Sym. on Mining, Hydrology, Sedimentation and
Reclamation. Lexington, KY.

167

Girts, M.A. 1986. The hydrology of Sphagnum - Dominated

Appalachian Wetlands in West Virginia. M.S. Thesis College
of Arts and Sciences of West Virginia University. 203 p.

Golet, F.C. and J.S. Larson. 1974. Classification of

freshwater wetlands in the glaciated Northeast. US Fish
and Wildlife Service. Resource Publ. 116.

Grim, R.E. 1968. Clay Mineralogy, 2nd ed. McGraw-Hill, New
York, NY. 596 p.

Guckert, J. B. , C. P. Antworth, P. D. Nichols, and D. C. White.
1985. Phospholipid, ester-linked fatty acid profiles as
reproducible assays for changes in prokaryotic community
structure for estuarine sediments. FEMS Microbiol. Ecol.
31:147-158.

Guckert, J. B. , D. B. Ringelbert, and D. C. White. 1987.

Biosynthesis of trans fatty acids from acetate in the
bacterium Pseudomonas atlantica . Can. J. Microbiol.
33:748-754.

Guckert, J. B. , M. A. Hood, and D. C. White. 1986. Phospho-
lipid, ester-linked fatty acid profile changes during
nutrient deprivation of Vibrio cholerae ; increases in the
trans/cis ratio and proportions of cyclopropyl fatty acids.
Appl. Environ. Microbiol. 52(4) :794-801.

Guertin, D. , J.D. Emerick and E.A. Howard. 1985. Passive mine
drainage treatment systems: a theoretical assessment and
experimental evaluation. Dept. Environ. Sci. and
Engineering Ecology. CO School of Mines, Golden, CO.

Gustafsson, E. and C.E. Klockars. 1981. Studies on groundwater
transport in fractured crystalline rock under controlled
conditions using nonradioactive tracers. Swedish Nuclear
Fuel Supply Co. Stockholm. Div. KBS. 64 p.

Haag, D.A. and C.J. Covert. 1987. Controlling acid seepage

from buried coal wastes through groundwater inundation and
wetland. IN: (F.F. Munshower et al., eds)

Hem, J.D. and W.H. Cropper. 1959. Survey of Ferrous- ferric
chemical equilibrium and redox potentials. USGS Water
Supply Paper, 14 59 -A.

Hewitt, E.J. 1975. Plant Mineral Nutrition . J. Wiley & Sons,
NY.

168

Kiel, M. T. and F. J. Kerins, Jr. 1988. The Tracy wetlands: a
case study of two passive mine drainage systems in Montana.
Presented in symposiiim to Am. Soc. Surface Mining and
Reclamation, Pittsburg, PA. April, 1988.

Huntsman, B.E., J.G. Solch and M.D. Porter. 1978. Utilization
of Sphagnum species dominated bog for coal acid mine
drainage abatement. Geol. Soc. America, 91st Ann. Sym. ,
Toronto, Ontario, Canada.

Hutchinson, G.E. 1975. A Treatris on Limnology; Volume III
Limnological Botany. J. Wiley and Sons, New York, NY.

Hvorslev, M.T. 1951. Time lag and soil permeability in

groundwater observations. Bull. 36. U.S. Army Corps of
Engineers, Waterways Exp. Stn. Vicksburg, Miss.

Ingram, H.A.P., D.W. Rycroft and D.J. A. Williams. 1974.

Anomalous transmission of water through certain peats.
Jour, of Hydrology. 22:213-218.

Ingram, H.A.P. 1983. Hydrology. IN: Ecosystems of the world,
vol. 4A, Mires: Swamp, bog, fen and moore, general studies.
A. J. P. Gore (ed.) Elsevier, New York.

Ingram, H.A.P. 1967. Problems of hydrology and plant

distribution in mires. Jour, of Ecology. 55:711-724.

Irwin, R. W. 1970. Soil water characteristics of some Ontario
peats. Proc. 3rd Int. Peat Congr. Quebec, Canada (1968)
pp. 219-223. From: Chason and Siegel (1986).

Jaynes, D.B., A. A. Rogowski and H.B. Pionke. 1984. Acid mine
drainage from reclaimed coal strip mines. 1. Model
description. Water Resources Res. 20:233-242.

Johnson, M.S. et al. 1977. Revegetation of metalliferous mine
spoil contaminated by lead and zinc. Environ. Pollut.
12:261-277.

Kaneda, T. 1977. Fatty acids of the genus Bacillus : an example
of branched-chain preference. Bacterid. Revs. 41(2) :391-
418.

Kim, A.G., B.S. Heisey, R.L.P. Kleinmann and M. Deul. 1982.

Acid mine drainage: control and abatement research. Bureau
of Mines IC 8905.

Kleinmann, R. L. and M. Girts. 1986. Constructed wetlands for
treatment of minewater: successes and failures. Proc.
Ann. AML 8:67-73.

169

Kleinmann, R.L.P., T.O. Tiernan, J.G. Solch and R. L. Harrxs.
1983. A low cost, low maintenance treatment system for
acid mine drainage using Sphagnum moss and limestone.
Symp. Surf. Mining, Hydrology, Sedimentation and
Reclamation, Lexington, KY.

Kleinmann, R.L.P. and D. A. Cierar. 1979. Thiobacjllus

ferrouxidans and the formation of acidity in simulated coal
mine environments. Geomicrobial J. 1:373-388.

Kleinmann, R.L.P. 1985. Treatment of Acid Mine Water By

Wetlands, pp 48-52. IN: Control of Acid Mine Drainage,
U.S. Bureau of Mines Information Circular, 9027.

Knight, A.H. , R. Boggie and H. Shepherd. 1971. The effect of

groundwater level on water movement in peat: A study usxng
titrated water. The Macaulay Institute for Soil Research,
Craigiebuckler, Aberdeen, pp. 633-641.

Kohler, M.A., T.J. Nordenson and D.R. Baker. 1959. Evaporation
maps for the United States. U.S. Weather Bureau Technical
Paper 37.

Kruseman G.P. and N.A. DeRidder. 1983. Analysis and Evaluation
of Pumping Test Data. International Institute for Land
Reclamation and Improvement/ILRI, Wageningen, The
Netherlands. 200 p.

Ladwig, K.J. , P.M. Erickson and R.L.P. Kleinmann. 1985. Alkaline
injection: An overview of recent work, pp 35-40. IN:
Control of Acid Mine Drainage, U.S. Bureau of Mines
Information Circular, 9027.

Lang, G. E. and R. K. Wieder. 1985. Modification of AMD by

Sphagnum-dominated wetlands and the effect of stream water
quality, pp. 176-184. IN: R. P. Brooks and D. E. Samuel
(eds.), Proc. Conf. Wetlands of the Chesapeake. Easton,
MD.

Lee, R., J. A. Cherry and J.F. Pickens. 1980. Groundwater
transport of a salt traces through a sandy lakebed.
Limnol. Oceanogr. 25(1):46-61.

Lee, R. , I.R. Jonasson and W.D. Goodfellow. 1983. Metal

accumulation by bryophytes in some zinc-rich blanket bogs,
Selwyn Mountains, Yukon Territory. Can, J. Bot. 62:722-728.

Lenda, A. and A. Zuber. 1970. Tracer dispersion in groundwater
experiments: IN: Isotope Hydrology (Proc. Symp. Vienna,
1970), IAEA, pp 619-641.

170

r

Lohman, S.W. 1979. Ground-Water Hydraulics. U.S. Geol. Surv.
Prof. Paper 708. 70 p.

Lovley, D. R. and E. J. P. Phillips. 1987. Competitive

mechanisms for inhibition of sulfate reduction and methane
production in the zone of ferric iron reduction in
sediments. Appl. Environ. Microbiol. 11:2636-2641.

Lovley, D. R. 1987. Organic matter mineralization with the
reduction of ferric iron: a review. Geomicrobiol . J.
5:375-399.

Martin, J. P., G.K. Helmkamp and J.O. Ervin. 1956. Effect of
bromide from a soil fumigant and from CaBr on growth and
chemical composition of citrus plants. Soil Sci. Soc.
Amer. Proc. 20(2) :209-212.

Martin, A.C., N. Hotchkiss, P.M. Uhler and W.S. Bourn. 1953.
Classification of wetlands of the United States. US Fish
and Wildlife Service Spec. Sci. Rpt - Wildl. zo.

Mathur, S.P. and R.S. Farnham. 1985. Geochemistry of Humic

Substances in Natural and Cultivated Peatlands. IN: Humic
Substances in Soil, Sediment, and Water. (G.R. Aiken, D.M.
McKnight, R.L. Wershaw and P. McCarthy, eds.) Wiley-
Interscience, NY, pp. 53-85.

Mattson, S. 1929. The laws of soil colloidal behavior: 1.
Soil Sci., 28(3) :179-220.

Mayboom, P. 1967. Groundwater Studies in the Assiniboine River
Drainage Basin, Part II Hydrologic Characteristics of
Phrearophytic Vegetation in South-Central Saskatchewan.
Geological Survey of Canada, Bulletin 139. 64 p.

Mayer, A.M. and E. Gorham. 1951. The iron and manganese

content of plants present in the natural vegetation of the
English Lakes District. Ann. Bot. (London), 15:247-263.

McClernan, H.G. 1983. Metallic Mineral Deposits of Lewis and
Clark County, Montana. Memoir 52, Montana Bureau of Mines
and Geology. 73 p.

McGrath, D. , R.F. McCormack, G.A. Fleming and D.B.R. Poole.
1982. Effects of applying copper-rich pig slurry to
grassland. I. Pot experiments. Irish Journal of
Agricultural Research, V.21(l) :37-48.

McHerron, L. E. 1985. The seasonal effectiveness of a sphagnum
wetland in removing Fe and Mn from mine drainage, p. 385.
IN: R. P. Brooks and D. E. Samuel (eds.) Wetlands and
water management on mined lands. Penn. St. Univ. (abstr.).

171

McNaughton, S.J., T.C. Folsom, T. Lee, F. Pock, C. Price, D.
Roeder, J. Schmity and C. Stockwell. 1974. Heavy metal
tolerance in Tvpha latifolia without the evaluation of
tolerant races. Ecology 55:1163-1165.

Merck Index. 1983. The Merck Index. M. Windholz, ed. Merck
and Co., Inc. Rahway, NJ.

Merrill, S.D., S.J. Smith and J.F. Power. 1985. Effect of
disturbed soil thickness on soil water use and movement
under perennial grass. Soil Sci. Soc. Amer. Jour. 49:196-
202.

Miller, J.F., R.H. Frederick and R.J. Tracey. 1973.

Precipitation Frequency Atlas of the Western United States.
Volume 1, Montana. US Dept. of Comm. , National Oceanic and
Atmospheric Administration. Silver Springs, MD.

Montana Department of Highways. 1975. General Highway Map of
Lewis and Clark County, Montana. Helena, Montana.

NAS. 1974. Water Quality Criteria, 1972. National Academy of

Sciences, National Academy of Engineering. U.S. Government
Printing Office. Washington, D.C.

National Oceanic and Atmospheric Administration (NOAA) . 1987 .
Climatological Summary of Montana. U.S. Department of
Commerce .

Naymik, T.G. and M.E. Sievers. 1985. Characterization of dye
tracer plumes: In situ field experiments. Ground Water,
23(6) :746-752.

NBS. 1982. National Bureau of Standards certificate of

analysis, standard reference material 1645, river sediment.
U.S. Department of Commerce, National Bureau of Standards.
Washington, D.C. 5 p.

Oborn, E.T. and J.D. Hem. 1962. Some effects of larger types

of aquatic vegetation on iron content of water. USGS Water
Supply Paper 1459-1.

Onken, A.B., C.W. wendt, R.S. Hargrove and O.C. Wilke. 1977.
Relative movement of bromide and nitrate in soils under
three irrigation systems. Soil Sci. Soc. Amer. J. 41:50-
52.

Orion Research. 1982. Instruction Manual, Halide Electrodes

Modle 94-35 and Modle 94-53. Cambridge, Massachusetts. 28

172

Paivanen, J. 1973. Hydraulic conductivity and water retention
in peat soils. Acta For. Fenn. 129:1-70. From: Chason
and Siegel (1986) .

Pardee, J.T. and F.C. Schrader. 1933. Metalliferous deposits
of the greater Helena Mining District. U.S. Geol. Survey,
Bulletin 842. U.S. Government Printing Office, Washington,
D.C.

Parizik, R.R. 1985. Exploitation of hydrogeologic systems for
abatement of acidic drainages and wetland production.
Proc. Wetlands and Water Management on Mined Lands. Oct.
23-24, Penn State Univ.

Parshall, R.L. 1950. Measuring water in irrigation channels
with Parshall flumes and small weirs. US Dept. of Agric.
Circ. 843. Washington, D.C. 62 pp.

Peccia & Associates. 1987. Final report: Wetlands for

treatment of acid mine drainage, Johnson Fields area. Sand
Coulee , Montana .

Pesavento, B.G. 1984. Factors to be considered when

constructing wetlands for utilization as biomass filters to
remove minerals from solution. IN: Brooks, J.E. (ed.).
Treatment of mine drainage by wetlands, contribution No.
264, Dept. Biology, Penn State Univ.

Reed, S.C., R.K. Bastian and W. Jewell. 1980. Engineering

assessment of aquaculture systems for wastewater treatment:
an overview. IN: Aquaculture Systems for Wastewater
Treatment, Seminar Proceedings and Engineering Assessment.
U.S. EPA Washington, D.C. EPA 430/9-80-006. pp. 1-11.

Reichmuth, D.R. 1986. Fluvial Systems in the Wetlands Environ-
ment, preliminary draft. Geomax, P.C. Bozeman, MT. 45 p.

Robinson, R.A. , R.H. Stokes. 1959. Electrolyte Solutions.
Academic Press Inc., New York, NY. 559 p.

Romanov, V.V. 1968. Hydrophysics of bogs. N. Kaner (tr.)
Heimann (ed.). Israel Prog. Sci. Transl., Jerusalem.
From: Chason and Siegel (1986) .

Ross, C.P., D.A. Andrews and I.J. Witkind. 1955. Geologic map of
Montana. Montana Bureau of Mines and Geology.

Ross, R.L. and H.E. Hunter. 1976. Climax vegetation of

Montana, based on soils and climate. USDA, SCS, Bozeman,
MT.

173

Sauty, J. P. 1980. An analysis of hydrodispersive transfer in
aquifers. Water Resour. Research. 16(1) : 145-158.

Schmotzer, J.K., W.A. Jester and R.R. Parizek. 1973. Ground-
water tracing with post sampling activation analysis.
Jour, of Hydrology, 20:217-236.

Shaw, S.P. and C.G. Frederick. 1956. Wetlands of the United
States. US Fish and Wildlife Service Circ. 39.

Shaw, R. 1966. The polyunsaturated fatty acids of
microorganisms. Adv. Lipid Res. 4:107-174.

Shea, T.K. 1947. Geology and Mines of the Upper Blackfoot
Valley, Montana. B.S. Thesis, Department of Geology,
Montana School of Mines, Butte, Montana, p. 24.

Shearer, R.E., W.A. Evenson, J.W. Moustellar and R.P. Simmerer.
1970. Characteristics of viable antibacterial agents used
to inhibit acid-producing bacteria in mine waters. Nat.
Coal Assn./Bitum. Coal Res., Inc. Third Symp. on Coal Mine
Drainage Res. Reprints, Pittsburge, PA, pp 188-199^

Shumate, K.S., E.E. Smith, P.R. Dugan, R.A. Brant and C.I.
Randies. 1971. Acid mine drainage formation and
abatement. Water Pollution Control Res. Sci. Post-42 14010
FPR 04/71.

Simola, L.K. 1977. The tolerance of Sphagnum fimbriatum
towards lead and cadmium. Ann. Bot. Fennici 14:1-5.

Singer, P.C. and W. Strum. 1970. Acid mine drainage: the rate
determining step. Science 167:1121-1123.

Skibitzke, H.E. and G.M. Robinson. 1963. Dispersion in

groundwater flowing through heterogeneous materials. U.S.
Geol. Surv. Prof. Paper. 386-B.

Slichter, C.S. 1902. The motions of underground waters: U.S.
Geological Survey Water Supply Paper No. 67. 106 p.

Smith, S.J. and R.J. Davis. 1974. Relative movement of bromide
and nitrate through soils. Journal of Environ. Quality,
3(2) :152-155.

Smith, S.G. 1967. Experimental and natural hybrids in North
America Typha (T yphaceae ) . Am. Midi. Nat. 78:257-287.

