 

 

llllllllllIlllllllllllllllIlllllllllllllllllllllllllllllllll

3 1293 02074 1876

LIBRARY
Michigan State
University

 

 

 

 

This is to certify that the

thesis entitled

THE EFFECTS OF AN INDUSTRIAL EFFLUENT
ON THE QUALITY OF A MICHIGAN
WARMWATER STREAM

presented by

Ethan Jay Nedeau

has been accepted towards fulﬁllment
of the requirements for

M. S. Entomology

 

degree in

 

    

Major prof I

Date 16 December, 1999

0-7639 MS U i: an Afﬁrmative Action/Equal Opportunity Institution

PLACE IN RETURN BOX to remove this checkout from your record.
To AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE

DATE DUE

DATE DUE

 

 

.L

.1

.
I
z
‘r
A
r

I'JQ v

aggm

 

 

 

 

 

 

 

 

 

 

 

 

 

mm mm.“

THE EFFECTS OF AN INDUSTRIAL EFFLUENT ON THE
QUALITY OF A MICHIGAN WARMWATER STREAM

By

Ethan Jay Nedeau

A THESIS
Submitted to
Michigan State University .
in partial fulﬁllment of the requirements
for the degree of
MASTER OF SCIENCE
Department of Entomology

1999

ABSTRACT

THE EFFECTS OF AN INDUSTRIAL EFFLUENT ON THE QUALITY
OF A MICHIGAN WARMWATER STREAM

By

Ethan Jay Nedeau

I studied the effect of an industrial efﬂuent on the quality of a second-order,
warmwater stream. The eﬂluent increased the total stream discharge by 50-150°/o; this
signiﬁcantly improved downstream physical habitat quality. The eﬁluent carried high
levels of iron precipitate, and caused a considerable increase in water temperatures.
Artiﬁcial substrates were used to compare macroinvertebrate colonization upstream and
downstream of the eﬂluent, with no effort to standardize habitat quality. There were few
signiﬁcant trends in colonization. Macroinvertebrate community composition and
colonization was then compared among sites with comparable habitat quality (riﬂles with
gravel/cobble substrata). There were clear, signiﬁcant differences in macroinvertebrate
diversity and colonization among these sites, with riﬂles immediately downstream of the
efﬂuent supporting the lowest diversity and highest proportion of pollution-tolerant taxa.
Growth and mortality bioassays were conducted with the mayﬂy Stenacron
interpunctatum. Though not statistically signiﬁcant, results ﬁom the two bioassays
suggest that food quality, not water quality per se, likely explains the scarcity of mayﬂies
immediately downstream of the eﬂluent. This study illustrates the importance of
considering both habitat quality and water quality when assessing the effects of point-

souroe pollutants on stream ecosystems.

 

DEDICATION

Well, dad, I’ve earned a lot if titles in the last 24 years - CHAMPION Runner,
EXPERT Fisherman (no laughing), BACHELOR of Science and Woman, ROYAL Pain
in the Neck ...... Sorry you leﬁ before seeing me become a MASTER of Science. I

dedicate this to you.

ACKNOWLEDGEMENTS

Through the drivers-side mirror of my 1984 Dodge station wagon I watched the
sun rise over the hills of western Massachusetts. I was heading west. A passenger seat
full of junk food wrappers, one new alternator, and 900 miles later I arrived at Michigan
State University. Three days later I arrived at Site 001 - the outﬁall of Pharmacia &
Upjohn’s industrial wastewater into Portage Creek. My leg sank deeply into black mud
whenI stepped intothecreek fortheﬁrsttime, andmyheart sankwithit. DidIreally
come all this way to solve a riddle of how an industry can possibly make a severely
degraded stream even worse? My impressions ﬁom tlmt ﬁrst afternoon at Portage Creek
were only reinforced through two years of intensive investigation. As the sun rose over
the broad ﬁelds of eastern Michigan, and the engine of the U-Haul seemed to be roaring
in eager anticipation of a late-August sea-breeze on the coast of Maine, I was left with the
question of what I had gained ﬁom my midwest experience. I thought of all my friends —
sharing a beer with Loren at the boathouse as the sun set over Gull Lake, stumbling home
from the bar with Jeremy, waking in my tent to the sound of Nate yelling, “ETHAN,
GET UP. YOU HAVE TO WRITE YOUR THESIS!” You all know who you are, and I
can’t believe you are reading my thesis - get a life! Some of you helped me above and
beyond the day-to-day maintenance of sanity, and deserve special recognition — Jeremy,
Kelly Nate, Eric, John, Keri, Bill (the ﬁnest bottom picker around), and Julie (not so bad
herself). I also thought of the lessons I had learned, such as how to jimmy a car lock with
a landscaping ﬂag, the value of patching neoprene waders before spending a January
afternoon standing around in a stream, and how to make a graph in MSExcel without

having the tail end of the y-axis label get cut off (a lesson I still struggle with). I also

iv

thought of the guidance and training I received from my committee members — especially
Mike Kauﬁnan for being practical, anal-retentive, and tolerant. Thanks also to my
primary advisor Rich Merritt for granting me a chance to represent Michigan State at the
NABS 5K, and making the lab a more festive place to work in. I also learned to deal
with loss during my time at Michigan State (beyond data, equipment, and sanity),
because myfatherwaskilledonthesame morningthatlwasinaquagmireofrawdata
and SAS© printouts. I thought about canning graduate school and returning east to be
nearer my family, but ultimately it was the friendship and support of my ﬁ’iends in
Michigan that allowed me to weather the ordeal. I thank you all.

Finally, I must give thanks to the Department of Entomology for research and
travel ﬁmds, the Pharmacia & Upjohn Company for funding my research assistantship,
Kellogg Biological Station for summer housing, lab space, and support, and the Michigan

Department of Environmental Quality for project oversight.

TABLE OF CONTENTS

LIST OF TABLES ........................................................................... vii
LIST OF FIGURES ........................................................................ viii
CHAPTER ONE

POSITIVE AND NEGATIVE EFFECTS OF AN INDUSTRIAL EFFLUENT
ON THE BENTHIC MACROINVERTEBRATES OF A
MICHIGAN WARMWATER STREAM

INTRODUCTION ........................................................................ 1
STUDY SITE ............................................................................ 6
METHODS .............................................................................. 8
RESULTS ............................................................................... 12
DISCUSSION ........................................................................... 18
SUMMARY ............................................................................. 24
LITERATURE CITED .................................................................. 27
CHAPTER TWO

GROWTH AND SURVIVAL OF ST ENACRON HVT ERPUNCTATUM
(EPHEMEROPTERA: HEPTAGENIIDAE) EXPOSED TO

AN INDUSTRIAL EFFLUENT
INTRODUCTION ....................................................................... 52
METHODS ............................................................................. 54
RESULTS ............ , .................................................................. 58
DISCUSSION ........................................................................... 59
LITERATURE CITED .................................................................. 65
APPENDIX 1

QUALITATIVE LIST OF MACROINVERTEBRATE TAXA COLLECTED
BY THE DIFFERENT SAMPLING TECHNIQUES,
JUNE 1996 — JUNE 1998 .................................................................. 74

LIST OF TABLES

CHAPTER ONE
TABLE 1 . Water quality and habitat quality variables, adapted ﬁ'om
Karr (1991) ........................................................................................... 32

TABLE 2. Explanation of metrics and indices used to compare

macroinvertebrate communities collected by the three sampling

methods (HI) = Hester-Dendy artiﬁcial substrates, Hess = Hess

samples, Tile = ceramic tiles) ..................................................................... 33

TABLE 3. List of aquatic macrophyte species found within the study

area. Relative proportions of these species are indicated for upstream,

downstream, and within the industrial effluent (0 = Absent, 1 = < 20%,

2 = 20-50%, 3 = 50-80%, 4 = >80°/o) ............................................................ 34

CHAPTER TWO

TABLE 1. Temperature and ﬂow data for in situ bioassay with the mayﬂy

Stenacron intetpunctatum. Experiment 1: May 16 — May 22 (6 days), 1997.

Experiment 2: May 18-May 23 (5 days), 1998. Degree Day = 2‘. daily mean

temperatures > 0° for duration of experiment. Velocity not measured in

Experiment 2 ......................................................................................... 67

TABLE 2. Summary data for the in situ bioassay with the mayﬂy Stenacron
intetpunctatum. Experiment 1: May 16 — May 22 (6 days), 1997. Experiment 2:

May 18-May 23 (5 days), 1998. Mean Growth: A Dry Weight. The p—values

for the overall AN OVA are indicated; tests are considered signiﬁcant of the

p-value < 0.05. See Figures 4-6 for graphical summary, and Table l for

temperature/ﬂow data .............................................................................. 67

TABLE 3. Surmnary data for the in situ bioassay with the myﬂy Stenacron
interpunctatum: Experiment 2. Dates: May 18 - May 23 (5 days), 1998.

The p-values for the overall AN OVA are indicated; tests are considered

signiﬁcant if the p—value < 0.05 .................................................................. 68

LIST OF FIGURES

CHAPTER ONE

FIGURE 1. Map of Portage Creek. A) Broader view of Portage Creek (Scale:

1 inch = 1000 meters), showing the three riﬂe “sites” used for Hess samples

and ceramic tile colonization studies. The dashed lines perpendicular to the
stream channel are locations for habitat and ﬂow analyses. B) Narrow view of
Portage Creek (Scale: 1 inch = ”200 meters), showing 5 locations where
Hester-Dendy artiﬁcial substrates were deployed, including the discharge channel
itself. The circle in Figure 1.3 indicates the location where mussel shells were

found following scouring ﬁom the storm .......................................................

FIGURE 2. A) Pharmacia & Upjohn’s discharge into Portage Creek. B)

.35

Conﬂuence of the discharge chamTel with Portage Creek ..................................... 36

FIGURE 3. A) Trout habitat restoration Site, and the upstream Site for Hess sampling
and ceramic tile colonization. B) Short riﬂe section created by railroad
construction, and upstream site for Hess sampling, collection site for mayﬂies,

and test location for mayﬂy grth experiment 1 .............................................

FIGURE 4. A) Riﬂe section located just upstream of the Kilgore Road bridge, and
downstream site for Hess sampling and ceramic tile colonization. B) Portage Creek

in the lower reaches of its watershed, about 5 miles downstream from the study area. . .

FIGURE 5. Mean depth (m), and mean velocity (m/s) upstream and
downstream of the industrial eﬂuent (+ Standard Errors). Values are
taken ﬁom microhabitat measurements of the Hester-Dendy artiﬁcial

substrates (See Figure l for sampling locations) ..............................................

FIGURE 6. Two representative depth proﬁles from upstream and downstream
of the industrial eﬂuent. The X-axis is inverted to illustrate the proﬁle as a

cross-section of the stream at each transect. See Figure l for transect locations ........

FIGURE 7. Percentage of substrate types upstream and downstream of the
industrial eﬂuent. Values are taken ﬁ'om microhabitat measurements of

Hester-Dendy artiﬁcial substrates. See Figure l for sampling locations ..................

FIGURE 8. Percentage of instream-cover upstream and downstream of the
industrial eﬂuent during the growing season (June-September) and
non-growing season (October-May). Values are taken ﬁom microhabitat
measurements of Hester-Dendy artiﬁcial substrates. See Figure 1 for sampling

locations. A list of aquatic macrophytes is presented in Table 3 ...........................