Soil Survey Staff. 1975. Soil Taxonomy: A basic system of soil
classification for making and interpreting soil surveys.
Agriculture Handbook No. 436. U.S. Dept. of Agriculture.
U.S. Government Printing Office, Washington, D.C.

174

Stanley, G. 1987. A Scientist's Dream. Ducks Unlimited
51(6) :72-79.

Soil Conservation Service (SCS) . 1981. Average Annual Precipi-
tation, Montana. U.S. Department of Agriculture, Bozeman,
Montana .

Stewart, R.E. and H.A. Kantrud. 1971. Classification of natural
ponds and lakes in the glaciated prairie region. U.S. Fish.
Wildl. Serv., Resour. Publ. 92. 57 pp.

Sturger, D.L. 1968. Hydrologic properties of peat from a
Wyoming mountain bog. Soil Sci. 106(4) :262-264.

Subramonian, D.C.S. and T.J. Sorg. 1986. Removing dissolved

inorganic contaminants from water. Environ. Sci. Technol.
20(11) :1072-1080.

Taylor, G.J. and A. A. Crowder. 1984. Copper and nickel

tolerance in Tvpha latifolia clones from contaminated and
uncontaminated environments. Can. J. Bot. 62:1304-1308.

Taylor, G.J. and A. A. Crowder. 1983b. Uptake and accumulation
of heavy metals by Tvpha latifolia in wetlands of the
Sudbury, Ontario region. Can. J. Bot. 61:63-73.

Taylor, G.J. and A. A. Crowder. 1983a. Uptake and accumulation
of copper, nickel and iron by Tvpha latifolia grown in
solution culture. Can. J. Bot. 61:1825-1830.

Tennyson, L.L. and CD. Settergren. 1980. Percolate water and
bromide movement in the root zone of effluent irrigation
sites. Water Resources Bull. 16(3) :433-437.

Thomas, G.W. and A.R. Swaboda. 1970. Anion exclusion effects
on chloride movement in soils. Soil Sci. 110(3) :163-166.

Thurman, E.M. 1985. Organic Geochemistry of Natural Waters.
Martinus Nijhoff/Dr. W. Junk Pub., Dordrecht. 497 p.

USPHS. 1962. Public Health Service drinking water standards.
U.S. Department of Health, Education and Welfare. U.S.
Public Health Service Publication 956.

Vinogradov, A. P. 1959. The Geochemistry of Rare and Dispersed
Chemical Elements in Soils. Consultants Bureau, Inc. New
York, NY. 209 p.

175

Wallace, A., E.M. Romney, G.V. Alexander and J.E. Kinnear.

1977. Phytotoxicity and some interactions of essential
trace metals iron, manganese, molybdenum, zinc, copper and
boron. Communications in Soil Sci. and Plant Analysis. V.
8(9) , pp. 741-750

Walsh, F. and R. Mitchell. 1972. A pH dependent succession of
iron bacteria. Environmental Sci. and Technology 6:809-
812.

Wehr, J.D. and B.A. Whitton. 1983. Accumulation of heavy metals
by aquatic mosses. 3: seasonal changes. Hydrobiologia .
100:285-291.

Weisel, G.F. and R.L. Newell. 1970. Quality and seasonal
flucuation of headwater streams in western Montana.
Missoula, Montana: Montana Forest and Conservation Experi-
ment Station, School of Forestry.

White, D. C, W. M. Davis, J. S. Nickels, J. D. King, and R. J.
Bobbie. 1979. Determination of the sedimentary microbial
biomass by extractable lipid phosphate. Oecologia (Berl.)
40:51-62.

White, D. C, 1986. Quantitative physicochemical

characterization of bacterial habitats, pp. 177-203. IN:
Poindexter, J.S., and E. R. Leadbetter (eds) . Bacteria in
nature, vol. 2, Plenum Press, N.Y.

Whitehead, N.E. and R.R. Brooks. 1969. Aquatic bryophytes as
indicators of uranium mineralization. Bryologist 72:501-
507.

Widdel, F. and N. Pfennig. 1984. Dissimilatory sulfate- or
sulfur-reducing bacteria, pp. 663-679. IN: N. R. Krieg
(ed) , Bergey's manual of systematic bacteriology. Vol. 1.
Williams and Wilkins, Baltimore.

Wieder, R.K. and G.E. Lang. 1986. Fe, Al, Mn and S chemistry

of Sphagnum peat in four peatlands with different metal and
sulfur input. Water, air and Soil Pollution 29:309-320.

Wieder, R.K., Lang, G.E. and Whitehouse, A.E. 1985. Metal

removal in Sphagnum-dominated wetlands: experience with a
man-made wetland system. IN: R.P. Brooks (ed.). Wetlands
and Water Management on Mine Lands , Penn State Univ . ,
University Park, PA.

176

Wieder, R.K., G.E. Lang and A.E. Whitehouse. 1984. The Use of
Freshwater Wetlands to Treat Acid Mine Drainage. IN ;
Burris, J.E. (ed.), Treatment of Mine Drainage by Wetlands,
Contribution No. 264 of the Dept. of Biology, The
Pennsylvania State University.

Wieder, R.K. and G.E. Lang. 1984. Influence of Wetlands and
Coal Mining on Stream Water Chemistry. Water, Air, and
Soil Pollution 23:381-396.

Wieder, R. K. and G. E. Lang. 1986. Fe, Al, Mn and S chemistry
of sphagnum peat in four peatlands with different metal and
sulfur input. Water, Air and Soil Pollut. 29:309-320.

Williams, W.J. 1979. Handbook of Anion Determination.
Butterworth Inc. Boston Massachusetts. 630 p.

Winer, B. J. 1971. Statistical principles in experimental
design, second edition. McGraw-Hill, N.Y.

Winter, T.C. and M.R. Carr. 1980. Hydrologic setting of

wetlands in the Cottonwood Lake area, Stutsman County,
North Dakota. U.S. Geol. Surv. , Water-Resources
Investigations 80-99, Denver, Co. 42 p.

Yamamoto, S. 1970. Study on permeability of soils, especially
of peat soil (in Japanese with English summary). Mem. Fac.
Agric. , Hokkaido Univ. 7:307-411. From: Chason and
Siegel (1986).

Zaburnov, S.A. 1987. Helping nature help itself. Coal Mining.
May, 1987. pp 42-43.

177

APPENDIX A

Analytical Methods and Quality Assurance and
Quality Control Statistics

178

m

(0

e -p

■p

U -rt

•r-i

\ c

c

M D

□

O rH t r-l

XIX-D \

e cp-p tj>

D e CO B

c
o

-P
O
(1)
4J
CI)
D

O

-p

s

03

■r-l
Ui
>i

rH
(0
C

<:

M

•p

(0

3:

I

C

O

(0
(U

-H
4J
U
Q)

CM
o

M

a<

(0

o

0}

>i

<S!

Si

S) CS <S! Si

CM m in >£)

< <C <: «C

Oi CU Oj CXi

H H w ta

u

o

in

00

<SJ O
U

O

u-i

(0

u

to
(0

o

<3)

Cht-\
CO (0

-p
■> o

0) £->
O

C *

(0 to

•P to

O O

D C

C M
O (0

U tc

CO

Q
E-<

0)

rH

(0
W
0)
-P

Co

(0

O

8-1

0)
3
•O

•H
CO
Q)

W WW\W WW WWW

eggeBEeEeeEEegeeeee

m

rH

in

in

in

s>

S

CM

iH

rH

m

rH

fS

n

ts

s

rH

rH tS

• •

s>

•

•

•

is

•

Si

•

•

s

•

•

s

•

SI

•

Si

•

rH

•

s

•

tStSSiSrHSSltSlSlSiHSSliHSlSiHSSl

r--r-i^r^r~r~-r^r-r~r~~(~^

r^ (~~

53 SSSlSSlSlSlSlSlSlSl

Si S Si Si

S (SSlSlSSlSSiSlSSCS

s s in s

CN fMCMCNCMOacMCNCMfMrMCN

<N oa rH (N

\ WWWWW\

W \

rHCMCMrHrHrHrHrHiHCMrHiHrHrHCN

<N rH • rH

(N"<!ri£)rOin<30(TiSVO<Ti(NrOcT\
«SSlSlrHrHrHrHCMrOrO">r'^"5j'
CNtNCNtNtNCNCMfMCNCNtNCNfM

OuChCLiChChChCUChCuCUCLiOjOu
WWWPlIWWWWWWMWPl!

00 s
in r^

CN CM

<d <

Oi CU

tN en

CM CN

< <:
Ou cu

C3>

in CO

CM
CO

U <C
CO a,

Djo;

e

D

c o

e -H

.H C

C O

E

D

■H 4J

E E

HDD

C -H -H E

<U E O O (0

to "O .H M J2

u (0 m jc o

u

Ou C t3

a o (0
o u

<<<:uuuuuiH

E (D

D to

•H to

to C rH

0) 1X5 1)

en C O

(U (0 CO -H

J s: s 2

E
P

Ie

•H D
to -H M E -P
CO C 0) D C

(0 o; > -H o u

4J rH iH "O W C
O 0) -H O 4J -H

Oi to CO CO CO tS3

179

in

•1-1
c

eeeeseeee

c
o

•w 4J
+J ■

u

4J
Q)

Q

O
-P

Q)
3
C

-H
4J

c
o
o

I

C
0)

a
a
ft

0)

jj
o
g

(0
M
(0

(T5

s
I

c
o

z

01

u

•r^

C

(0

D^

M

O

C

in

is> si in i-H

rH SI rH 1-1 ® «Sl

• • • • • •

i-H«Si-l(S!iS(SlS®r-t

rO CM CM

CN

ro

(SSsLDLDSLncoiriLn
iHCNcsco'^'^Lnvor^

WWWWWCdWWW

a:
o

c

(0

ro
O
O

00

O
U

X

res
o

+J

•H 0)

C dJ "O

•H t3 -iH

i-H -rH M

ro e O

AJ O -H

1-1 w x;

< CQ O

0)

o

M
4J

rH O

(0 QJ

4-) rH

O OJ

-p

0) TD
fO •'-'

■H M

c o
(0 a

O fa

(0
0)

•rH

1-1

a

<D O

4J s:

(0 -P
JJ o

■r-i

Z "

03

•• 3

c w

(POO)

D^x: -p
o a. m

O rH

JC 3
Ou CO

cs

•iH

•o

O -rt

i-i 2

■P

0)

u

•H

c

(0

o^
o

tn

iH
CU

w

(U
>

rH

o
m

05

■iH

C
O
XI

(0

u
o

-1-1

c

(0

p
o

180

Appendix A-2. Procedures for calciilating accuracy and precision

Procedures fog calculating accuracy.

1. From the summarized information on the accuracy form
the percent recovery of the blind field standard and the
recovery of each lab spiked sample pair are calculated
separately as:

%Recovery of BFS = VA x 100; or %Recovery of Spike = SSR-SR x 100

VK ' SA

where:

VA = analytical value of BFS

VK = known (or certified) value of BFS

SSR = spiked sample results

SR = sample results

SA = spike added

Perfect accuracy would be 100 percent recovery.

2. Calculate the standard deviation of all pairs.

1/2

SD =

(Recovery^ - Recoveryp,vr, . )

n-1

where recoveryj is the individual recoveries, recoverygyg ,
is the average recovery, and n is the number of values.

To validate recovery data, the individual recoveries are
compared with the average recovery value to identify individ-
ual values that lie outside the range or reasonableness.
Chauvenet's criterion is used to identify individual recovery
values that lie outside this range.

To use Chauvenet's criterion, the screening variable is
computed for recovery values that are suspected of lying
outside the range of reasonableness.

Screening Variable = (Recovery^ - Recoverygvg . ) /SD

The calculated screening variable is then compared to the
maximum allowable value (Table B-1) for the appropriate
number of recovery determinations. The suspect recovery
value is set aside (set aside values are called "outliers")
if the calculated screening variable equals or exceeds the
maximum allowable value.

If outliers are identified using Chauvenet's criterion, a new
average recovery and a new standard deviation are recalcu-
lated using the remaining "good" values, and Chauvenet's
criterion is reapplied. This procedure is repeated until all
surviving recovery values pass Chauvenet's criterion.
(Usually one application and one recalculation are enough.)
The final average recovery and final standard deviation are

181

Appendix A-2. Continued.

calculated from the "surviving" recovery values. The final
average recovery value is used to eliminate any bias from
the laboratory data.

4. The range of uncertainty (R) in the recovery is then calcu-
lated.

+ R = + tSD/(n)l/2

where;

R is the range of uncertainty expressed as a
percent

t is the value of the t distribution for the

selected confidence level (90 percent) and (n-1)
degrees of freedom

n is the number of samples

SD is the standard deviation.

The range of uncertainty, is used in conjunction with the
average recovery to determine if bias adjustments are
required.

Together, the final average recovery value for BFS and lab
spike and the corresponding range of uncertainties constitute
the QA statements of accuracy for a particular sampling
program.

The completeness of accuracy data is that percentage of the
total number of samples that remained after outliers are
identified and set aside with Chauvenet's criterion.

182

Appendix A-2. Continued.

Table B-1. Chauvenet's Criterion,

CHAUVENET'S CRITERION
FOR REJECTING A SUSPECTED VALUE^

Number of

Samples

n

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

40

Maximum Allowable Values for
(Recoveryi - Recoveryayg) /sD

1.901
1.983
2.015
2.111
2.164
2.195
2.214
2.228
2.279
2.318
2.348
2.373
2.393
2.409
2,424
2.435
2.445
2.454
2.462
2.469
2.475
2.480
2.485
2.502
2.517
2.530
2.543
2.555
2.634

^Based on "t" distribution rather than the traditional "normal"
distribution.

"Individual Recovery = Recovery^

Average Recovery = Recoveryg

vg

183

Appendix A-2. Continued.

Procedures for calculating precision.

1. From the summarized information on the precision form the
relative percent difference (RPD) of each replicate pair
(blind field replicates and laboratory duplicates) are
calculated separately as:

RPD = |S-D| X 100
(S+D)/2

where:

RPD = Relative Percent Difference
S = First sample value (original)
D = Second sample value (duplicate)

Perfect precision would result in 0% RPD.

2. Any RPD value exceeding the control limit of +20% is evalu-
ated as an outlier using the Dixon's Q method. This i^ethod
compares the difference between the suspected outlier and the
value nearest to it in size with the difference between the
highest and lowest values. The ratio of these differences
(without regard to sign) is the Dixon's Q.

Q = [ suspect RPD value - nearest RPD value|
(largest RPD value - smallest RPD valuej

The critical values of Q for P=0.05 are given in Table B-2.
If the calculated value of Q exceeds the critical value the
suspect RPD value is rejected, and not used in the following
precision QA statement calculations.

3.

Calculate the relative standard deviation (RSD) of each
duplicate pair (field duplicates and lab duplicates are
treated separately) .

RSDeach pair = SD/Mean

Calculate the RSD for all the pairs
duplicates are treated separately) .

(field duplicates and lab

RSDoverall =

n

z

i=l n-1

pairs

1/2

184

Appendix A-2. Continued.

5. Calculate precision as a percent.

. . ^n-lpairs ^ RSDoverall
Precision (%) = x 100

(n-lpairs) ^/^
where: t is value from the 90% probability level

Calculate the maximum uncertainty for any individual test
result as:

X + xtSDoverall/(n-lpairs)^/^

where: x is the reported test result

t is the value of the t distribution at the 90%
probability level.

7. The completeness of precision data is that percentage of the
total number of RPD values that remained after outliers are
identified and set aside with Dixon's Q ratio.

185

Appendix A- 2. Continued,

Table B-2 . Critical values of Q*

(P=0.05)

Sample size

Critical value

4

0.831

5

0.717

6

0.621

7

0.570

8

0.524

9

0.492

10

0.464

4

E.P. King. 1958. J. Am. Statisti. Assoc. Vol. 48, 531.