.37

.38

..39

..39

..40

..40

LIST OF FIGURES (CONTINUED)

FIGURE 9. Log (# Individuals) colonizing Hester-Dendy artiﬁcial substrates
at each sampling unit over the three colonization periods (+ Standard Errors).
N = 11-14 (Upstream), 4-7 (Discharge Channel), 10-12 (Downstream). Overall
ANOVA p-values: October 0.0429, January 0.0001, April 0.0029. Sites
marked with the same letter are not signiﬁcantly different (T ukeys LSD
Method, p > 0.05). See Figure l for sampling locations ..................................... 41

FIGURE 10. Mean # Taxa colonizing Hester-Dendy artiﬁcial substrates at

each sampling unit over the three colonization periods (+ Standard Errors).

N = 11-14 (Upstream), 4-7 (Discharge Channel), 10-12 (Downstream).

Overall ANOVA p—values: October 0.0019, January 0.0001, April 0.0001.

Sites marked with the same letter are not signiﬁcantly different (Tukeys LSD

Method, p > 0.05). See Figure 1 for sampling locations ..................................... 41

FIGURE 1 1. Mean # EPT Taxa colonizing Hester-Dendy artiﬁcial substrates

at each sampling unit over the three colonization periods (+ Standard Errors).

N = 11-14 (Upstream), 4-7 (Discharge Channel), 10-12 (Downstream). Overall
ANOVA p-values: October 0.1070, January 0.0011, April 0.0002. Sites

marked with the same letter are not signiﬁcantly different (Tukeys LSD

Method, p > 0.05). See Figure l for sampling locations ..................................... 42

FIGURE 12. Mean # Ephemeroptera individuals colonizing Hester-Dendy

artiﬁcial substrates at each sampling unit over the three colonization periods

(+ Standard Errors). N = 11-14 (Upstream), 4-7 (Discharge Channel), 10-12
(Downstream). Overall ANOVA p-values: October 0.1791, January 0.0372,

April 0.0333. Sites marked with the same letter are not signiﬁcantly different

(Tukeys LSD Method, p > 0.05). See Figure l for sampling locations ................... 42

FIGURE 13. Mean # Trichoptera individuals colonizing Hester-Dendy

artiﬁcial substrates at each sampling unit over the three colonization periods

(+ Standard Errors). N = 11-14 (Upstream), 4-7 (Discharge Channel), 10-12
(Downstream). Overall ANOVA p-values: October 0.431 1, January 0.1487,

April 0.0429. Sites marked with the same letter are not signiﬁcantly diﬂ'erent

(Tukeys LSD Method, p > 0.05). See Figure l for sampling locations .................... 43

FIGURE 14. H’ diversity of invertebrates colonizing Hester-Dendy artiﬁcial

substrates at each sampling unit over the three colonization periods (+ Standard

Errors). Oligochaetes and chironomids were excluded when calculating H’.

N = 11-14 (Upstream), 4-7 (Discharge Channel), 10—12 (Downstream). Overall
ANOVA p-values: October 0.0001, January 0.0001, April 0.0001. Sites marked

with the same letter are not signiﬁcantly different (Tukeys LSD Method,

p > 0.05). See Figure 1 for sampling locations ................................................ 43

LIST OF FIGURES (CONTINUED)

FIGURE 15. Biotic index of invertebrates colonizing Hester-Dendy artiﬁcial

substrates at each sampling unit over the three colonization periods

(+ Standard Errors). Oligochaetes and chironomids were excluded when

calculating the biotic index. N = 11-14 (Upstream), 4-7 (Discharge Channel),

10-12 (Downstream). Overall ANOVA p-values: October 0.3632, January

0.0001, April 0.0054. Sites marked with the same letter are not signiﬁcantly

different (Tukeys LSD Method, p > 0.05). See Figure 1 for sampling locations .......... 44

FIGURE 16. Log (# Individuals) collected with the modiﬁed-Hess sampler at

riﬂe sites upstream and downstream of the industrial eﬂuent (+ Standard

Errors), exclusive of the chironomids and Oligochaetes. N = 5 per site.

Overall ANOVA p—values: June 0.0079, November 0.0010. Sites marked

with the same letter are not Signiﬁcantly different (T ukeys LSD Method,

p > 0.05). See Figure l for sampling locations ................................................. 45

FIGURE 17. Mean # Taxa collected with the modiﬁed-Hess sampler at riﬂe

Sites upstream and downstream of the industrial eﬂuent (+ Standard Errors).

N = 5. Overall ANOVA p-values: June 0.0006, November 0.0029. Sites

marked with the same letter are not signiﬁcantly different (Tukeys LSD

Method, p > 0.05). See Figure 1 for sampling locations ...................................... 45

FIGURE 18. Mean # EPT Taxa collected with the modiﬁed-Hess sampler

at riﬂe sites upstream and downstream of the industrial eﬂuent (+ Standard

Errors). N = 5. Overall ANOVA p-values: June 0.0027, November 0.0004.

Sites marked with the same letter are not signiﬁcantly different (Tukeys LSD

Method, p > 0.05). See Figure 1 for sampling locations ...................................... 46

FIGURE 19. Mean # Ephemeroptera individuals collected with the modiﬁed-

Hess sampler at riﬂe sites upstream and downstream of the industrial eﬂuent

(+ Standard Errors). N = 5. Overall ANOVA p—values: June 0.0077, November

0.0093. Sites rmrked with the same letter are not signiﬁcantly different

(Tukeys LSD Method, p > 0.05). See Figure 1 for sampling locations ..................... 46

FIGURE 20. Mean # Trichoptera individuals collected with the modiﬁed-Hess

sampler at riﬂe sites upstream and downstream of the industrial eﬂuent

(+ Standard Errors). N = 5. Overall ANOVA p-values: June 0.2907, November

0.374. Sites marked with the same letter are not signiﬁcantly different

(Tukeys LSD Method, p > 0.05). See Figure 1 for sampling locations ..................... 47

LIST OF FIGURES (CONTINUED)

FIGURE 21. H’ Diversity of samples collected with the modiﬁed-Hess sampler

at riﬂe sites upstream and downstream of the industrial eﬂuent (+ Standard

Errors), exclusive of chironomidae and oligochaeta. N = 5. Overall ANOVA

p-values: June 0.0332, November 0.0012. Sites marked with the same letter

are not signiﬁcantly different (Tukeys LSD Method, p > 0.05).