186

APPENDIX B

Influent Flow Rates, Precipitation and Water Chemistry
Data from The Swamp Gulch Wetland

187

jle B-ln

Swamp Gul

ch daily

■f 1 ow < m3

) at tail

ings dam -f

1 ume ,

Date

June

July

August S

eptemtaer

October

Novembet

1

36.84

73.97

18.92

24,95

18.92

'".>

36.84

7 1 . 68

1 7 . 95

24,95

18,92

-^

36.84

73,94

1 7 , 50

23 . 39

18.92

4

36.84

70.55

15. 15

21 ,86

18.92

5

36.84

63,52

14.75

24 ., 95

18.92

6:,

36 . 84

62.79

17.50

29.85

18.92

7

36 . 84

74.74

18.92

29.85

1 8 . 92

8

36.84

68.68

18.92

29.85

20.36

9

36.84

40 . 1 6

18,92

28. 1 8

27.05

10

152.73

1 7 , 50

18.92

28. 24

17. 11

11

122.37

16. 11

18.92

3 1 , 58

2 1 . 86

12

86 . 57

16.82

18.92

32. 43

2 1 . 86

13

76,27

22.37

17.50

34, 15

23.39

14

71.68

42. 17

1 7 . 50

32,43

23.39

15

27, 16

64.97

38.69

16.11

35.03

1 2 . 23

16

160.89

59 . 1 9

3 1 . 58

16. 11

37. 78

20.25

17

99.94

418.74

29.85

1 7 , 50

38. 06

18

171.62

731,79

28. IS

•| 8 . 92

26 . 99

19

127.75

358.76

23.39

20,36

17.50

20

78.90

238.74

21.86

20 , 36

16. 11

21

64.63

210.76

18.92

18.92

16, 11

.-■y4"\

57.43

213.56

18.92

18.92

1 6 , 1 1

23

54.29

363,88

18.92

20.36

16. 11

24

53.27

191.33

1 9 . 88

20.36

16.79

25

51.23

142.48

74 , 82

2J ,86

1 7 , 50

26

45.00

120.47

38.23

20 . 36

18.92

27

38.69

113.70

29.59

2 1 . 86

17,50

2£3

36.84

HI. 04

29.57

23.39

18.92

29

36.84

105,80

23 . 39

26.54

1 7 , 50

30

36.84

98. 13

22.37

26.54

16.79

31

85.92

20.84

18.92

»... ,~.. «.. ~» »~ »«..». w. »»«..-.«.«. ..... .«« -,

— _« — «.- ™« «.. .™. —

4470.74 1204.00

578.78

759.32

319.93

731.79* 74.82

26.54

38,06

27.05

36 ,84 16.11

1 4,75

1 6 , 1 1

12.23*

Total 1141.32
Ma;c 171.62
Min 36.84

Note: * Record maximum and minimum for the pe^riod

188

Appendix B-2. Precipitation at the Swamp Gulch wetland study.

Date

cm

m

Date

cm

in

6/09

0.25

0.10

6/10

0.25

0.10

6/16

1.02

0.40

6/17

0.30

0.12

6/18

1.27

0.50

6/21

0.13

0.05

TOTAL

3.22

1.27

7/01

0.13

0.05

7/05

0.13

0.05

7/09

0.13

0.05

7/10

2.34

0.92

7/16

0.71

0.28

7/17

5.72

2.25

7/18

0.89

0.35

7/20

0.08

0.03

7/21

0.10

0.04

7/22

1.42

0.56

7/30

0.18

0.07

TOTAL

11.83

4.65

8/07

0.20

0.08

8/10

0.13

0.05

8/14

0.71

0.28

8/15

0.25

0.10

8/16

0.10

0.04

8/24

0.05

0.02

8/25

1.07

0.42

8/26

0.03

0.01

8/27

0.03

0.01

8/28

0.05

0.02

TOTAL

2.62

1.03

9/27

0.08

0.03

9/30

0.10

0.04

TOTAL

0.18

0.07

10/15

0.08

0.03

10/16

0.38

0.15

TOTAL

0.46

0.18

11/02

0.13

0.05

11/03

0.20

0.08

11/14

0.25

0.10

11/19

1.07

0.42

TOTAL

1.65

0.65

12/02

0.36

0.14

12/04

0.20

0.08

12/08

0.18

0.07

12/09

0.25

0.10

12/11

1.52

0.60

12/12/87-01/11/88

4.24

1.67 +

1/13

0.18

0.07

1/14

0.38

0.15

1/15

0.10

0.04

1/20

0.31

0.12

1/22

0.23

0.09

1/23

0.13

0.05

1/30

0.20

0.08

1/31

0.15

0.06

TOTAL

12/87 -

1/88

5.92

2.33

2/01

0.05

0.02

2/02

0.18

0.07

2/03

0.48

0.19

2/04

0.20

0.08

2/06

0.05

0.02

2/07

0.64

0.25

2/08

1.35

0.53

2/09

0.28

0.11

2/10

0.08

0.03

2/11

0.10

0.04

2/12

0.30

0.12

2/13

0.05

0.02

2/14

0.41

0.16

2/15

0.66

0.26

2/17

0.05

0.02

2/21

0.76

0.30

2/22

0.13

0.05

TOTAL

5.77

2.27

189

— • c: O

est czi c-j

t£!

o -~i -o CO CM ro o ^t- h^ — « -pH CO «s- CO o <::• ■^■-

i%t o o o o- o i*o r-. o CM ■*•<:> c^ -!»■ c;. txi r-j

•c;- "»»■ ij::i

h!" C-^J C'sl

i

■^ ,~i o

C> -O '^

■^1 l>4 C-A *^ LTJ to t>4 <Z> 'Kf C-i C J -<J I Ti

c> <:> o- o o" <:><:><:><:> o" *o -"-^ .c>

C-'.i ■«*

t'-i ^•^^ "*■ 1." ■

to
en

U"S CP- -"• ■—*
*~^ r-" r.-^ o"

ira LO o- — *

u-3 o-

k:» r--. t:-t ->*■ tia u'l

-43 I— ^ -~t

•":■ C--J ■::::■ oa i: -i r.-t ix> cjo o~^ (■-■

o- r-~ •^■» cr-

<x: cni aj

•— • C-^l "*• O^ O' ct--

n*- ^-. r^. ^ < tri ho

C-.I C-4 I>J CJ ir-A C--J

-o K-> ■*•

t-M CO O- OTJ

r-.r u-t r-- cci i:--.f cr-

ir-.F U'l U"3 — ' C-l --'

'Z> O' cca

-o -,0 -ur f-^ c-nI t^. c-4 O' r-

i:-^ *4- t--:' o-- int- 1 —

CM "ir CO o-

CM

u"3 CM »r--4 "^r- 0-- <:> •"»■
oo o- -hO r*^ psi oa -o

.^i -.-. r-.

"

.^.'

0-- a -

-o ~o C-4 "*"

in ■■::::

0-"

c-i o-

•«*- C-4

CO

r-.

c-j !■•':<

«*■

ira ci--

«id- r--

-CI r--j

r^-j

(■.■-;. .r^

--.

c-M f-o -*■ c-4

M-i- ..-

»~i

-■^ ir'-j

O" O O "O <3'

<:> •;=• o-

a

CM r-- 1^ -^i Cr-

•^ CO

a--

tr>

ro

rti r-4

t*-

-O CO

■■-D

*^ 0--

U':i

C-l

en

U•"^

f-i h-1. !-■:<

r--

c--t

ir-f

o-

o--

I." ■(

-*■

1 -.«

r--j ir 4

<1

■o o c;- c:> O'

<> o-

■o

o

en-

■c:' -c:^

o-

•::> o-

c::-

O- •::-.•

o

o.

•zz-

.:::> .-:> <:>

o-

•-'*

o-

o-

■:r:'

•::.:>

<::•

.:::::■

■::::■ -n:;-

o- c:> o o

o- o-

o-

<r>

o

o o>

C:-

o- o

c::*

o- •:::;■

■~:

<.>

•o

o- c-^ -::::.■

<::■

<=•

■r::'

•c::*

C~'

c:>

<:.•

<:>

■o c>

0'

•~0 r-^ cr- if'S

ixi t-y ^■■> r~-- ■«»
■— « u-n CM -r-. r-.

CM — t —• — • ■^--

•~Ci •«»■■ •«t" t-"l'

-o "Bt- -o •ea- CO -xa r^. *=:> i--
r- t>4 ■— « CM -ci o- ho -<n ij-j
,_, t>j r<' r-j ■— ■

10

ro

1—4

4J

C3

(0

ac

m

'D

-^

>i

l=n

M

-P

CO

^

• r-l

on

Cn

e

3K

Ei

0)

JS

rr-

D

c;ri

an:

«;

)M

0)

4J

«?

C-W

ro

CJ

ey

»

c»

^3

UJ

C

10

rrl

i-H

C.1

4J

UL,.

ixi,

0)

I==I

s

UJ

w:

U.i

U-

t—

JZ

o

tXI

I-H

<i:

;x;

D

--I

o.

U

—I

UJ

a

%JL.

cu.

E

(0

UJ

$

...J

w

UJ

1—

£k..

•xr:.

<s:

<r

esi

era

•

ro

CO

c

0)
Ok

<Z> O O "O O' O O O* O <!:• <^' O- O O* 'w* O* O*

1 K> c-t «t- ro •*• -er ■— » r'> -*■ e-4 li":i -^■'

•r^ ••.^■' "s^" "^^^

Of <2' O* 'w* O O O O' O" O <■> O O*

■~- <:> o '■::::•
i:m C--4 —■ • ■— '

f^ -r-t ,-- 1 *-~l

O' ■:!;• cv

s-.' "C;* <i>

CO -^t- -JD -JLi U '

0~- -M CD 0~

CO o-i -c:- CO r- CO r-

irt Cf- --o r^-

O O- -O O- O C'

o* o- -o o-

O- -O O"

o- <:> o* -o -o "O o- o c::' <::> o- •■■::.■ o-

•o -o o- ■o-

fo 0-- r^ r-^ <r> r-'T' -.—t ■«»■ ira r~^ 0-- -•--< o- ■•— I c^- i>^ -nO --ci o~' CM ■«r- cr- ■»*■ --o •*■ r'^ to o- c*v -o co •«»• j ■
■.-^ CO CM CM CM ■--. "tr UT ir5 ■^" •«i^ r~* -o m3 i-T -^ o" cm co , cr" c 4 r--" cm cm •'-t 0:3 CM r--. to ■« LO r~- -c

■o o

•:::::• CO CO CO
■«r -43 •<! j;:!

O- C:- C> O" O-

CO 0-- Mil CM MD Ma

m:i cm tea

i:--4 <:::• ■■--' o- m:i o~- ■;::■ co -r-" --a u":t -ir;- 00 r--
r-" Ma -ca ■<» t.o -n »■■•:< c-4 m:i tra M""t -j:j u-a K'

r-- to M3 m:i -^ u-t

■J.:. -CI -o -J^t fj':t UM U"^

m::i m;:i -rt -t) u"i

h-^ -::;• rt*

':»«: LiJ ::«:: 1:^
LU fr— cc: «a

LU 03 C;i

■o U.1 c-3 m; 3L- o:r

t— H

en

0::

ca

om

3e:-

era

a..

sX

C_l

LU

a..

31:

ar

<t:

->

.. 1

LjJ

r'"i

fflt^

iXl

■••B-i

TTT

<t:

==-

=>■

r>-

oiy

10::; oc ct:: cm:

fe ^ '■'-

<■: <r :m:

3IC 3K: _J

crs CO oa

ina CO C'-f

I i CM CM

X. <L { I

_i <c «x <c -cr

ro CM r-j I t I

I ca aa ir,i ca tri ica era t:::^ t-O i:;:a crt

<i; <c <r <r

._« *_ CM ?M r<i i^> *sf -wi" u~.i u*.i
t I t I t I ( I I t

C-i CJ ILJ C_"t CJ C,-l K-J C-^i i;~l t-1

CJ C.5 1:"

0"3 c:u cjrj

hZ:;< C2t cui

190

-ja — * c-4 CM — •

m o« u..
-< b... i-(i
i:ra C3 c'J

■:::■ f-":' r-.i cr- ■■■■■' -a lo •;;> o^
■"■ <^ "S*" ij-j --■• o- '■'*■ I"- o-

Cvi cr- ._-. t-ut oci u":i CM o^ c"M f--i o^ -o -^s o- -j:^

<~! ,JI .4" O- C^4 r-4 t'Nl CM o -c

<:> -a*- "-CI r-J "«*■ o

r--t r-- --^ <-• '■-' -"• ■— • '■■^ 'w- 'O

•O •^•* ■:"■ -^^ O c::- <~- •:-> O* CI'

c-4 «-< ir-4 CM --• r-J ^'> o o- c--^

■c> <:><><> o ^^ -w- o- •=> <::>

O- --^ O- -"« o

o o o o o

(-v„ r-^ r-j

f-> »•■■■:* — '
UT US- ir--J

r-- CO CO

—1 -rt- 0--

|..-, f.^;, |j:-j

iXi CO -o I3--

r--;" ■«*- u"3 o r^ 1-^ r--

Cf- "d" 1^' --O
-ip u"^ tiT cr- ^';•

CO ^. tw

<:> •::::■ -c:' o-

o- <":> ''J^ o <^'

■!::;• -d -xi
CO ■<! 03

'■a txt cr- r-'

r-j -nt ao cir- -at ir-4 ct- in t-o txi **■ Cf~ -ja
•ej c-M r-j tn r-4 ro o- to — » !>■ -o c-m -^^

ex) —1 r--4 --1 C-J CM CM •— ' c-4 —.««*-■<*•■'-<

O' f-^ o--

—H -^ CO

C-^ C--4 — •

Ma r-" ir-3 x* r-- ^w

o- cr •^'

CO »- o-

r-- r^-- <o

U™! CM

— ■• r-- CO

«a~ o- MP -*- M3 o- — • irj C--4 ir.i

— t M3 o- csi

to ii"i -o -D u""j ■--' o r^' t-'j cr- -o cm o- cr^ r-'s
•nf ?-■ o- CO o- r-- -43 r-. •=> -^^ -j3 u~a O; co r*--

•o MTi CM C-4 Mia -o

o CO "*■ Cf- «*• 10

O- «*■ CO O CO

*-. —H *-. CM ■— •

••■::> -o o."

"O O' c^*
.c> -o -o o en- O' o o-

o" <:><:=• o- -o

o-

"!t- CO «=*■

tKi oa O"- CO cr- •— « O- O"-

•o o o- <::> en* -o -c:' -en- o

tC:- o o- -c:- <:;■ o o o o

■nt" -Bt- O CO 1^
O O O' o o
•o c> -o -o o-

<:::■ -CI o-

o- -£< CO M3 CM b-i f ■> CO in ho f^- fM <;• km ry- c-4

U-3 .^ .^ .^ ira O O- 03 iO IT? --♦■—♦ CM r-4

■^ CM --O

O- Oh- O-

cr.3

«it- "4- ■«- -c:- c~* ■^-« <:> C-i O

o::i itKi c3:.i co c-4

f."! -^o--o O'O'U'acM-^-'OOtraco
t-o -^-4 cf- un 1:30 u-5 r^ ao CM

*-.. i:--i C'-.i

c::. O" <:::• o- o- <:::• <> o- o- -o <:> •c^'
■M*- r*? "Br* ro t-*'-' ••=•" "*" •■~* •••-* -^ *'"•' '^

-^ t -4 f-' CM *-•

-4::i (■■- I-"'

o- o -o o* -—^ <:::• o 'O -o -o o o o- <=•

~^ hO CM h-^i «*- C-l ■*■ ■^•' ■-- ' *■■• "* '■■^■■' l""-" "^

— " r'" -" CM --^

■j:* u:i u~j tj"*:i "*■ ^^x^ -ci

:■■* c-4 u':i oa

r-~« c-4 CM CM r-'ii

L--4 c-j CM C-4 I---:.

•z> o o-

u-j -- < -o ro <:::• r- •■ ixi co -w o^ u~i o- r-o r •■•

c-4 r<f CT3 U-3 c-4 C*4 CM CM C-4 ■-— C-4 CM V-'It CM

" c:> -o c> <z:' -o o o •"■ o -o •■■:.> •z>

:-4 fJ- t -4 •■Z> t-'.'

-■■4 c-4 C-4

r-- C'4 r-4

•«*• urt »:» r--

4J

c
o
u

t- 4 1-:'

Csi «3" r-

-o txi tn M-.I

-Ct c-4 r-4 I---:

r ■

*f ---.

<:.»

■*■

tea

■0

o-

■-0

tr3

or-

C.4 (--"

(..-■:,

r--:-

C-4 t|-f

C^4

u-i

L--4
CM

0-.

ro

U"3

CM

-0

CO i;t-

i:i^-

ir--4

•0
-0

■0

•3-
CO

0--

•0

o-

-*■ CO U'U -,0 CO ■!*•<=' -"O

-43 03 -o !-:• --o CM ti-n

C-- r-^* to r-

~4:j --CI -~o -JO t'

••JO o- r-- "is- C-4

-o u~J ir--i r---.