See Figure 1 for sampling locations ............................................................. 47

FIGURE 22. Biotic index calculated for samples collected with the modiﬁed-

Hess sampler at riﬂe sites upstream and downstream of the industrial eﬂuent

(+ Standard Errors), exclusive of oligochaeta and chironomidae. N = 5.

Overall ANOVA p—values: June 0.0539, November 0.1494. Sites marked with

the same letter are not signiﬁcantly different (Tukeys LSD Method, p > 0.05).

See Figure l for sampling locations ............................................................. 48

FIGURE 23. Functional feeding group composition of samples collected with

the modiﬁed Hess sampler at riﬂe Sites upstream and downstream of the

industrial eﬂuent. Figures are based on presence/absence of taxa.

See Figure 1 for sampling locations ............................................................. 48

FIGURE 24. Log (# Individuals) colonizing ceramic tiles at the three riﬂe sites

over the four colonization periods (1: Standard Errors). N = 4—5. Overall

ANOVA p-values: July 0.0717, August 0.0229, September 0.0332, November

0.1765. See Figure 1 for sampling locations ................................................... 49

FIGURE 25. Mean # Taxa colonizing ceramic tiles at the three riﬂe sites over

the four colonization periods (:1: Standard Errors). N = 4—5. Overall AN OVA

p-values: July 0.1060, August 0.1392, September 0.2238, November 0.0618.

See Figure 1 for sampling locations ............................................................. 49

FIGURE 26. % Midges and Worms (Oligochaeta) ) colonizing ceramic tiles at

the three riﬂe Sites over the four colonization periods (:1: Standard Errors).

N = 4—5. Overall ANOVA p-values: July 0.0004, August 0.0009, September

0.0001, November 0.1502. See Figure 1 for sampling locations ........................... 50

FIGURE 27. Mean # EPT Taxa colonizing ceramic tiles at the three riﬂe sites

over the four colonization periods (:1: Standard Errors). N = 4—5. Overall

ANOVA p—values: July 0.0688, August 0.2242, September 0.1445, November

0.0127. See Figure l for sampling locations ................................................... 50

FIGURE 28. Mean # Ephemeroptera colonizing ceramic tiles at the three riﬂe

sites over the four colonization periods (:1: Standard Errors). N = 4—5. Overall

ANOVA p—values: July 0.0223, August 0.0066, September 0.0838, November

0.0657. Figure 1 for sampling locations ........................................................ 51

LIST or FIGURES (CONTINUED)

FIGURE 29. Mean # Trichoptera colonizing ceramic tiles at the three riﬂe Sites

over the four colonization periods (:1: Standard Errors). N = 4—5. Overall

ANOVA p-values: July 0.2862, August 0.0581, September 0.2743,

November 0.1666. Figure l for sampling locations ........................................... 51

CHAPTER TWO

FIGURE 1. Map of Portage Creek ﬁom Portage Police Station (Upstream Riﬂe)
to Milham Park (Downstream Riﬂe), showing locations used during the course
of the mayﬂy growth experiments. Scale: 1 inch = ~1000 meters ........................... 69

FIGURE 2. Elements of the mayfly growth experiment. A) Ceramic tiles placed in the
discharge channel. B) Ceramic tiles after 1 month in the discharge channel, displaying

a heavy accumulation of ferric hydroxide and bacteria. C) A pair of growth chambers
used in Experiment 1; l is open to show the arrangement of food tiles. D) A growth
chamber used in Experiment 2, showing the divider used to separate the chamber into

2 compartments. E) A pair of growth chambers suspended within the cage and
immersed in the stream. F) All growth chambers (6 chambers, 3 cages) at the

industrial outfall during Experiment 2 ........................................................... 70

FIGURE 3. Experimental design for the rmyﬂy growth experiments. See Figure 1 for
a map of test locations, and Figure 2 for photographs of ceramic tiles, growth
chambers, cages, and the experiment in progress .............................................. 71

FIGURE 4. Mean growth (g) of caged Stenacron interpunctatum for the in situ
growth/mortality experiment conducted May 16-May 22, 1997 (+ Standard Errors).

N (cages) = 4 (upstream), 4 (discharge), and 3 (downstream). Overall AN OVA not
signiﬁcant (p-value 0.5918). See Table 2 for summary data, and Figure 1 for Site
locations ............................................................................................. 72

FIGURE 5. Mean grth (g) of caged Stenacron intetpunctatum, corrected for total
degree-day accumulation at each location (+ Standard Deviation). In situ
growth/mortality experiment conducted May 16-May 22, 1997. N (cages) = 4
(upstream), 4 (discharge), and 3 (downstream). Overall AN OVA not signiﬁcant
(p-value 0.8677). Degree days = 2‘. daily mean temperatures >0°C. See Table 2 for
summary data, and Figure 1 for Site locations .................................................. 72

FIGURE 6. Percent survival (+ Standard Deviation) of caged Stenacron interpunctatum .
for the in situ growth/mortality experiment conducted May 16-May 22 1997.

N (cages) = 4 (upstream), 4 (discharge), and 3 (downstream). Overall AN OVA not
signiﬁcant (p-value 0.6019). See Table 2 for summary data, and Figure 1 for site
locations ............................................................................................. 73

LIST OF FIGURES (CONTINUED)

FIGURE 7. Mean growth (g) of caged Stenacron interpunctatum exposed to different
test locations and food sources (upstream versus discharge channel) (+ Standard

Errors). N (chambers) = 6 for each combination. Dates: May 18 - May 23 1998.
Two-factor ANOVA not Signiﬁcant (p-values: Location 0.1978, Food 0.1332,

Location x Food 0.9284). See Table 2 for summary data and Figure l for site

locations ............................................................................................. 75

CHAPTER ONE
THE EFFECTS OF AN INDUSTRIAL EFFLUENT
ON THE BENTHIC MACROINVERTEBRATES OF A
MICHIGAN WARMWATER STREAM.