~.o C-- I — c- r--- r-'

tri -*■ -*■ ^■■■' r-> -sf 10 ■«*■•«*• u-i 10 to "*•

■-0 -* tn trt K> to M3 -sr- -o -43 in --CI

—* CO o
C> "St" o^

-Ci •«*■ 10

I--, (-.-■. r-- r^ r — c--- r~- r-- r-- r-^. r~~- 1 — r^ r-- c~- i — r^- f-^ f- f^-

CO CO i:xj txi CO ci::i 0:1 03 »xi oa 133 as i:i3 00 tri m CO CO a;i TO

cr' cr'- cr"- rr o^ •'' •"■• "^ ■"■• --^ o -o -d- o o^ m oci ca tp co ^

c-4 C4 C-4 C-4 C-4 f.' l^'J> r-^' k:" t

IJ-3 uM ij":i iD r---

r -. r~- r-

r-- c-.. r--- r--

|v^ ho ro ro c-4 CM CM CM C--4 c-4

f-_' f-^ 1-^. r-^ r — r-^ i — ■ r— r~-- c-

o- <z< o cr:-- o c^' o o* o -o

■-CI --O 1/3 -O 10

M3 to O- CO CM CS|

■-0 --o --o u'a --o -CI

c-., I-:, (-~.. r-- r--. r-^

CO m CO CO CO 00

cr- Cr-- tr- cr" <=> O-

irM C-4 c-l C4 t'Tt CM

r-~ r--^. r- C-- r-^ r^

o o- o- o- o ■o

I

i:^:: Cct i:xi

■::»> iA.1 c:i tjn JW Cc::

r.cj 3E.* c:a ce;

C

a
<

<X'. ii..t ci. ci... 'JK: ''X
:h: :::=- l-"-i

,..., (^ .:aj 31. i:ia.) ar;:

1 iLii CB.: cc: ct:

iX. Ct.. IJ. Ix.

iJ Li..( I I

a.„ li.. ac: «:
u.i ui 'O: -"nc
iM.1 uu ist. .:»&

( 1 c-4 C-4

trra iX* c-4

I ca era u-3 «n=i co kt-j era cz:, tj

fr^ Oy i-f^ tx» tx» iXi ecu

„i -^r -cc <S-

lU t-l C_t I

CM C-4 tTf f ■-:

■■x .■:T. <T <x: act act iXi ipct ix-t

191

s s-

loo-— 'C^f'j— ia-OD «>)

<S' (>- o CM

iri oo in -o

i>o l-» IfJ — <
<:> lo -^ <^

r-^

^

1^

o "*• in

*-. c^^ r--

O^ C--4 O

•=::•

1

MTI C-^ lO
1

CM <:> o
1 t

<=■ --I --
t 1 1

-a --o t*o r-~. *-i o> r-. -e:?
r--" r^-H c^3 o •-- t--' ^o —^

o <^ o

c^ csi r->

o o o <=> o

I--3 "*■ t-ri r-"< h-':*

CMiraooch-i>-r-.t*oirt

fM

OP ^- cr-

■sf-^ r-or--o^in^r-rt-o-d

to — < H--.

u'ti iri r-4 o r--> o-
u-3 I---"

*■-* c-t

OOOOOOOO-OO

o* o c:? o o

:;:■<:> O O 0> <:>

■o

s

era

O
CO

«i

t

"O

i.,.i

■4^

m

tiS»

na

1—

i-j

zj

— • cvj un -.-^ -"t •.»■-•-> c>4 r<> —-I ..— 1 *-i CO ■--• -.--I c^j !«■:■ ■^■« "*■ C--4 c^j ,-. •-h ._^ i>j *-.i .^-i *-« ■.— • -.-■» -—i cm

-— • CM

<:> CO •— • M3 -^3 ch. ro

-o r^ r--. lo

r-- CM u~i

t^o OP r-- ho CM in i"^

I — CM r— ■

r-.cM— •coKioar-.O'Oaoa roir>

CMr-~. --or— u"3r^ Opr-^cx30'-'--»CMo-i*--ouraO'0'ixiixi

f^> ,__,-, C*^ .,—. Ki CM r'S ^< .r~« — « C-'J ■•—• -— I -^-t ■— * C-J ■— * CM

o- o o- o

s

i

03

MCI

■«f

CO

oo

CM

to

to

"sr

<9

K>

-43

CM

•^ f-->

to

o-

tr-4

s

in

03

M3

in

■-0

r— -

CM

to

U"3

--a

Cr-

CM

o

o

•o

•o

o

o

o

o

•o

o

o

O*

O-

<::>

•O

O" o

o.

o

c?

o- <:::•

o-

•=:■

o

o

O

o*

•o

O

O

■^

O

<;•

o

o

o

o

o

<»

o

o

o

o

«>

o-

O

o

O

o o-

•=>

o

o-

o o

o

o-

•o

o-

O

o

Ci-

o

•C:-

o-

O

<;;•

l*0 MO CSI -O
O- Ml O -O

CM in O- <3 o

o- ro DO fo — •

r-- 03 o-

coo-ocsi-oo-r-. — -^-^
boa3-^o--"8r'cr^mcMCM--»

■— ' — ^ CM — • -^ -.-. -.—

•'^t>4-^mf-^^c^-^c>4-~*''~-* <«a-*-4CMr-4ro-r-H«rcM-'-'---^-'-«-'--*-^''~'---*'-~«03in'«rinp^

■o <^ o o ^o o o o o o ^S" O" o- >^ <^ -o o <:> o

•O O' o o* o o

fo ro o h-j CM

to m -o c»

o..^.,.^aaro*-4«^ m-^rtoc-Jroto— ^oatOLncaLn-jao

O" ro ■*■

r-:. ■(»■ —« CI- -^o CO u":i U":i <:*•■ r--. r~- f~- »:»- ^o r-~- r--

jr ?

r~- -o -*> ■«r
CM i*o r^ -

CO ro MO o- f'-:'

— « CM CM CM K"

CM .H

r-4 C-4 CM C-4 -er

-.::> u"3 -*3

o

pa .r: IS

aal- "CI- m C--4

o oc^oo-ooooo

o -o o

O <S' o c^

LU ac

-1 -^

O
-IT

ro
ro

P>

0--

to

0--

tn

;?

03

in

■**"

CM

to

to

CM

-a

o-

ao

CM

o-

CM

o

03

^4

o-

CO

u-i

in

■ar

o-

CM

CM

CM

-JTJ
t-'S

CM

CM

CD

ir:*

C--4

CI

r-4

IM

t<'

to

•-0

»n

> o

o

O

o

■=>

<::•

cs

o

•o

O

O

•o

■O

r-j

^

o-

o

o*

■o

o

o

•=::•

o

O

O

■IT'

o

O

O

O'

o*

■O-

o

o

3 r-j

CM

■Bi-

■d-

r-.

■-0

o*

to

•iT

to

O

oa

O^

■!*•

0--

o-

^-:

«":

■u-

■Mf

■O"

r^-

u-j

oo

to

1^-,

m

r:

ro

o

o-

U"5

c;-

r--.

0D-'-4Ot0CMr^h0C^

— •--•^••r--.oirtr^-ci

in o-- o "«»■

■o o o

ho r- J

CM cr-
-J3 CM

to CM M3 -sO -O -O t-"^ «*■ to l^«

■c- -o* o* o

r^ K-M cr- ^

in*>om-o-«..oinM3-«in

r-. fo .JO

f>. «st- O^

-xs -XI •<» --JO ■<% -^ -^ -Ki -a -n m -JO ijn -tf t^.i

43 irM -i3 i>4

-43 M3 -43 in -hO "W- •«*- ~J:i l/TI -O -40

4J

c
o
o

r^ r-^ r--

I— JC

eri to

r^ r^ h-H.
•o o o*

fn t^.' fyj

r--. I- - f — r^ r — r^.

o c> o- o <p* -o

,-. —I *-* -r-» rM rH r-4 r^;t ro C--4 ra CNi ■—«■•-• —«

CO
I

CO

no

C
0)

a.
04

at U3 ga tn
(tit

■— « *^ CM CM ro •— » CM

A.X

i=a en (» en

J-I-. Old Uu

c/a t 5 en %

t in u» I cjtt

— . - L •— 1

►■-» u':i cn tt::

:>- 3K:. tn C'j c.:i

<r -a:

3m ijj o... <c <:r

!--»::> Ljj UJ

ar: CD iKi ac 3R::

«u <r. r*- ««.• '31'

ji ix: oc cc

<Xr CJ I

I

U~ U_ U^ 1

CM

':^ ::^ 'M^ I

1 I « I I I I I ■ I I l.JU t-»J ■•*. •— ^ I I »-* • ■>"• -^^ —-' r-^ --* ~a

esiisaeeiUJi4biuJU.u.u»uu.u»tncn<c: i3Qc^cJie-:> u~ tn tn ao ixt 0:* <c

c=i iLn CM

cn I I t

I ..^ ._ „.-,

CM I I I

I ILJ C3 CJ

192

?; y

.— • IT O

ura ira o cr- ff"

-Bi-"sr-o-*--o™*-e^r--.

r^u'3o--'.-«c>c--4 0--cvico<:>c>r-~car^O'CN<

csj i>4 ro r-j c^i 1^4 CM -^ cNi iv-^ i>4 r<" r-4 k:« ro €--4 cni i>4

o- cr- r>4 ro o "*■ to r'> -JO -(*■ oo oo pa K-i

Cs( u~j -JQ tfj ira o

^.^ ._, .,-1 ^^ -r-) ■r*^ tr^ ••— « CM CM CM

CM O OO O

_ —4 •-» CD

•-* *-i C^l CM —-t CM OD

O 'Z> •":■ -O

M3 r-^ c^ -o o

.=> cEi r- ■—<■*• CM --o CO ho CM o- eg ■;:::• CO
^■^ r-M o- cr- u^ •=:> M» f^4 -o c-4 -^ r^ <:■ -^

C-4*-HCM--~*-- •— •CM-'-*—- "f^JCM-— •— -<

mf ^Z" CO r-— CO i-*~

-03 1^- !•'■:• a- K' r-o

--« C»- M3I CM t-O t~-*

O^ ■"»" OO '«»■ CM -O

o- m:j oa
u-j o- cr-

c-4 ■— — *

.__,-, -j3 .p-t CNl r-. CM

.J3 r-'j o~ oca ■•-« oo or

»-^ *-. CM •'-^ C^ C-4 C^

ira »o ITS ITS or- tra -^

ira r^ ho o -"O o OD

.^ —1 Osl CM ■— • CM CM

CM *-t ■—»••-» —•

o* o o o- o o -o o

o o- o o o <=• o-
z> <:> o o <^

•vr* "•■^ "v^

-C3 r-- CO r-. o- MO
cxri •<! -^o -CI ira -o

•o <:> o o o o
O' o- ■=;• -o <:> o
^~-tXl^•^cMr-.CMCOCo^o«»-oC•

0PO2I— •u~ao--r^ oca mdco-o
M3--Ci-<a-a--o-ci-<3r-''*o-— •

•^ M3

c? o o -c? o o- o

•o o •=>

O O O" O' c> o

c> o c> o- o o- o
ira o -SI" ira

CM

•c? 0-- -o o-

Ki «*• C:' r-

ro o- o- <:>

CM CM -o OO CM
~i •—■ CM

.3.0''n='C>'C>'C>'C>'C>c3'C><:><=>c>
r^ i,~J ^ CM -.-•■—• c^-J .--.*-* fi" -"t --« —•

c;:. <n> <:^ <:>

<::• .r:> •:::> O O o- C:- O* ■=> o <r> C> •:;:> O O-
ira •«" MD uT M3 iv^ ira r-^ tf^ oo r-- r--. lO r-w aa

O' "O o*
.3j" r-- o-

O'C^co-efcoKao-Maoa-^j^il^l^SS^

<-;• f-'a .o u"j t3ta «*■ --o --• -^o r-- cxa f^- CM go
(:;i ;-M ,?::; tL; c-4 -i-^ ro CM ^'> -4- K> CM CM CM

Ota I*- CO CM

»«

o^

t-a

&

ua

o

c»

<C

C.J

«ffi 1-

.._j

U.J

m ■--

ca

Wi

....J

o

UJ

j=: «

LJl

« L

u.

LU

e ■"-

uTt -.ts c-4 ■"-* in -JO v^ ey^ '^ o ■^■* ^ iJt'i ^ t^.

•o o o-

O- O O O O' o o- o- o- o o o o o — •

•CSOO'OOOOOOOO'OO

to o

•0 CM

— •ou-atrao<:>c>oO' o--«<:;'U^ira-«t-

O" •=> r-^ uni

--t c-i O' f-"' o*

u,. I— ca

.— >

cu>

<:•

<r>

o

O

O"

•=>

o

f""

r^j

CM

CM

r^

o~-

-ar

li-a

f>:<

r-o

-O

ira

•><-

!•-■

oa

o-

■o

o-

r-:t

-:*■

c^

-t:a

-o

ro

-tj"

•XI

ira

■o

OCMl^-^oc^t-OKsc^-r-^lorohoi^'Jrof^

u-3<«r*-o~<o-^-Jca -jniro-o^o

■^a»r3--a~<a'^o«'^^'3'^'-'«>^f^

C

o
u

COfXIOCI COCOIXIIXIOOCO

coooa3ixio3|0ooaQaroDOoo
cocoixicooncciixiixicriiDocti

■.^ i>j oj Kl ^-3 h"^ «>j c-^ f^ "^ »^ ^^ "^ f*'^ "^ '"'■' "^ ^ ^ ^

o o o

oo-o-^oooooo

I

I
I

•iH

c
a

Cu

lura LO c=a en «:» en
I - I III

..^ ..^ i>4 i>i .«i- .«r (

193

I cat tn
-^ I I

r en ca <n ca en 1 i

.1 I I t I a. Ou.

-^CMCMCM— « ■— <CMCMUJtAJ

icai UJ (£51 ca

Uh. Lfc.. u. uu exa en

•O O O O -O o

•O O O O O" o

o o o*
o- o- <:>

<:><:>%*•• •V-

O O O' c? o o

" ^ r^ CM CV4 c-^

o o- o o

■v^ '-v' "-—^ -V-

—I <::> <:3i o.

<5> O"

CM C--4

•c? o- o

CM CM r-4

•O ■=:• -O

•o

•O O O"

O* O" "O*

ir:i uTi tri M"3 in u"j ---^ u"3

OO-OOOOCMO

•~> o- *— • -o c> <:;• o o

un unt vn
o o o
o o <:>

<=>

o-

o

o-

o

<:>

<;;:'

o-

<"•

<=■

o-

<:::•

t^

o

o

o

•o

o

•o

o

•o

o

o-

o*

o

-v^

•o

•o

o

<:>

<Sf

o

o-

■o

■~^

V

N^

c::-

■rrs

o-

o-

o-

CM

r-4

C^l

CM

i:-.i

t>*

CM

CM

l>l

.r::^

o

o-

<•:»

<■>

o-

c::*

o-

o

o-

■*-.■-

".x-

■-.■■■

-^'

■■■.-■

-■^-

-•.•'■

■--..-

•^■'

■^^■

o

•c

^^

•-—■

O:-

•o

en-

'Z>

c:>

•o

<::■

O'

o

o-

■o

.:n>

O

•C-'

o-

■o

■:!;•

■o

■O

-^■

•%^

■■V

■'■--■•

■-...•

■■..^•

■•v-

■-^■

-.^

-^■

■■•^•

■^

U-3

irt

UM

in

UT

tn

^.

<~J.

•.-:;•

«-•:.

■Ct'

•O

■c

o

O-

o

O-

o

o>

o

•o

'C:-

o-

o

o-

O

o>

o

<=. ■:::i;- ';:r- -c^ ■=::>
i~-j j*^ CM rj c--)

O O O- C:- <3

c;:* c:* o*

o- 'O o o- c;- c? c?

f^ — , ^.^ .,-. CM -^ -—♦

Cf o o- o- o o -o

o o> o c::- -o <;> o-

CM CM CM CM CM CM CM

O- C> 'O o ■ o- o o-

o o- o* <;>

■■.,- O ■v-'

<3' O O

o o- o*

■^ r^. CM

^ O O"

o- <::> o- <=• c;. -r-i

to

in

in

in

m

in

in

m

0.

•0

--,

0.