INTRODUCTION

There are ﬁve primary classes of water resource variables that determine the
structure and function of lotic ecosystems: water quality, habitat structure, ﬂow regime,
energy source, and biotic interactions (Kart 1991, Allan 1995). From a conservation or
management standpoint, it would be valuable to know the relative importance of these
classes of variables. This would aid in establishing protection criteria, or allow focused
restoration or remediation efforts to achieve more efﬁcient and beneﬁcial use of public
resources (Karr 1990, 1991, Allan and Flecker 1993, Scrirngeour and Wicklum 1996).
Unfortunately, these variables operate at different spatial and temporal scales within and
among lotic ecosystems, and most are confounded, such that it becomes difﬁcult to tease
apart the relative importance of each to overall ecosystem structure and function (Power
et al. 1988, Allan 1995).

Water quality has traditionally been the variable of interest in water resource
management and biological assessment (Prati et al. 1971, Karr 1991). Water quality
includes an array of variables of both natural and anthropogenic origin that collectively
make up the physico-chemical nature of the water, including dissolved and suspended
materials (Table 1). These variables provide a means of quantitatively comparing water
quality within and among water bodies. Early legislation emphasized the importance of

water quality because it was thought that clean water would ensure the biological

integrity of aquatic systems (Karr 1991). Much of the focus was on regulating point-
sourcec of pollution, and developing water quality standards based on laboratory toxicity
tests (Cairns and Pratt 1989, Maltby and Calow 1989). Little or no emphasis was placed
on examining non-point source pollution, or human activities that altered other classes of
water resource variables such as habitat quality and ﬂow regime. The combined effects
of non-point source pollutants and other human activities on aquatic systerm precluded
the ability to associate water quality standards with biological integrity (Kart 1991). In
the 1980’s, there was a strong push for water resource assessment and protection
programs that emphasized direct measurement of biological integrity, rather than relying
on water quality standards as surrogate measures (Cairns and Pratt 1989, Kan' 1991).

Direct measures of biological integrity have focused on a variety of organisms,
most commonly invertebrates (Plaﬂdn et al. 1989, Rosenberg and Resh 1993), and ﬁsh
(Karr 1981, 1991). Macroinvertebrates are commonly used as surrogates for biological
integrity because of their importance in the structure and function of lotic ecosystems
(Merritt et al. 1984, Rosenberg and Resh 1993, Wallace and Webster 1996). The
growing understanding of the biotic and abiotic variables that determine
Imcroinvertebrate community structure (Power et al. 1988) has increased the ability to
detect biological impairment in aquatic systems (Rosenberg and Resh 1993). Yet,
identifying the speciﬁc cause(s) of biological impairment has remained elusive due to the
confounded nature of environmental variables, and the bias towards water quality as the
variable of interest in pollution assessments.

Among other water resource variables, habitat quality is of critical importance to

the biological integrity of aquatic systems. Habitat quality is deﬁned by a number of

variables that describe the physical/structural nature of the system (both within the
aquatic system and the adjacent riparian corridor), and collectively determine the degree
of niche heterogeneity and ultimately species diversity (Table 1). Niemi et al. (1990)
reviewed the literature for studies that documented resilience of aquatic systems to
disturbance. They found that recovery ﬁom most types of disturbance were generally
less than three years, unless the physical habitat of the system was altered. Nearly all
cases in which recovery times exceeded ﬁve years involved some sort of habitat
degradation or simpliﬁcation. This suggests that habitat quality is at least as important as
water quality in determining biological integrity, and should perhaps be given equal
consideration in water resource management and biological assessment.

This is not to suggest that habitat quality is ignored in current bioassessment
procedures, or considered unimportant in water resource management (Rankin 1995). A
description of habitat quality is a component of nearly all bioassessment procedures
(Platts et al. 1983, Plaﬂdn et al.1989, Rankin 1995). However, the mere description of
habitat quality variables does little more to predict biological integrity of aquatic systems
than did the mere description of water quality variables called for in early pollution
assessment programs. The question remains the same: how well do descriptions of
variables reﬂect the actual biological integrity of the system? Water resource
assessments have taken an important step to directly measure biological integrity rather
than rely on water quality measures. The same philosophy should be applied to habitat
quality measures. Both water quality and habitat quality Should be expressed in terms of
their contribution to the overall productivity and diversity that the stream ecosystem can

support (Karr 1995, Rankin 1995).

A more holistic understanding of the factors contributing to the biological
integrity of aquatic systems becomes especially important in watersheds that have been
severely impacted by human activity, and receive both point and non-point sources of
pollution that degrade both water quality and lmbitat quality (Karr and Schlosser 1978,
Karr et al. 1985, Karr 1991, Roth et al. 1995). The distinction between water quality and
habitat quality is critical ﬁ'om a restoration or remediation standpoint. For example,
habitat restoration programs that seek to improve biological integrity of aquatic systems
by increasing habitat heterogeneity would not produce the desired effects if water quath
were the limiting factor (NRC 1992). Likewise, targeting point-sources of pollution in
order to improve water quality will not improve the biological integrity of the system if
habitat quality were the limiting factor.