■0

•o

0,

■0

on an

<c «s

-s,-*

■v*

-^

•V

■s^

-^

■^.■

■v-

•%v"

■•v.-'

in

m

m

tn

10

■0

•rs

■0

.— t

•0

o r^j -o

in m u~j m ii"3 in u":< tn

d?" "w* ^I!I' 'h!2' C^ O^ ■''■^' 'Ctl'

.C> -O O O O" O' •"•

O- O- <=• O O- O- C:'

<:? o c:^ o

u~i tn ya u-j ut

O' «' o- o- o-

tn ir:! ii":i m u-j u"3 m

<;. o- o o- o- o- o

O' o o o- o o -o

■::::;■ o* 0- o- -^

ii~i tea g"j m y~> g-i

<=. <:=. O- -rr. <=• -O

c> ■:r' o c;-

o o* -o o-

■o <:? o-

tn in tn uT K"? tn tn u^ in
<;:> <:;• o- o* o o o o* -o

O O- O O* O 'O o o- o

<!«

<*.<:><:>c>c>'C>c>c>c><>c>c>o><:>o><:::'C>

•.^,' ^.^- ^.^• •^.•* •^^• %x* ■■^■' "^-^ ■-•^■' ''■^■' ■■*■■■' '-V

to ro ro

O' o c::-

•o -o «o

in m u~i u":! ut in

•c;:- <^> o ■:::-• o o-

o- .0 -r::- O' o- o-

c> o" o c"^ ■Z> 'p

h". r--":' b-"^ ^■ J r"' h:*

•ni:^ <_-■ -L-'

CD .^ -^ ..^ -^ ._-. ._^

^o o o o o o -o

O' C> CT' O O "J^ O

o* o <:> o

o* o o o*

CM O O' O 0«

o- c» o •=• o- <:> o

ro CO ro r--

O- O" o-

o- "O o o- "O o <."> <z> c:- ■^;>

Is-, ^, .^ -^ .^ „.-^ r^ ■.-■.
-o o- o c> c> o* -c? o-
o- o- •'-• -o o- o c? o

o" cr"

._i..~>.,.-4-~irM-»-i-.-4f^-~4

r--. c> o "O -s- o o o o o*

o o o o o
c> o- O" 'O c:>

•O -O •=> O' o-

c> o- o c:- -o o -o <:=• O" O' o -c:;' o
o o' o- o- o O' o- o o* o* c:;* -o o-

CO CM r-o •«*■ -a

CO O O' o* c

Kit O -O O O <=> C>

cr-"*-ocr-rM"«»'ro-<5 -r--* cr^ —-t r--~ cr- 00 en --■•'«*■ co •—• cm o*
iXic>ixiun(^4M3^ooo--j3-^u-3incci<:>--oc'-Nic-4r---ca«*-i^j

o -o — • c> o o* cs- cm -ir CM r-":*

O O* •«*■ l^» U"3 f^- r--4 K*.'

CM -o r4 u*J ro

■*• -nO -CI

00

r--

'i^I

■0 o-

K.t mD

M»

CM

CO

r^

C-4
ITS

r—

r-> CM
ir -r C-4

•o r".' ro CO

o- <=■ o- c

irj o- cr- r~- urt c> -o "*■ -"•
<:" -d^ o- c> <^ c:> o" o" --o
:■ c::. c;- o -r::- o- •=» o c:- c:- o •;:;■

IX' ixi rxi •"> o- u> <-j>

—t C--J -v-i ■<»■ •«»■ r--.f ~-t
o o- cj> -o -o ■::::■ o- ■;::::•

■c;- cr.' -c^ c:>

C--.I C-4 1^-4

c-4 irii f-i CO -'■-• t*^

4J
C
O
U

CO
I

T3

C
0)
Q.

r- r~^ r-

r^ r-^ i"..„ r-..

O^ O" CM C-l C^J C-J CM C-4 C^J

CM c^-j .:z> ■:::> o o -o o- o

-*- «d" in in ii'> ir» in in in

■--:::■ ■=:> -o <:> c;- o- o- o •=:•

1:1:: »-D

o- r<' t-")-

ixi en CO
I---;, r-^ r-'^

I--. I—, r^ r~-

ixi CO as en CO ta;:* oa

k's O -w.

u^j in in ■* "*■ -ot- "kt "«*■

tn ■'<*• ir".i ■*■ ■«*■ tn M"\ mtw \r\

iit" Lu -cm:

LU I— cc: ixt

L4J era c-j

cr. <t: «=a

x^ iKz zm oc^

>-■ CD

CO <r cna ix:

0:1 Cl- «x t->

U4 O- 31:

IJJ

<r

3:n C3 iBJ laa (B.I x:
"Sij ci: rj- 3^ ~>- «.'
_,j ...,-i la:: ex.: <2c. ct-:

tao tiu cici «» -^.-c'

,..j «:t' «.^j: -x -X

I I

I -I CI

•«r -* •*>:■ *»

^

194

eg

CM

o

O

o o

o

o

-s^

o

.^ ■s^-

CM

o

CM
O

CM
O

CM
•O

O O O O O- 'w-

■— « ■— I --t r~- (Xi *-•

?

to

lJ-3

lO

o-

o

O

■o

•s^'

c::*

•^v

o-

o

o

o o ir-j

^— .

O

■r-^

•r>

-.-:>

o

o

o

o

o-

o

tf-'

•o

O

o-

o

o

o

o

■o

o

■o

U"3

CI-

C*4

o-

r->

■•"•

•—4

*-H

r»io.O'CHC>c>c>'C>'0'0'0<><»0'CS<:>c>'C>iOC>o

■o

ITS

U~.t

U~.t

o

-.^■•

■,^-

•^-■■

o*

c;-

-..-

lO

o
o

ITS

o

U-3
O
O

trt

u-3

e-3 fi-

•O <Z> O O C^'

.,—4 „_( .^^ »-i (*o •»-* CM ••— « —-• ■»-* ""^ O*
•C:. O- O O <=■ -O — ' O' O* C::- O" •— *

o o- o- o o o- c;- o* o-- o- o <:>

•o c o o

••»■ CM Cm cm •*-* ■•—* ■•■-' *"• ••"• ■•■"• *•"' "•"• ••"* ***< *-*

— ooooooooooo-ooo

OOOOOOO'O'OOO'O'OO

•o o- o

O* O O* 0> O O "S^

c."^c>0'<3'C><-"C><3.0'<:>oi>0'"C>'0<::><:>'C>0'0'<C'C>c>

-v^ -s.^ -s.-

C--1 0-- CM o o <3' o o- o> o- o <::;■ o ,o 0'C>0'C>0'C>c>c><3<<r'<oc>0'C><>'C'C>c><>c>

<:::><n><— .c5-c:>c><3'0'0'C>0'C>c>c><><3><3'<:>0'C>0'0'C><2'<>0'C><30'0

o> o> o

0*0

<:>c>c><5c>o*c>0'0'<:>c>c>0'C><>0'C>c:>c>'C>0'C>'

-^ -s^ -s^'

•s^ -S." "v^ •%^ ■•J^.- S*' -^^ "s^

o> -o o

O OO -"T C^-

o^ r-^ M3 K"

oooo^cMCMO^irai^r-^ioo*
*-.<*.-r-io-.o"*'rocMOO'«--« •^■»

s

■— •hOooPCM roo-o

0'0"<3-0'<><>0'C>c><:><>-^'«r-BhK>---*c>0'<=>c><:>'«rhotounc^

?

O'Oroo-co <:>--aMaM3M3"«tc><::><r>c><:>0'<:?oaoc>c>o.oc>ooc>
CM i^"? ■"■ CO CM eta c>. o <:> o o u"i o CM r«* c^4 r^ -CI -jp CO c? cm ro co r^ ho OJ

oo O"- o r-- "«*" CO o" <> o 'd

0-- -o -*• o-

-v'

-S^ s^ s,^^

4J

C

o
o

I

t3
C

a
a.

1— Kr

C»0 O'-O oo-o ^^^

mm "- ■•- ""i^

"u^' ^.^ -^^ '•-y* "^H^ '*v^^ *■*•

f-~. |-^ r— , F"— r^. r^ r~- r^^ r^ r*^ r-»- r~- r~~- r^ i—~- r^ r^ r~-- r^ r*~ f^ f^ r** r^ r*^ r^ i*^ r^ f"^ f^ t>^ f^ f^ f"^ •"*• r^ r**

crio3a:«iXioocooDtx)criCDa3030300coa3COCo«»crii»ixiixicoooiXiix>^ Oft oo

ooo OO" CMCMCMCMCMCMl^•l<»ro^o^o^o^■:troI^^^CM^MCS(CMCMCMCMCM^^cM^ocMC^|v^lr^

-«_ -^^ •--». -— ■— ^ -*~ ••-». --^ ~-„ -*^ -~^ ■•»-. --,. --«, - — ---. -^. ■>-. "-^ •*-™ *^^ "-^ •«*, -^ --^ ■*^ "-~. -^ ■ "*-^ -^ *— • ■*--. '—- "*"-■ ■*-* "■■*"■ ■*'^

uj ex:. C)C ca

UJ UJ CJi

CM CM f^

Cu a... -x:.

O UJ CJ CO

C--J •— t CO SET

ac CO >"

>~ <r <c

•=3: UJ ix. ex.

CD isiu se

i=Q ™l

<C r> 133

—J cc: cc:

:3C 0=:'

CD

■sr cj

UJ

UJ

mjf :acz

ixi

IX. LL. LX,

t

1

i::::a

CO

t:=»

CO CM

CM

U-

U^

r'->

r-o

t

1

c-a CO
1 1

t

1 1

ica CO

u^

1

CO

t

T

*f

1

C::! CO

1 1

1

da

1'

<a

•ta-

•w

«a:

<r

1

1

fv^ CM CM

1

1 1

CM

C>4

t^

r^>

■"T

PM

si

^

s

^ ^^

^

^»

ii

LJ

U

1
CJ

1

CJ

CJ

CJl

icl,

a

B A

ea

CSX

135

<3 ^ <? O <? O*

«■« 4^i^ vN ■*-• vx -r^

<& o o o« o o

s

ar

4e

ff

«

i"

s

-s,.

?

»

en

o o o

O <^ O O O Oi o o o o o

<5 O O O- O O •i'> o o <«•
O O" O O <3* o o o o o

«;> o o

o o o o o o o- o o o

■O O* O" O O O- 0> O O"

c-ooooo-oo-

o csi c;> ■•--* o lo o- O' •i^ o- o o h'-j ■—* o o o* <:> o o

•^x' ■^^' -v-' "■^.- ■■■^* ■^^•' ••■•■* --^ ■■%-•■

in

o in K" ro 1^ ro
o ro o o <z> o-

•O -O O" O" O <Z'

■—•—•—» *-t •—• -o o-

o o o- •=> o o»- "^

"" O- O <3 O ■— • •«*•

■*-^-.~l|^..--<t^..-^>M^OJfO-.-4-<~l|^-->~>->~4---H-.--l>««C^I^

O- O O O -O O- l>4 O' -O CS- O- <=• O O O- O "Kf 1--"J -O O O -O O'

o o o *o o o O' o •i;^ -o o -o o >::::• o o o o* o* o o o o-

.c^ <:> o o* o

•o o c;> o o*

O O- O C> 'O-

OO'Cl'OOO.O'O'OiO"*'

<^ o o- o

o o <:> o o "O O' O'

O O' O O' O'

O "O o« -o o

•O O- CS- O- O* C> O- O O -^ -O O O- <=:•

•O O O -w* O'

o o o-

■— • o i<i p-. .JO ro cvi

r^o*- «CNiO"<r"'*roo <::>oc>0''00'0'0'Oc;*c-sii.-Nif.^

o o cs-

£ ?

■<J^or--■<>^^iln<:>aoc>4-o^^4lnoac>c>o•o■^Xl■«tt•

o o o-

; J

■s^*

-...- •%..- -■^^■

-t CM

o o- o-
o o o-

o o- o

•o o o

r-^ CM -.o -c;-

.—• -^ -^ lO «4~ to

<n> <:> -cs o o o

■o <o o o

^. <:^ ^ ,;:> <;:> <^ O" o cs o*

o* o- o o-

O' O- O -O O "O o

o o- c& o o •=>

CM -—•CM CM — * CM

■O O O O' o o- o o

<:> <=> C:'
C--4 -^ r-^

<Z> O O- O'

O <=> -c:' -O -O
■^ -^ -H c-< •^'

4J

o o o
(^ r-o ro

r-> r-. r-».

'O o o

C

o

t

ro

CO

t^-^ r*-. r~«. r**^ r--. f-^ i~-~ r-^ r*^ r*^ r~~ r*^ i-^ r*

•O* O O O" O"

oo-o-oo-oo-o-

r-j tM r^ r-^ t^j r-^ ir-4 c-j — »

•"• O" •;::> o -o

04 rsi CM 1*^ f-o

.» ■=:■ O

c

0)

C^ tn c::st tn

Cr.' Oi cni !=:»

o uj t-3 ctr oc

CM -K

•— < to os: cc:

>- 3e: CO cj t-T

3C uu cl.. MX <c:

rx:: CD mi acn ac

u, u- aw <e: r>- aaj au

ce iA_ ce:: lj- •—•

=3 ce r3 u.

CO =D en cr _

c=i ert I CO I lo is-3 z:< i~- i- i~ «=*:3 »:/:• i=a era

\r<i tjn \ t.r.t tx. ex.. u li . u- i t r::^ en

oc ua

ckt err:

IX. O-.

- I

mat iO %M

t'-* UJ UJ <t. _ . . _. ,

l.'^'^-i^.'. ,L S ,L a*, S isl

196

■:m.: ::>-:: I

C--4
I

I

I -.-^

r-j

I

KtL "it:. i»* tm

fici ia'

citu cj> tn ■::» CI? cji era

cji cji i_ji i:»

^ s*

m

■O C;' o o o o o o o o- o

OOO'OOOOO'OO'O

OOO'OOO'OO'OOO'O'*-"

o o o o o o
o o- o o .o o

.^ .,-4 —4 *«i *-• r^

o «> o <? <:> irt

o o c> <» o o

"O o o o o o

%•• "S^ -s,^ NX- "v^ N*^

<:>

o

o

o

<»

o

•»

■o

o

•o

•o

CV4

«•

o

<r>

o

o

o

<=•

o

o

o

o

o

o

•^.^

•v^

•s--"

•NX-

■s^

NX*

^^

•V

■^•

oo

r-.

0-- i>i ••"« ro "*• o CO

■J3 '•r r»i lo m

ro -^o CNi h--^

o o o

"*■ CO CM !>■ *♦" fo ""f r^ CO <:> oo o- --^
•^rroroi^oro— <cMCsir^CM*-<c>4ir3

en

u..

en

t3

-o

•o

CM

o

o

o

<:>

o

o

o

C<l

— 1

ro

.M

^->

— 1

4J

c
o
o

«=. o c- o o o

04 ■•--•■•--•'•»• -^ *-

ooooooixiooixiccicoaaaoaocciQooo
ooooaocomcoooixiaoaDcoaatxicD

OCS'O'O'O'O'OO'O'OO'O'O'

CO

I

CO

c

0)

a

*^ -^ CM tNI CNI

I I I I 1 -_,

ca en
ca en 1=1 ixi t I

I I I I a. a.
— < — • rvi CM uj UJ

I .' .1 .! iy {}^

197

o

D
O

a

E

(0
CO

(U
■p

■p
CO

CO

c
o

•iH
4J

nj

(0
0)

a

e

(0

0]

•

m

a)

o

■p

•i-i

c

(0

o

-H

TJ

JJ

C

<0

«5

>

iH

Q)

4J

iH

0)

b]

^

•

■*

1

M

c

0)

04
04

<

c

3

^

n

> (B

01 CD

"H >^

uj m

»-»

OT -^

3 -<

C

C

■u

(9

> t^

01 CD

'-* "»s

UJ n

<-*

in V

3 CD

D t^
E O

C ~v

in

0. 3

O QD
U N

m
in X

3 rv

>

_ OD

m
in X

3 IC

c
o

>

-, 03

UJ -^

(Jl

01 nj

tn V

3 r

-<r oDuioj.iiajrnj-ii^n

(Dt\j f/irfiE^r^nict^omcn

030) ^^^N^^^[^^Dto

1/11X1 liitni/iipuii/iirinmui

o in
n i\j

mooiiD-'r iDin

!^ ?• iT ifl

r !• UJ ID

iDicinininx t ^ ^ m cu citn
uiinininuiui inui in ini/i i/iin

Sl^

in o 3" IV r M
r ID ic 01 CD m

in in

rv IV |v ID ID cE

rv IV |V pv (V tv

in in in in in ifl

to roo -n

|v tv O 01 IV IV

m 01

Oi "

10 10 10 in in in

|v |v IV IV IV |V

in ui in ui in in

f-t (H v-( «-l (-^ v^

r r

IV IV

in in

IV tv

in in

♦1 <P4

o

to IV

m ID

03 01

(71 r^

IV IV

tv tv

in ui

IV tv to (0

|v rv IV tv
1/1 ui in 1/1

3" rt !•
IV tv B

tv -<
01 m

o

to 10 in

IV IV S

tn in in

ui m

tv

1/1

mtviD-inrmrrKntvCDOiivmaiaj-i/iiviofflrtrrtOim-ojoj-
i/iroDytninmoioimtvCDoiiDrnooinuii/ii/iaiCDOOiivivojmnji/i

a)tDic(Divrvtviv[vivf^iv(DiDiflioioinin3'T'iflioiDtntnininj'rfn

[vtvivlvrvfv(vfv{vrvlvivfvlvfvlvfvfv(vlvlvtv(vfvtvfvrvlvfv(vis

inininirii/iuii/ii/ii/ii/ii/ii/iini/iininLni/iini/iinininini/tini/itni/ii/iin

n (V IV 01 (L
T T T 01 01

01 Oj 01 01 01
(V tv IV [V rv

in LP in tn i/i

a u

-a

c
a

■O 4J

c n

3 >

O o

Q U

IvflDinOUJOlOl — IDIfi

inriOT3*inriinrvm

i-tOOiivOirvrin'^(rivoi'-^j*iO(^oifnyo
oioa3nioaini/ifvu3a3iv.Ha)tvtv«iTioiiD

or inoj to
in in 1/1 01 o

maJODODIvlvivtDIvIv

IVlVtvIVIVlVrvIVlViV

intnini/iintni/iinintn

tvfviDiDtotDini/iryuDUJi/ii/iinini/is'T'n

tvivIvlvIv(vN|vIvlvfv^^tvlvIvr^Iv|v|viv

ifitnini/itnjiini/itninintnintni/iuiinuii/ii/i

01 01 01 01 o
IV N tv rv CD

tn 1/1 ui i/i tn

oitDOi^mxj'ODo-'iotuotDiflivoiiDtvi'incDoiinmoiaiDioooiiviD'-'tflguii/i
njrt^,^rtrt_itninM3"3"a)oooi!7imfnooiuitoiv!vin3'ai>-<'-<ivoioiiDinoDoioi

oiiiioioia3a)SD(BaD(D(Dniivivivu:uiuDtoifli/iivivifliDiDUJi/iini/irt^n<^mt^oo

I-vIvlv!v[v|vtvtv|vIvlvIviv|vrvtv[vN[vIvlvlvlv|vivIvlvtvlvlV|v|vivtvlvrvai(D

i/ii/iinininuiini/itntninuii/iininifli/iinini/ii/ii/itni/ii/iinini/itnuii/ii/ioii/iintnintn

oioiffliva3T«-'OiODtvnoiinoj'-*TrifltDoiT(DtDmmo©CD
fumcirt — oinntviv3"eDNOOjruoiaiinj"Oi"(DiiiiJioiu)iD

O'-'iDfnmroiOirin
inniv'*— <tDuiffi — »"

ivQiaDtDtDrvpvNtvivivrvtviviouai/iini/ii/iintnioiDininuii/i
rvtvrvtvfv(vivfvrvrvrvivivr^rvf»vivtvtvivfv[v|viv(v|viviv
tni/iini/iini/n/iini/)ifiinini/iintni/ii/ii/ii/ii/itninininininini/i

rrmmnojoioioo
tviv^rvrvivtvfvgaai
i/iinini/iininininuiin

u a i/i

t. E I I a ui

o o m n I I . .

> U I I tU 01 I I — r< rt

■-< u d a I I rn 01 OJ I

3 □ <c d a c I I

o a <i<L <z <L <L m

a in 01 OJ

ami I lauiuiacnauiainauii auiatn
* ■ • I I I I I t 1 (•-•«-< ■

a in a 01

UJ I t I u UI UI UJ tn u Ui UJ UJ 1-^ UI • b^ UI I— i UJ i— t ui i-j UI

i ^ ^ ^ t ^ ■t|iiftlt<-«<-<tlii till

(\ji I i^^HrHoioii^i^J'J'inini i'-».~toioin^oifTi— '^oiru
icjuuiiii iiiiiiiaaiitiiiiiiiii
mmmo]uuuuuuuuuuuuuaaaaauuiuu.u.b.iu

a 01
I

0. u.
m UI u

I I I I I I I 1 I t UJ UJ

aaujujuu.u.b.iuu.(nin

a ID

c u

1 B

in lu

10 id

u J

Ui

IL.

Q U

a

a n

>~ 3
- OJ

198

APPENDIX C
Chemical and Physical Data from the Swamp Gulch Wetland Sediments

199

ro
4J
(0
'O

>i

4J

m

•H

V

o

c
e

•1-4

(U

a.

a

E

1

a
a

g

1
a

E

a

ve

a

3

t
a

'-i

^ s.

g

?

E

a

b.

a

-1

V

•H

CR

u

E

m

-H

a

V

u

a

X

E

H

xu

u

K

H3

u a
in tt
mi*
i. n

au.\*

u 6 u
ui 6 ^

□)

-0

u tro

c
rtJ

<t i rH

rH

4J

V. O

5

u o-o

It) BO

U E <-<

X

iH

o

X

o

n-3

W D

Oi

a>j E

5

X
H X

O. tD

w

U *-'

ax e

^inoNcinoui

mON»H»HlU«-ilD

oooooooo
oooooooo

{nODloaiROItDOD

cnooiD^oDivin
(urn:fnj*iv±'in

oooooooo
(urirH(uo*'ntH
ajivr-.t-iiuin(nN
airHRoinjaifuai

oooooooo
njoi<j<ajiDNin

m (u ^ nj .H <-i

mmEmP{nSS2PP°°PO°°°0000O0(nOOO000O0O00

gooooooooooooorH^

O O O O O O O O rH O rH O O o • •

ooooooooaim,-(ONtD

rH in m
en m (n

o o
m ID

o
o

o o o o o o o

u

•H

01

u

•H J in

(/) J tn J

rH

m

IN O 01 PI
lU rH

01

in

m 01 (\j o

o

rH

O rH OIN

rH J* in r-

oomoivaiom

rHrHd^J'ODJ'inO

iuNa)iDiniu:*^(vj

n in rH rH

!lJ!P!fi52S9t:'^lQ'" rHtumoi^ innm^Ninj^oo inoiiurooNino
i* N OI rH j< m *• o 01 o rH lu N -0101 m 01 in o in 01 ID ^ m (u in ui oi o ** rH o

pioorHoiuiDnruiUrH ininnino

mrHrHOIOrirHOm

OI

inrHDlNOOrHO

OOOOOODI/IOJ

rHrHOJUJJ'UlrHrn
d m rH tH rH V

oj'^'^o^oiNoirHooo cui^rHOihj moiaiNoiNinoru inmiD3^iDiD*'iD

^■f:t<fntnmmi':r:r:t' mm aiin:t< j'mj'ma^oimajri rnd^pima^ina^ui

?Ir:I!P'^31!yto'^'o<-«njiDrH:t< NminmiDinj'ini'rHrHajr-.inNd^iDftjinmo
aiix)airH:t<inuiin:t<tnm3^oJ' lUrHrHj'ivajooiJ'aitDrurvOiorj'iDinrHrHn

" D- -

OOrHriJaim(Tl**00>-HrHaiai :t<rHOOOO>-frHa(OOrHrHrHnJnjr)OrH(UO

o^'intDmi^njiDoiDrHnjiDrHrf'rHmointDiDiDj'inorHrHiDNinr-.d^oruTOO

OIVJrH3<rH3^(TilDOi'0iaii^O:f*'d^OrHj^|v{DO01Ot\JUlJ'N01O3^OinrHO

OOOrH(UaimnOOOrHrHrUnjr)OOOOOOrHrHOOOrHrH,-H(\jaiOOrHO

o

X

bJ

-l-l

H

■o

(n

c

(U

9*

?

<

bJ

a a: CO a QDa X X X a

i'iii'ii<r<na:aaaiaaijxxxx xxx

•I^ IP •? T 1"? "^ '? ' I I I I I I I I <r <r a: cc QQ p <r <E a aaaao

iJl-iiii' ituiunjiumnjiuaiiTii i i i i i i i i i ii i i i

SSSSSSSS' I I I I I I I I n rn n in mm n ftjiu fu ai lu lu nj lu

S S S S S S <i^ ^"^ <i: <'^ <i: <'^ <i <E « <n a; I i i t i i i i i i i i i i i

<[<t<E<E<r<r<r<i:<E<i:<c<E<E<r<E<t<r<r<t<r<t<c<L<raiiDaiootaoac]am

UlUU b)
KOCOiOi

a a o a
uuu u
I I I I
m mm **
III)
en CD m m

rHmmj'iniDNiDoiOrHnjmi^tniDNaioiOrHfumJ'iniDNODcnorHium^'intD

rH rH rH rH rH rH rH rH rH rH OJ lU lU (U OI nj lU rU lU OI m m m m m m m

200

4J

c
o
o

I
o

c
a

E

a
0. a

E

■ a
z a

c a

E

a
u a

E

10 a
u D.

o a
in E

E
0.

u. a

U E

a V

u a

X E

Ul 0::

u ,s«

•H

in .\*

Ul .\*

z

<E CO

K <r
a U.N'

in
a

£ E
U E U
Ul E ^

■^ CB

u tro
u no

« E rt

u tro
u uo
u E •H

X

a

X
H

a. 3
u o
a J E

X
H X

a. (3
iti >-•
a X E

Ul
H

in

o

z

o o lu u) u> o n
•-< rH :r m r^ «-< 10

V V V

o o o o o o o

N N m (D in en i"
lu lu m in in m m

ooooooo

h- O N (U m :f 01

rn m in ni (D (U N

PI m VMD

o o o o o o o

m oj ^^ o o ru o
ID o 01 m 1^ o J*

rH ri t-i lU 01

ooooooo
(^ r- in «H o in m
m 01 :f h. cH i-t m
(n «-• 01 in cH

01 rinmoi

lUNUllnN

r-l

ooooo

10 O N Iflr4

fi rl PI in PI

ooooo
pim i-N S"
PI IN frtm

01 N to NO

8?S8S

10 <H :r> <-i o
10 in :}< IflPI

ooooo
lu o o r4 in
o (urv (flo
lu ro m •-< ri

oooQooooooooooooooooooaooooooogoLnoooQ
*<3*piincHa)Pi pioin o oiiom inms'oooiooiDNPiruturuinj'iB
<H d< 10 :t< 01 m CD o in m in nn nn oi oi in oi .-i a* n m iu a* iu Si .^

01 cH OJ OJ «-l OI O <H 01 T rH OD PI (U<-iAJIU

. 01 rt OJ

CD (0 m
«-« o in

m o

O 01

*• N N

in 10 3*

in pi N o ^

«H PI rt rt 01

o oo

o o

-1

J .-1 u ^

•H ••-* -H W

uiinui tn

UlOl

PI 01 PI «-<

r-l r^ PI <-<

to f.

o 01 m ^
in in 3^ N

in r^

N 01 01 (D
PI 01 .H r^

N in

PI IV.

o O O .H

a< oi m OD

O O PI
c^ PI PI

o o o ^o

PI *• PI PI 01

N Pio i* .H

to

»H oooo

J .J J J
Ul Ul 01 men

01 PI 01 ID h>
w OJ »-< «-l

01 OJIO OIO

Ul 10 in uioj

OJ Ul Ul rt tv
01 <-iOJ PIIO

O OUINUI

Ul »H J* PI.H

01 o r Ul CD m .H

rH <-< <-* 01 <^

01UI0l:t1<HOaiO0J0J<HOOO<Ha}<-<U]OPI0J:rf-tOP1>-<P)O

owJ*Olul^-ulOolJ<ulOoululorvp^ooJ3^PlP^ul:t*r^p^l^

w ID PI in «H

OJ

01 i< CD .H
<-< Ul U1UI

00 n PI <-4

O 01 10 0)

PI01r^OPIr^c^OP10JPlOOOPIPIPIOOln rH O 01 a<rH:fOOrtrHcHOPIrtPlftl

m m

01 OJ .-i

01 O O 01 O o »•

T T CD at tnm s*

rH OJ OJ tH

O O 0} O OJ

Ul OJOl OJPI

OPioj»-<rvOJOPiu)«Hh. oorNNuiui

uia*uiaipipi:r*'ai3'pi oia^pia'Pi

uioaimu)ai<-<ojoaiAiu)>-ia]T-)P)in

Pli*PlOJOJ:fJ'*<3'»<Ull'3<PlUlUlP1

muioioojo:roj«HOja)Piroi3<o<HOJuit-i:r<OJO}<uiopiOj u)<H0JU)>-4r-iPi

D1Uim:f<0JPlO0J010JO0J^U101PlUI0J«-lU)O0J(D:t<NPl.-lUI :t<oioim«-)nj:t*

• • • • ^

Oi-<<HO«-40<-<<-IO<-«PIPlP1PIU100r<OOTHi-l|UPlPIO>-<rH OOrH<-t PI O O

;:!ffiH!9!fi992!OrJ'D'D'^*''-<oo'^oi'''^^no3<oo oi oi o ui v^ ui m o <-<

(001U]OaiOP10C>01P)OOJ:1^0JOPlU10<HtOOr4a):»<OP10JU10:t<01^(DO(U

o o rt oo o o »H o o OJ Pi PI PI Ul o o o oo o ^oi OJ pio o •-• »-i o do d '< d o

XXX

uiki uj a: o; <c
o: o: ck: XX t i i

DDOaCOJOJOJXX

XXX

<c a: <c

I I I X X XX

oaaa<i:ojoj(uxx iiixxxx xx xxx xx

uuuiii I I i<ta:anpaouiuiui<n<r<r<tc3aQc:<tao<to:<caQa:<E

' • I •;<'H '^.^'^ I I I I I I I • • • I I I t I I I I l l l I l l i i i i
T7^.'..', .'. ' ''-•'-'■-''-•'-''^•-"-•'-t'HOJOiojojojoJoiPiPiPipia^a'it^a^j'uiui

I I I UUUUU I I I I I I I I I I I I I I I I I I I I I I I I I I I I

BQCQiamKimmmuuuuuuuuuuuuuuiJuuuuiJuuuuuuuu

N CD 01 O

PI PI Pi i<,

iojpi:ruiu]|va)oiOriOJPi:ru)ioivcDoio<-<iyp)^uiiarv(Doi
> :t^ :t'if 3* i^ 3* i^ 3*\nutu\in\nunn\ntnmuiunSunauimmtDUi

N (D 01 O r4 lU

N N N

201

SI

0. a

z

a

E a

>/

E

a

a

m
u

E

a
a

o

a

E

k.

1

a

m^ ° <H o (uh; m (D:r<ui M^ o ro o m aj axninmt^ mm m In
<^^ r^r^r^^^lnm men (n.^m m

?

m

r-(

a

•Vs

u

n

X

E

H

X u

WK

H

3

U .S"

in .\'

in .NT

z

<r «

a: <t

O U.N"

o
x: E

U E U
U E ^

V. a
u cro

w |0

<i: E rH

or^mom3^inma^uimNmtnpi^rNtu^taioj^oinu):f<rn:i*oiDLnoinooI-5
oooooommor-iminmmo.HininwootnomorHo.HornrHcJrnooa'

4J

c
o
o

u cro

U ffiO
U E <H

X

a

X

Qh 3
U) O

a J E

X
H X

a. tn

W i-i

n X E

i^inuj fnoimminmmmj'init'tvoajininmu)
a^ *• in in in m 3^ j< m IT m 3»a< :r :t* a^ in 3^ 1^ ui (n

tH(Ti<H<Hnmma^mwmu)iD

m mmmSm!H9r;!!}^SK!'^ rvOr-faj3^(n0)otDrHm*<mrHmmmm(nto
- p.. ^'^^*I"!'^^""^'^''"u mcn(naj:fcnom:MDmj^^iDmiDinoSjio

00<H00<H00rH00<H

^'oo^mmoooorHmoovHmmmmfn

3^minomoioru)oa^tDornminrtO(n(n(ntnoon3^u)oo^HiDinmin»<(u
o.HrHooooooooooooomooommooooomoooainJmnJt»i

U

•rt

Ul

TJ

H
*-*

C

cn

<a

a

Q.

a

<

z

XXX

www a: a: a:

K K K X X X I I I

PQQUuui I I I I ic:a:<t

inininuiiDU)! i i i i i ^^^
' I • I I ipnoaaai i i

w w u w w w

^, _„ „ K ft: a: K Ik; K

XXX X X X X X a X X XO ODD o

<t<c<nnaa<ra:<c<cu<E<i:<ruuuuu

I I I I I I I I I I I I I I I I I I I

|^aJmaJmmlnf^(nmm(n<^^.^r^rt^^wr^

1 I I I I I I I I I I I I I I I I I I I

uuuuuuuuuuuupQaapaQQaaaapaaawwwuiwiiiww
lI!?!!t{'IS[:^^S'SsJ!y'I!3!'''iS'^°""0'-'nJtnrinu)NiDtnortmfn:j'inu)tvm

202

4J

c
o
o

I
u

X

•ft

c

(U

a

<

a.