Portage Creek is typical of streams throughout the midwestem United States in
that it’s watershed has been subjected to intensive agricultural, industrial, and urban
development over the last 150 years (Karr et al. 1985, Larnberti and Berg 1995). In
addition to a myriad of other point and non-point sources of pollution that continually
threaten both water quality and habitat quality, a pharmaceutical company releases its
cooling water into Portage Creek. The non-contact cooling water is drawn ﬁom a single
large aquifer, distributed within the plant for cooling purposes, and is then discharged
into Portage Creek at a rate of 5-8 million gallons per day. According to the Pharmacia
& Upjohn Company, Portage, MI, the only alterations to the water other than thermal are
chlorination processes to minimize bacterial growth within the plant, and then

dechlorination (sulfonation) processes prior to its discharge into the environment. The

eﬂucnt also contains input ﬁ'om the storm water drainage system at the manufacturing
facility.

Potential negative effects of this eﬂuent on Portage Creek are ferric hydroxide
precipitation and deposition, and elevated water temperatures. Groundwater in this region
of southwestern Michigan is very high in ferrous iron (Fe2+), and when this is brought to
the surﬁrce and exposed to oxygen, it is oxidized to ferric iron (Fey). In neutral water
most of this ferric iron is hydrolyzed to ferric hydroxide, which forms a yellow-orange
precipitate (‘yellowboy”) and settles out on the substratum. It is known to inhibit the
growth of benthic algae (Sheldon and Skelly 1990, Wellnitz and Sheldon 1995), depress
macroinvertebrate diversity by interfering with feeding and respiration (Koryak et al.
1972, Rasmussen and Lindegaard 1988, Wellnitz et a1. 1994), and has consequences for
higher trophic levels (Letterman and Mitsch 1978). The eﬂuent is also thermally
constant, ranging ﬁ'om 17-24 °C throughout the year. In winter it ranges from 15-20 °C
above ambient stream temperature, and once mixed with the stream water increases the
stream temperature 8-12 °C. Temperature is a critical factor inﬂuencing the life cycle of
most aquatic invertebrates (Ward and Stanford 1992, Sweeney 1984). It has direct
streets on hatching time and survivorship of eggs, larval growth period, timing or ‘
emergence, and adult Size at emergence. It can also aﬂ’ect the population and community
ecology of aquatic invertebrates (Ward 1976), and ecosystem processes (Cairns 1976,
Paul et al. 1978).

Although the industry consistently meets water quality standards for discharges
into local surface waters, the impact of the eﬂuent on Portage Creek is not fully

understood. Previous biological assessments carried out by the Michigan Department of

Natural Resources (MDNR unpublished reports), and independent consulting ﬁrms
(F ishbeck, Thompson, Carr & Huber, Inc. 1991) were unable to conclusively determine
the overall effect of the eﬂuent on Portage Creek. Among the shortcomings of each of
the studies was a fhilure to adequately characterize in-stream habitat quality, provide a
quantitative analysis of the macroinvertebrate communities, and a failure to recognize the
importance of habitat-speciﬁc sampling and confounding variables. The challenge was to
make a fair assessment of the impact of this point-source of pollution on the quality of
Portage Creek, while considering other historical and contemporary inﬂuences on
ecosystem quality. The objectives of this study were to:

(1) assess the immediate impact of the industrial eﬂuent on the quality of Portage
Creek, considering both habitat quality and water quality,

(2) incorporate this assessment with the larger story of historical and
contemporary inﬂuences on the quality of Portage Creek, and

(3) offer insight into the problems associated with conducting upstream-

downstream pollution assessments in an urban/agricultural watershed.

STUDY SITE

Portage Creek is a third-order tributary of the Kalamazoo River in Kalamazoo
County, in southwestern Michigan. Its 52-square kilometer drainage area is a glacial
outwash plain, with up to 350 feet of sand and gravel outwash overlying a Shale bedrock
valley. It originates from groundwater seeps and wetlands, and travels north for about 15
kilometers to its conﬂuence with the Kalannzoo River in the city of Kalamazoo. It

receives substantial recharge ﬁ'om both the groundwater and adjacent wetlands.

Origimlly, the watershed was classiﬁed as an oak savannah, and probably
included patches of dry prairie. The dominant tree species included oaks (Quercus spp. ),
and hickories (Carya spp.). Marshes lined the creek along most of its length. The
watershed was converted to an agricultural landscape in the 19‘” and early 20"n centuries.
In the upper reaches of the watershed, most of the agricultural activity has been
abandoned, and much of the area is now undergoing rapid urbanization. Land use in the
lower reaches of the watershed is primarily industrial and urban. Several large industries
are clustered along the creek in the city of Kalamazoo, including two large paper
industries that are now closed down. Much of this historical industrial complex is now
condemned, and listed as an EPA Superfund Site due to terrestrial and groundwater
contamination (Figure 4B). There are many documented cases of chemical spills,
contaminated groundwater venting to the creek, and other toxic discharges into Portage
Creek over the past few decades. The industrial eﬂuent which is the focus of this study
is located well upstream of the urban/industrial complex in the city of Kalamazoo, but
still downstream of a myriad of point and non-point sources of pollution in the city of
Portage, and outlying areas.

Though not accessible to the general public, people can rent canoes upstream of
the discharge channel and paddle past it. There was concern over the yellow-brown
precipitate that covered the substratum within and downstream of the discharge channel,
and a more general concern that such a conspicuous pollution source detracted ﬁ'om the
aesthetic beauty of the creek. This prompted the Michigan Department of Environmental

Quality to initiate an investigation into the effects of this eﬂuent on Portage Creek.

METHODS

I. HABITAT ANALYSIS

I focused on the following variables to describe the habitat quality upstream and
downstream of the industrial eﬂuent: water depth, water velocity, substrate type, in-
stream cover (i.e. macrophytes, woody material). Water velocity was measured with a
Marsh-McBirney portable ﬂow ureter. Substrate was examined visually and placed into
one of two categories: ﬁne material (silt and sand), or coarse material (gravel and
cobble). Macrophytes were identiﬁed to species using keys of Voss (1985). The habitat
assessment was conducted within 5 12-meter stream sections upstream and downstream
of the discharge channel (2 upstream, 3 downstream) (Figure 1). Habitat variables were
measured at six randomly selected locations within each of the stream sections. In
addition, four sites were selected in order to determine ﬂow heterogeneity and total
discharge (Figure 1). At each of these sites I randomly established two cross-stream
transects, and measured depth and ﬂow velocity at 30 centimeter increments along each
transect. From these data I computed total discharge (m3/s), and depth proﬁles.