E

a
a

m
z

E
O.

a

g

a

a

>:

E

a
a

E

a

a

u.

i.

a

u

a

E

tn

o
u

X

CI

E

X Ul

U .\'

01 it

in ."f

<E in

to

K

Oi <t

au.\«

m

n

C E

u

E I)

w

e N.

>s CD

u

cro

Ul

eo

<c

E <-i

•^ rn

u

tro

u

on

u

E rt

X

a

X

H

a.

3

Ul o

a J E

X

H X

a. tn

w

»-H

a X E

Ul
H

01

otaioaimt^aDmio
tur^njiupifncumfti

ooooooooo
ino(urninm>H<H(u
mmmmtuainmrn

ooooooooo
moDNi^aiNruiuo
o(uui<-iiu<-4oaim
r^(T^J'mullflU)lncn

ooooooooo
01 m in o OD lu 10 oi ^
mcniDftiod^^Dai'
(u<-i(uinium(nnjin

ooooooooo
(uiDOinomtumN
iv[D<H|vaiuJin<-ioi

in fs. o N o

o

?a

(D N (0 U) r-l <-t

V V

O OOOO O

CD m N O J< 01

in 3< If J* 01 oj

O OO oo

oi n;r in m

10 luio ojin h. «-<

j-inininrH rt*

OOOO o

01 tooiinoi in
o PI OD N 01 m
a« 3<i<moj m

o o oo o

01 N 01 O O ID
N m N 01 OJ N
OJ tn ** ID ID

ooiniDoiooooooooiDoooooooooooinogopgoo'

moiii^moP^^ojom isoiai -'Wh^j'iDitjNOi^oirta^mfeS*
mini inmoaoojo] m «h o oj i* o n id n oioioj o o :r m m j*

nj OJOIOJ ln«^r^±'OJ•-^ OJ.^^^a<^

0.3

<0.1

<0.1

0.19
0.H9
0,57
<0.1

am :f

m OJ ^^ o

.nm in

OJ *• inm

rt in

in <-t m • n

. . . o •

m r4 r^ V O

O
. V

o h-
mm

01
10

in

O O O OOOO

— 1 rj ,-1

U U U U J J

oiinuiinuiuuioi

ml^rHOlrHODOJ:**

m m :j< m 01 oi OJ

t^minoiinffiNin
inmina'inj'uioi

OO:J*0Ji^m<HrH
rH <H .H 01 01 01 h-

oommiDit'J'^'n
oi'OJiniDomj'N

nilDlDin:riD:f±'«-<

01

o oo 01 o

01 >JU)

01 N «H

:»< mm
01 mm

N O 10

ID in tn

J
u

01 01

01 «H

m

^^ m
01 in

o to

:t< 01 m I
m N m ■

I 01 o
:t< m m 01 ID

01

m

01

m

o

01«^m*'<-^«^^^01ln<-•'^ r^J'mO0JcH010l<HCD01>-<010J|va3N

m 01 3* OJ OJ o cH in cH rH (D i^ oi in o d^ <-h m cd id m n ^ loio lo r^ rn

(D o o o
oio m o

o t^ o

04 90 <-i

mvHooorH.Hrtmmm .^oo o in oi oi J* m ^ oi o oo oi »h m r^^ oi oi (UN «h

rH m

ooooooooin

aD.-imiDr^:foiNin
mmO10J'^«^«-^r^

N OlD o in

01

ID «H OD O rH cH

inmmmmiooiommm iD*'moiioooiomo<-<oiivt^>HO)ai«-<oii^i-t

oiin w

3^-i':n':f:t'i':t<i<\nT :t<3^d^*'mj'minininiDiDioiDmmminojmojmmJ'

in»-<<-ioiincHCDUiiDJ'0]0]<-toii^oiiD<HO«HiniDr^oj

«^lD0101lDOlDmtD(Vl:*<01D01^-a^01^(Dr^c^tsOJln

m N 01 O 3* .H ID
CD OJ OJ <H ^ 10 01

(audio

ooo«H«HOioimmoooo«-tcH>HOioioimoooo

rH o .-• 01 01 o m

b m o

oui«H<HOiin<-<iDinoj<om»-<ojr^oiiD<HOotnoi^ooioNoiooa'd^uJ:»'0

0<-^lD01tDlDOtDm00100lD01^(DOla^(DO«^OOJOOJOalnJ«-^0:fa^3^*'0

00000^010imOOOOO>H<H^0101(VIOOOOO>HOOrtOJOI\JOJO(UO

Ul Ul Ul Ul lU Ul

xxxxooDoaoxxxx
<r<i:<r:<Euuuuuu<i:<Ea:<E

X X X X

a; <r <i: <r Qp

<r c flc <E u u u u uu a: <E a: <E a c <i: <i: (3 (3

llllllllllllll 1111110.0.0.

01(U(U(UnJ0J0J0JOlmm«-^«^r^.^^^r^^^^<^0J(\Jmmmmu^U^W
I I I I I I I I I I I I I I I I I I I I I I I I I I UIUIUJ

Ul Ul u) Ul Ul Ul Ul Ul uiujuiu,u.u.u.u.u.u.u.u.u.u.u.u.u.u.(n(noi

mo.-iojma^iniofvaioit-immj'iniDi-^mgiojHOimauiniDi^mgio ^ ^
o<-)«-<'-(<-)rir4fHr-ir-i<-tojiviiuiu(u lu (u nj(ummmmmmmmmi<i:}< mp

01

*-*
a, a

H 1

m
. 0. >• I
I u <£ <i:
I Ul oi <E
I m (D a

, X

PX «
I C P I
01 I: I rH

I m OJ I

<C li I u

a: <ciin PQ

•^(yffi

S

203

c
o
o

I

o

X

••-I
V

c
a

E

a
a, a

E

10 a
2 a

E

a a.
c a

E

a
>£ a

u a

a*

^

o

n

tf)

E

P

a

u.

a

»-i

v

.-4

n

U

E

m

-~-i

u

\

u

n

X

E

H

X U)

UJBC

H

3

U .>f

•H

in

.\"

(n

.\'

2

<c n

(i) 0^

K

<r

a

u

m
n

.\'

x:

t-

u

p

ij

u

E

\

■^-..

rn

1.)

cro

lU

0) o

<L

E

r-i

X.

ni

u

cro

u

0) o

u

E

^i

X

a

X

H

p.

3

Ul

n

u

..J

B

X

H

X

a.

I'l

u

Ul

X

E

Ul

o

2

(\J

r-

o
ai

o o
PI ai

o

in
m
m

o o
(n o
in rv

o

o
m

o o

m o
oi m

o
m
m

o o

ID in
o tn

(UlD

**oino:t'Ooro

oiOfucHNfninruo

J' ^ ai >-i in o

ni oi

o

cH (T)

ni

• on

•

o •

o

V o

o o

H
(11

u u
in t/i

m

m
ai

m OI
in in

in
to

in
o

o m

ID m

ooiooin^tooio
(nNruommaiino

oa^or\loo<^o^

ai
m

o tn

(U V

oiincHaja^ommm

m (u

r

m

:t<

:P

:f

r

m

ai :r>

ai aj «H OJ >^
ai 111 ID ai in

r^mv^m^^^o>-lO

^ooiDui'-Mn^o
m 0) m ra in in r^ in o

o aj o .-1

o o o o o

Ul

X

»-*

a-

<r

1 X

3:

XXX

t-t

in n <r n <r

<r <r <r

- 1 1 1

1

1 1 1

>

.^ ru m a^

1 1 1 1

1

m ni --1
1 1 1

<r
a-:

u L) ij 1 J ri n M u.

Ui

o om o

a^tomoNCOrtaij'
tn I to t^ m 0) .H rt .H
aaaoannaa

204

c

N

00000(
(D N U) O l/ll

^<^r^al«^m»'(li>^

ppomri-opoooo^ggogogoogogggoggg

1^ (u in iniv m

aj<-«a<cniflcnivm»-irtQ

c
en

vvvvvvvv

oooooooooooooooooooooooooooooooooooo

t^<^^^«^^-tr^<^<^<^«^<-^<^«^.^<-^«^«^«^^^c^c-(c^r^r^«^«^«-l»^«^r^^^r^<^•-^•^•^

VVVVVVVVVVVVVVVVVVVVVVVVVVVVVVVVVVVV
VVVVVVVV

ommouitnu) ooinino minino «_.„

uiminmin3^iDO(no(nrH{TioiJioo3<inrHrt(uininm(oi0(\jujr.h-ruonju»

d^r^c^^^^^r^<-l(n^•<^«-^<^<^>-^^-lfl^•al(n»^<^r^^.lJlrU(UUJ^'JU1«^(Dd<U)(nNf^

o OD m to o
- - - - rH oi m N

oooooooooooooo:roooooj'ooui»H(nooOrttnoi>-<oOQ
iDma<rna^±'NrH3^(nvHinmh.u3.HOO(UfU03(uaia3v*<O(flUivi\i(DiUi<.-ttf)

I

(0

m

0)

E

a

X

3
U

u

u

oooooooo:roooo(nooo(u<Hr^(ucD0ioincDOh.ooooouiQi

~ " rt i< o ifl m ifl i'r> o I

rHtDmi/)i^(urvm:rs'mmm(Di^U30in:t*in(nNiJii/iin
in in • in in in in in

vvODvvvvv

or--inffi(nmtom

ru punin

4J
C
(U
B

•H

•o

0)

n

■n
c
m

+J

»

o
a

6
(0

I

u

c

(U

a

CQ

in

CD

<E

a 3
0) a
a J G

a a

U H

a X E

0)

in

ooooomoo
i-^aiOftjo(n<-i(n

<H .H •^ rH <-» nj

oooooooo
i^oDmffiinmoiD
rHiDmoi-iflininj^

x-l

UIOOOOOOO

CDininmiDomrv
<-i ni (n (T) (u m on <-i

OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOQQO
OOOOOOOOOOOOOOOOOrtOOOOOOftJr-OOOOOOOOOQ

iDcnminr^ujaiirininaiinmintnftjaimiDtDminmaimminomoinouia'rHm
^'inmtnnninooioinmajni^OcH.HUi.Hoa^njin.HinmiraiiDOD'HtoSiuia'B)
<H nj (u c-i (u (u ai cH rH ,H oj .^ fu nj m (u "H «Haj(uiu<Hai lu nj :»< ui lu r^ »^ w -n »-i

■Hinorvmoi-Hin

0<HfnrH(\JOtHO
V

*'rHajtn*'aiini^ifl<HftJiDrH:f t^miniou)inj<in:r>H<HnjrvinNJ'U3i\jmmo
minai.Hd^mtDLn^'infuj^o*' aj«-<rHj'N0Doin3^(uiflfur^tnoi'iflin^»-tci

p-

oo<Hajiumm3<oo^.^(um J^<^oooor^<^moortr^rt(\J(u^o.^(uo

oj<cniD(n:faiiDoiD<H ai to <H*'rH(noiniDiDa3:t<ino^»Hifli^tnt^:t*oruno
oru<H:f^:r(niDOj<(ncuj'0:t^a'**o^d^i^inomonju)a<Mnoi'Oin«HO

ooo<H(uaimmooorHrHmaionoooooo^<-<ooo.H^rtaj(uod«HO

XXX

<r <n <t n

I I I I

m mri m

I I I I

a: <! c a

<r <r a <r

apaaxxx d

I I I ia:<r<tcianaauxxxx

(1 m m ni I i i ^ ^ ^ ^

XXX

u ui u u
aaoa

iiiiu.u.u.l.ii--ii..iui(-ii-i;cak;i^ j;xj; uiulju

inmmnii i i i i i i i i<n<ca:a:aaa<t<ra:QDpaouuuu

I I I I OI (u iu (u oj III oj oj m I I I I I I I I I I I I I I I I I I I
0:0:0:0:1 I I I I I I I iinmmpimriniaiiunjaKutuajnjfnmina^
ao:o:o:o:o:a:o:o:<i:a:o:o:iiiiiiiiiiiiiiiiiii
a:o:o:o:o:o:a:o:o:a:o:<ta:oo:o:a:a:o:a:iaaiDQCQCQaJoaDQ mm m m

<-injiT)d^tnu}t^a](no<-iai(nnnu}r>-ai(nO'-«mrn3<(nu)Na}S)0'-*nim:>*tf)ffi
<H<H^TH.H<Hr4r4<-i<Hru(uninj(u(uiu(uiuiuininin(n(n(')j*t

I

205

c

N

ggogpooooooinoooooooooooooooooooooooo

c

XI

tn

0.

c

V V V V V V V

I «H tH rH
V V V V V

oooooooooooooooooooooooooooooooooooo

VVVVVVVVVVVVVVVVVVVVVVVVVVVVVVVVVVVV

V V V V V V V

o
o o

V V V V V

o o
in in o o m o

og^ p g in Ul g m o in m o o o o in in o in o o o in in o o m o oo

inpg(DN(n>HUJm(uin*<njna3min:faj^(UrtroinooinomiDinrtOoi^ui

luaJlDlp>^r^c^c^lnr^ruair^>-^allD^■c^<^rt3^<^^(U■^(Dr^>^^^^^(nalmln«-^>-^

O

I in o ■ in 01 m
ru fi in in m oi

CD rH in *• r-
<H ai ID cH rt

99Sr32990'^oo'noooooou)oooooo(uooo o o o o o o o
ininoin*'inninvinocnaii^mainjcDtn(n<-(OiD<HO(Dnjoruin:fmtij(nmrH

3^ <H

a

m

u

u

V V V V V V V

V V V V V

gggggggogooooooooooooooooooooooo
2SSP229Q900000000000000000000000
rKEr2f^?!°9*'Oo°"""0'Doooinoou)ooooh.:fOonjo
!2?l!fl'G&<!y3:'^''i'^*^'-'^'D''''^'''0"'i'ia).HOr-i^oajin(uaiinrt
(nfU(\jm(VJtn(D(v,Hinm.Ht^rva)mmoi(uajcn(najininoam^ainjj*h.

H2(099P299°°"i9oooooNooi^oooooooooouiooo
}D:ri<ip(UNi'ofnrvinNmiflCDj<o<Htnm3'morvin:r':fiflCDiD(ni^tfl(DJ*m
in<H i<«H<-<rucnNoa inininmNm «-<ai .-<(DiOrHrtN<HcHi^(n rv^n

oooo
oooo
o to in <-<
«-< 5 ^ 01
luiu tnm

tn in in in in in in

V V V V V V V

O O O (\J O O ID

nj o i< w rH o 1"
01 m «H ru <-t

(U

• in in in

h> V V N »H

rH o o o in
m o m (u o

Ol <H V

<s

I-

£

•M

4J

a 3

c

• o

O "

a J

o

x;

*J £

ao»

1

ss

u

X

•H

c

OJ

■tJ

<p

•H

9«

(n

S

<

.

d o o 01 o o o

o »H m ID in i-v in

J* 01 m *• :t< rH

m

O ftJO o o o o

o 01 »H m <H m n

<H m lu m in (u

o o o in in in in

3^ 01 in N Ol (T) Ol

m 01 m OJ OJ ■

o o o o r^

in 01 IV m in

^ ni T <-<

ID O O O O
01 «^ d^ ID h.
J< 01 <H Ul m

in o o o o

01 ID ai OJ •-<
OJ OD in OJ

OJ

01 ■ ID m in

<H «-< 01 •-< O

ffiH39'2!M9S'^d'^S'^*''"^0'-<'^"i'-<^ojo:j^ino(noi incnojiort^m
oiiDOii^oJinoojoiojooja^uioimiDoj.HiDooiai*' r-- m ^m :t<oioja)rtOi»'

fL

o<-i<HO.-<o<-<cHOf-immm(ninoo»HOo<-<r-iojmnOf-<«H o o <-i ^ tn o o

.'^SKJ9i2°9^0'-''DCDm3^<HOOrHOin'H3'(noj<ooojojoiDrHioujOtH
iDoiiDoaio(noooimoojj'oiomiDOrHiDo»HiDJ<o(naiino:^oi5'ffiofu

oo«HOOoo<-(OonjmnminoooooOrtOioinoort^oooortOo

XXX
Ul M U ' <C <C <[

K K OS X X I I 1

!i;SE2i'ii!l]3!'^""^'D'no<-toj(nd^iniDtviDoio<HOj(n*<iniDtv(DoiOTHiu
(nmfn3"i'*'*'3"»'*'9»*<*"inuiininininininininiDU)iou3U)iDiflU)iDiDNi>vtN.