The following large-scale habitat descriptors were assessed visually using ﬁeld
surveys and aerial photographs: channel alteration, channel sinuosity, run-bend ratio,
nature of the riparian zone, land use, and potential sources of pollution. These
descriptors reﬂect historical and contemporary landscape-level disturbances that could
confound our assessment of the impact of a single source of pollution. The large-scale
habitat descriptors, and other historical and contemporary inﬂuences on Portage Creek,

are addressed in the discussion section.

II. BENTHIC MACROINVERTEBRATE ASSESSMENT

The initial approach to assess the benthic macroinvertebrate community focused
on a 600-meter section of Portage Creek, ﬁom 100 meters upstream of the eﬂuent to 500
meters downstream of the eﬂuent (Figure 1). Artiﬁcial substrates were deployed into 6
stream sections (Figure 1), which included the ﬁve sites used for the habitat assessment
and an additional Site within the discharge channel itself. Each stream section was 12
meters long, and divided into 24 quadrats. Artiﬁcial substrates were placed into six
randomly selected quadrats within each site, and habitat variables were measured at the
exact location at which each artiﬁcial substrate was placed (See also: HABITAT
ANALYSIS). The samplers were deployed for ﬁve one-month periods from October 1996
to September 1997. I used Hester-Dendy (HD) artiﬁcial substrates; HI) substrates consist
of seven circular masonite plates stacked onto a large eyebolt with spacers in between
each (Hester and Dendy 1962). They were attached to a heavy basal plate, and placed on
the stream bottom (Mathers and Martin 1967). This sampler provides a uniform surface
area for invertebrate colonization, and is accessible to both drifting and crawling
individuals. Samples were collected by placing a pint canning jar over the sampler so
thatthejarﬁttightlyagainstthebasalplate,andinvertingitsothatthemulti-plate
portion was contained within the jar. The sample was eliminated ﬁ'om further analysis if
sediment or leaves buried the sampler.

It became apparent that Site 3 actually encompassed two distinct environments:
the west side (3W) was outside of the eﬂuent plume and more similar to upstream
conditions, and the east side (3B) was within the eﬂuent plume and more similar to the

discharge channel itself. For the analysis, sites 1,2 and 3W were pooled into a Single

‘ pstream” site, site 3E and the discharge channel were pooled into a single “Discharge
Channel” site, and sites 4 and 5 were pooled into a single “Downstream” site.

Since sampling sites and habitat quality were confounded in the previous
approach, I tried a second approach where I sampled similar habitats upstream and
downstream of the discharge channel. I sampled in ﬁve riﬂes located within a 6-
kilometer reach of Portage Creek (Figure 1). The ﬁrst upstream riﬂe is a 75-meter reach
that was restored for trout by the Michigan Department of Natural Resources. The
second upstream riﬂe is where a railroad bed and old service road cross the creek. These
two Sites comprise the “Upstream” site for the analysis. Samples were taken within a
400-meter stretch of riﬂe habitat immediately downstream of the discharge channel
(“Impact Zone”). To test for downstream recovery, I chose riﬂes at least 3-kilometers
downstream of the eﬂuent (“Recovery”). All samples were taken within 200 meters of
the bridge at Kilgore Road (Figure 1). All of the riﬂe sections chosen were similar in
terms of depth, ﬂow velocity, and substrate type.

I sampled these riﬂes with a modiﬁed Hess-sampler (Merritt and Cummins
1996). The sampler is planted ﬁrmly in the substrate, with a 250 um net pointing
downstream. Sediments are stirred within the enclosed area, and all ﬁne particles and
invertebrates are washed into the net. larger stones were examined for attached
invertebrates (such as the limpet F en'issia rivularis, or the caddisﬂy Psychomyiaﬂavida).
Placement of the Hess sampler on the stream bottom was non-random - I was careful to
sample locations with similar depth, ﬂow velocity, and substrate. I used this non-random
approach because I was only interested in assessing diﬂ’erences in water quality between

Sites, without having to take a large number of samples. If I were interested in assessing

10

the spatial distribution or abundance of macroinvertebrates, then a random approach
requiring large sample sizes would be necessary. Five samples were taken at each of the
“sites” (Upstream, Impact Zone, Recovery), on two sampling dates (June 1997,
November 1997), for a total of 30 samples.

Ceramic tiles were used to assess colonization processes in the same riﬂes used
for the Hess sampling (Figure l). Six-inch square ceramic tiles were placed within a
narrow range of depth, ﬂow velocity, and substrate type, with the corrugated surface
pointing down. The tiles were intended to sample clinging/sprawling invertebrates that
graze the surface bioﬁlm. These types of invertebrates would be more sensitive to ferric
hydroxide deposition. The tiles also allowed for a more quantitative assessment of the
differences in chironomid midge colonization between the three riﬂe sites. The
colonization period was 28—32 days, and they were placed in the riﬂes for four

consecutive months.

111. SAMPLE PROCESSING, SORTING, AND ANALYSIS

All samples were either preserved in 90% ethanol in the ﬁeld, or placed on ice
and transported back to the lab for processing. Hester-Dendy and tile samples were
rinsed through a 250 um Sieve, and Hess samples were rinsed through a 500 um sieve.
Samples were stored permanently in 70% ethanol. Rose-bengal stain was added to the
Hess samples to facilitate sorting. Taxa were identiﬁed to genus or species, except for
the following: Class Hydracarina (water mites), Class Oligochaeta (worms), Family

Chironomidae (midges), Family Ceratopogonidae (biting midges), and some early instar

ll

insects. Oligochaetes and chironomids were ennumerated for the HD and tile samples,
but not for the Hess samples.