206

c

IM

oooooooooooooooooooooooooopoopoooooo
(nal<^u)m<^m(UN3^J'ID(DO:rlfl<-^;3-•^:t*wullnJ'm(u«^«^u)ln^^nonlpo

c
in

oooooooooooooooooooooooooooooooooooo

VVVVVVVVVV VV VVVVVVVVVVVVVVVVVVVVVV

a.

2

o

o o

IX)

oj o o o in o in

O O O U O Ul

inin ino ooooinooooinooo ooinuiino nio-.^-.-..
ini^otninin<-<in>-<iD(uoouiiOoHRt0Oinor-no«-i(ncDin'-<oo<HOio«Hin
<-HH(nh-(Uh.mNinu)njinN<Hnm<H»-4to«-t(D«-(«-i(\jm«H«HOi(umru<H«HivrHCU

oooooomoooooooooooooooooooooo o o o o o o o
incDc-<±'njm:j*miDCDiD(na)mtDnjh-aj(nr-.'rHi'0(ur--inrnajNiD(DiDLnh.ptD
>-<Nntfli'<H mm njin<Hin*'j''H3^h-in«Hnjcni'i<-<«H<H«-<(U0Qfn«Hi\J oiru

D

0)

3

u

oooooooooooooooooooooooooooooooooooo
oooooooooooooooooooooooooooooooooooo
oom±'U)tDOinooocnoin^<±'00ooi^<-iOfUrHt^oi^onj(uaia)(nni/)
ajoonjinnj±'n(n(nj'Nruaia^*'oaj(D(0(U{n*'tD(ntDiD(nnjt-~r^i<u3O^N
naim(\j(\j<HoaitDininmininmrirnin en tt* ai^'iDNa^mJ'njONinmtnftjNin
rn<-<m<H(\i •l^^^>^ >H n

ooooo.HOooooooooooooooooooooooooogoog
p^lntooi'aJnln(n^^i^fu^^lD3^NO)lsaJu^(D(nlno:t<m^(n*'o<^al*'mlilcp
— -• a<CDinmin<-iino

I ai •

ru '

i^N:r'.H(Dinorn±'inLncn«n(n(um(noainj«H
(u <H m cH «H <H nj

m

u
u

•a
u

I)
n

DQ

VI

<t

oooooooooooooooooooooooooooooooooooo
ooooooooooooooooooooootoooooomoo o o o o o
'Hmu)ooNj'intD(naiLDiniD3<minoDomincnncDi^tDininm*<mo(nj'Ni^
•-ttna^^'iDaiiniDh-oinNoaiDaDa'moiina^o.-iCDrtinaicnaiNi^mmininniD
m^cHOid^^ mincHrum>H*'j'Ln>-irf±'mma< (vi(nmaj<H m^'iumrurn.-i

CD

c

sz

■ ■■

■p

*

ft 3

4J

0)

c

a J E

o

-M £

a CD

«

0) H

CM

a X E

u

X

•rt

-O

0)

C

■p

I

in

On

rt!

Q

CD .-H<-H<-(0»Htuo«Hajouiai

en ajiDma^iD(U*'iDfurninru

0- - p- c-

>H 0<H00rS00«H00<H

t^OTHruj<mtnoiDwaj:r'(UrHfummaj(nio
(umcnairfinoriJ'iDnjj'rHuiajiDcnoaju)

J*oor^nJal00oor^nJOo^nJaJ^^^^m

moimO'-irtOo.HOO.HOoinm'-toortajooinoiDrH.-ioai.Hinmm'Hcn
a^ruinoftJtnoifiDOJ'mominin^ocninminooinj'tnoOrHUJinoQina'ai

OcHrHooooooooooooonoooojaiooooonjoooiuiutuiuri

rc ^ 3^

www <t <r a:

OS K K X a: X I i i
Daa<i:<ra:ajajaixxx

QQQUUU

II'

in in

w u w u Ul u

K OiOiOiOiOi

XXX X X X X X D X X X O D D O □

c<cpa<i:<E<ca:<Eu<n<r<tuuuuu
I I I I I I I I I 1 1 I I I I I I I

aaouuui i i i i i <t a: <r <r c <c p a <i: <e <c a: <e u <n <r <t uu u uu

llllll<H<H<H,HrH^||| lllllllllllllllllll

Lnininuiioini i i i i l<^c^<^<^<^fufUlu^unJnntnm(nln«^<-(^^<^^^^^«^r^
iiiiiiaoaaaaiiiii lllllllllllllllllll
uuuuuuuuuuuuaapaPQCJaaaaaaaaawwwwwwmw

fna^iniDrvcDcno<-irun3<inuih-eDtnorH(u(n3^iniDiviDino<-<aitn»'in m n aa
NNNNNrvi^oDmmflotDiDCDDDffimcntnaicnincnintninoiooooooooo

207

c

N

mmS2mSS2SPS92o°o°°°oooooo o o UIOOOOOOOO
ISm^31mftiS2;S'^3i°''"^"'^ ^° Rinnan r :f r^^mt-- t--ai mm Uidi mm

c
in

JO

in

JO

a.

c

vvvvvvvvv

V V V V V

292S95°°°°°°°°°ooooooooooooooooooooo

vvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvv

vvvvvvvvv
m}SSmi9m2S92SU'S°2'"90°00'''"1'" UlOOOlhO uS

mtflr]iDUJ(Dd^iDOj^or>.ONOu)a)o:j<ocD:i«ONJ«*<h-oNCD(uminrainin

V V V V V V

O O O O 1

ui o p o in o uS

airfioa^mmiflcno
tnj'aimnjpniunj.f

cn rH <H m in m
■-) in (U vH in <-i

0000000000000(

in*'ininujainj<H<HOD(\Jmini

o g O O O O O O O O O OO O O O O O O <^ 3*

(nmroinoiJ'ifoouJotnNiniDrtDoiiid'tD oo
ajinrumrv3^tnaim>H:f(umLnLniDi£io<H

■H fn oj in .

J* ID

X

3

u

u
u

■a
u

0)

c

V V V V

<-i *H <H cH
V V

<H cH rH «-H <
V V V V

<r

s:

•»

+J

a 3

4J

1) Q

C

a J E

Or

£

U

■p x;

ats

"*:■

H

axe

299°9°°°°oooooooooooooooooooooooo

inmoN<H(n*'m(uioftimt^oinaiinh-r-:t^a^<HmaimairHuiDDU)0(u:f

S22922°°9°ooooooooooooo>hooooooooo
fnmoimJ'ininmtutnajmininintD^ini^inrfajit^rHrH^ounDa^Nnjm

tnzr'QQ.Hj'ffiinaiin

V<HrHajrH<Hv>HrH

ooo<-Hin<Hininin

in:*<d^momooo
::t* ±» V V V V

OOOOOOOOO

i^tnrvj^rHtDotnm

OOOOOOOOO

inioinaj±'inajinj<
ln>^a)*'r^c^^^^a)

«-< <H rH ai cH <H

inoooooooo

CDlnlOa<<^^^J^>^al

o o
o o

W) ^ (U
- N ID
3^ (D

• m aj cH >H in

N <H <-t rH to V

ID o <H(n J* 01
in in N ^ h- N

o oo o o

ID IV m r^ in m
nj <-4 lu <-i lu m

o oo o aj o
in IV m 01 in m
<H ai >H rt :t*

in o o o o o

03 01 N O O ID

m rum m tv lu

|v

01 N

in

22222222292°9°°ooooooooooooooooooooo
°9999999°ooooooooooooooooooooooooooo
SHmmmmS5^!Sfiftlf!aKiyrffi'^S"°^orln(uotnm3^o^SmSSS
n^-i^S3KSmf^. ^H!r^9"^'^'^"""""^'^'^"3'DaDOioinjoivmma)iDotvlr

o>^rvlnallD3^(lJlD

aJn<^<^o<^<^<-^o

O O

• (U :t< O • rH

:j< . . . ru

■-I in in fu 01 <-»

Cvl

15.';!S!}l!{!;:Ji5!fllS^'"'"'^"J'^'n"'^0'-<iDioivnj mivmoa^rHtn iduhd
'"I i^tn m to o in miD (u 3^ oiDftj IV moi :j*a3 ^ rH N m in mrum^s^infu t^mT

ooOrt.rHnjajmTioooo<HrHrt(iiaiaiiTioooo .-iorHaJajorn'^dpio

25r;3s^!iiHJ^'B!il9^c)'n"~'"J'^'"'D>-<ooinoivoajoivajoodHa<uj3<o
o^uioiaiiDoiD(noajooiDajiviDaj*'ODO^oaioaioajaj^oj'a<3'J<o

ooooo.Hajajmooooo<H<H<HaiaiajooooowoO"HaJoaJaioaJc>

•o

4->

c

•H

(U

m

i<01<

a

,

<

o

z

XXX

<i: <r <r
I I I
aj aj aj
I I I

Ul Ul 1x1

m o tH

O rH rH

Ul lU U bt u U)
Q:: D:: 0^ de: 0:: Oi

xoaoaaoxxxx

<t a

X X X x H I a

<r <i: <r <E a a mi

I I I I I ia.a,a.a.>iaj

ni ni ni ni n, o. 1 ^ ' ' ' T T T T I T 0. O, O. O. > V K

aiiufuaiajajini'irHrHcH^rHrHrH.HrHajairnrnfncnujujuiuitta: i

I I I ) I I I I I I I I I I I I I I I I I I I 111 1.1 1.1 I.I (V /r /I
bl Ul U) bj

X

<E Q
h 1

m oj

h I

iuiuiiii.-»r-ir-irtrHrHcHcHrfruajnmnribibJWbi<i:<r i noj i
.', ,'i .'.,'. .' .' I .' I I I I " I I I I I I bj bl bJ bl ft; <r a h i u
bibibJwii,ii,b,u.ii,b,b,u,b,b,b,iv,b.b.b.inininint3o:<r<i:paS

aima^intDtvajoi^iaina^iniflivaaffio.HaimJ'intDivoDoio

rt rH rt rH ^ .H rt rt m ai ai ai aiaj ai ai ai m m m m m m tn m m m ;r uioai

rH»HrHrHrHtHrHrHrHrHrHrHrHrHrHrHrtrHrHrHrH>-l,HrHrHtHrH»HlOrH(Uf'J

O

□ Q aaa

208

N

WOOOOOOOO

i* m >H >-i 3^ >H ifl tn oj

ooooooooo

1-'

«H i~l i-l c-l «H v~\ <H w^ «~l

in

vvvvvvvvv

ja
in

V V V

n.

tn
o in ui o o in o r--
j-mtDLnmh-niuim

v~i .- t v-l (U n <H <H *H J*

i

.1 ^ in

c:

^lOOOOOOOO

m <• 1 ai ai ±' PI m on ^

3^ ai >-i .-1 J' <H j< ru 01

3<

V V V

09
li.

OOOOOOOOO
OOOOOOOOO

t-- m a- N m id in m
a' ui m > I ±' a-i CD m in
m in in lu n ;f m o o)

1 - O O :J' O O O O O
O r 3^ .^ m J< ID <H O

m 0) ^1 J-' >-i fti in on

|j
u

•o
u

ai
m

10

nn

O

■ri ID

in

0) m

O

m ID

V

1^

o o

U)

ID 3^
(VI l\l

O

m
m

o r-
N in
m

c
o
u

CM

u

in

D

a 3
0) □
a J e

x:

+■> j:
a til

01 'I

n X E

•D

01

C

.H

0)

1/)

Oi

a

kC

a

2

o o o

% -1 J' V

OOOOOOOOO

i/i o o o o o o o o

iv ^ ^1 10 ftj ^, 01 nj r-

ri ID ai m m ID vH nj r-^

n m m ,-h in ru j^ j- ^

o

if) ID •

■ in
o rt ai

ai J" ni r-i aj ai <-! ai -H
ai J* >-i <H (u aj in ai ID

^ in w m <^ >H o >-^ o

.-1 O O ID in >-) ID cH o

10 ID m m in in .h ui o

o ai o cH o o o o o

a:
<r:

I :i: XXX

in o <i ci a: <E <r

- I I I I I I

<-i ai ni r .-i ri at

I I I I I I I

u u t.) LJ Q ri ui

i/i

»~t
<r

X H

<r:

I >•
th <r
I oc

li, ID

o o (1 o

-J' ID in o 1^ 01 ^ tti J'

in I m iv m on <-< ^ >^

a a a Q a Q n Q a

209

APPENDIX D

Precision and Accuracy of Vegetation Data, and Plant Species
Observed at the Swamp Gulch Wetland

210

Appendix D-1. Accuracy and precision (at 907, con-fidence level)
of vegetation data.

El ement Accuracy l PrecisionZ

Al 44.5+ 6.37.

As 91.4+36.27.
Cd NC

Ca 85.8+ 4. 27.

Cu 80.8+ 5.87.

Fe 60.0+ 2.97.

Pb 90.2+12.67.

Mn 75.4+ 4,27.
Ni NC

Zn 64.2+12.17.

+ ^.

Tj/.

± 5.

67.

±13.

57.

+ 2.

37.

+ 3.

57.

+ 5.

57.

+ 9.

47.

+ 1.

47.

+65.

67.

+ 7.

97.

1 Based on the analysis o-f MBS reference samples.

2 Based on field replicate samples.

NC = Not calculated because the NBS value and the reference

samples concentration were below the instrument detection
1 imi t .

211

Appendix D-2. Vascular plants and bryophytes of the Swamp Gulch
Wetland study site, 1987.

VASCULAR PLANTS
ASTERACEAE

Greene
B. & H.

Greene

Achillea millefolium L.
Agoseris aurantiaca (Hook.)
Anaphalis margaritacea (L.)
Aster junciforrais Rydb.
Cirsium arvense (L.) Scop.
Erigeron peregrinus (Pursh)
Hieracium scouleri Hook.
Senecio serra Hook.
Tragopogon dubius Scop.

BERBERIDACEAE

Berberis repens (Lindl.) G. Don

BETULACEAE

Alnus incana (L.) Moench

Betula glandulosa Michx. var Hallii (Howell) Hitchc,

BRASSICACEAE

Vicia americana Muhl .
Thapsis arvense L.

CARYOPHYLLACEAE

Cerastium berringianum Cham. & Schlecht.

CUPRESSACEAE

Juniper us communis L.

Carex aquatilis Wahl.
Carex illota Bailey
Carex nebraskensis Dewey
Carex rostrata Stokes
Scirpus acutus Muhl.

CYPERACEAE

ERICACEAE

Arctostaphylos uva-ursi (L.) Spreng.
Chimaphila umbellata (L.) Bart.
Ledum glandulosum Nutt.

212

^

Appendix D-2. Continued.

Pyrola asarifolia Michx.
Vaccinium sp.

EQUITACEAE

Equisetum sp.

Melilotus officinalis (L.) Pallas

JONCACEAE

Juncus ensifolius Wikst.

ONAGRACEAE

Epilobiuin glaberrimum Barbey

PINACEAE

Abies grandis (Dongl.) Forbes

Abies lasiocarpa (Hook.) Nutt.

Picea engelmannii Parry ex Engelm.

Pinus contorta Dongl. ex Lond.

Pinus ponderosa Dongl. ex Laws. & Laws.

Pseudotsuga menziesii (Mirbel) Franco

POACEAE

Agrostis sp.

Bromus carinatus H & A

Bromus ciliatus L.

Danthonia unispicata (Thumb.) Munro ex Macoun

Elymus glaucus Buckl.

Elymus trachycaulus (Link) Gould ex Skinners.

Festuca idahoensis Elmer

Muhlenbergia andina (Nutt.) Hitchc.

Phleum pratense L.

Scolochloa festucacea (Willd.) Link.

ROSACEAE

Fragaria virginiana Miller
Geum macrophyllum Willd.
Rosa woodsii Lindl.

Populus tremuloides Michx.
Salix boothii Dorn

SALICACEAE

213

r

Appendix D-2. Continued.

SPARGANIACEAE

Sparganium minimum Fries

NON-VASCULAR PLANTS

BRYOPHYTES

Brachythecium sp.

Cratoneuron filicinum (Hedw.) Spruce

Isopterygium pulchellum

Kopterygium pulchellum (Hedw.) Jaeq.

Leptobeyum pyriforme (Hedw.) Wils.

Polytrichum juniperinum Hedw.

Sphagnum tenellum Ehrh. ex Hoffm.

214