A number of metrics and indices were used to characterize the macroinvertebrate
communities collected by the two different sampling methods. An annotated list of these
metrics and indices is provided in Table 2. The metrics and indices are displayed
graphically, and analyzed statistically. Analysis ofvariance was performed on each of
the metrics and indices using the ANOVA (equal sample Sizes) or GLM (General Linear
Methods for Imequal sample sizes) procedures in the SAS® statistical software program.
Comparisons among treatment means were performed with Tukey’s LSD test in the same

software package. The signiﬁcance level was p s 0.05 unless otherwise indicated.

RESULTS
1. HABITAT ANALYSIS

The industrial eﬂuent increases the total discharge of Portage Creek by 50 to
150%, based on my ﬂow measurements and historical data provided by USGS gauging
stations upstream and downstream of the industrial eﬂuent. This has immediate and
pronounced effects on ill-stream habitat quality. The added volume of water did not have
an immediate effect on mean stream width, or mean velocity (Figure 5). However,
maximum velocities range from 25-100% higher downstream of the discharge channel,
indicating a faster and more Sharply deﬁned thalweg. Mean depth is 0.2 meters greater
downstream of the discharge charmel (Figure 5). Perhaps more striking than the mean
depths are the depth proﬁles illustrated in Figure 6. The depth proﬁles and ﬂow analyses

illustrate a greater heterogeneity of depth and ﬂow conditions downstream of the eﬂuent

12

due to the substantial increase in total discharge. Substrate was 100% sand and silt
upstream of the discharge channel (F igure 7). Downstream of the discharge channel,
there was a larger proportion of gravel and cobble substrate.

Macrophytes represented a large proportion of total in-stream cover in Portage
Creek. During the summer growing season (June-September), there was nearly 85% in-
stream cover upstream of the discharge channel, and only 60% downstream (Figure 8).
However, most of the in-stream cover upstream is the macrophyte Potamogeton
ﬁlifonnis, which is an important but transient cover type. It’s root system is weak, and
since it is rooted in a fairly unstable sand substratum, it is quickly uprooted and washed
downstream once it senesces in the fall. In the fall and winter (post-senescence), the
upstream reaches are nearly denuded of any in-stream cover, and what little exists was
woody debris that is conﬁned to the margins of the channel There was not such a strong
seasonal difference for in-stream cover downstream of the discharge channel (Figure 8).
Downstream, there was a larger percentage of the macrophytes Sparganium sp. and
Elodea canadensis, both of which are somewhat more resistant to senescence and
degradation than Potamogetonﬁlifonnis. Downstream, the macrophytes were rooted in a
more stable substratum, and exist in a more stable thermal environment. Each of these
factors may enhance their ability to remain active during the winter months. There was
also a Larger amount of large woody debris at the surface of the strearnbed downstream of
the discharge channel; this is in part due to the eﬂuent because the added discharge
allowed the stream to scour away much of the ﬁne sediments, thereby preventing the

rmterial ﬁ'om getting buried. I observed much woody debris upstream of the discharge

l3

channel that was buried under ﬁne sediment and thus not available for colonization by
invertebrates.

Overall, aquatic macrophyte diversity was considerably higher downstream of the
discharge channel (Table 3). Four species were found immediately upstream, though
Potamogetonﬁliformis comprises >80% of the biormss. Eight species were found
immediately downstream, and the two most abundant species were Sparganium sp. and
Elodea canadensis. Ludwigia palustris was dominant within the discharge channel itself.
This species was never found upstream of the eﬂluent, nor was it found ﬂirther than 75

meters downstream of the efﬂuent.

II. BENTHIC MACROINVERTEBRATE ASSESSMENT

The HD samplers were colonized by a total of 46 taxa (Appendix 1); though
worms (Oligochaetes) and midges (Chironomidae) comprised from 65-95% of the
individuals in nearly all samples. The worms and midges were so abundant that they
essentially swamped out all other taxa when computing community indices (H’Diversity,
Biotic Index), so these indices were calculated after excluding these two groups. The
data are shown graphically in Figures 9-15.

The discharge channel consistently supported fewer total individuals than
upstream or downstream sites; this was statistically signiﬁcant in January and April
(Figure 9). Upstream and downstream sites were never statistically different in terms of
total individuals, though there were consistently more upstream. The discharge channel

supported signiﬁcantly fewer taxa for all three sampling periods (Figure 10). Upstream

and downstream sites were not signiﬁcantly different, though in January there were more

14

taxa collected downstream of the industrial eﬁluent (Figure 10). The discharge channel
supported consistently fewer EPT taxa (Figure 1 1), mayﬂy individuals (Figure 12) and
caddisﬂy individuals (Figure 13) than other sites; in fact, no mayﬂies or caddisﬂies were
collected from the discharge channel in January or April. There was a high variability of
mayﬂyandcaddisﬂy individuals collectedatupstreamanddownstreamsites, andthus it
was diﬁicult to demonstrate a signiﬁcant diﬁ‘erence between the sites. Upstream and
downstream were never signiﬁcantly different for any of the EPT metrics, though there
were more rmyﬂies collected downstream of the industrial eﬂluent in January and April
(Figure 12), and more caddisﬂies collected downstream of the industrial efﬂuent for all
three sampling dates (Figure 13). The discharge channel consistently supported a lower
diversitythanupstreamordownstreamsites, andthiswas signiﬁcant inOctoberand
April (Figure 14). Upstream diversity was not signiﬁcantly diﬂ‘erent than discharge
channel diversity in January, and downstream diversity was signiﬁcantly greater than all
other Sites in January. Upstream and downstream sites were not signiﬁcantly different in
October or April (Figure 14).

The biotic index provided less consistent results than the metrics or the diversity
index (Figure 15). In October, there was virtually no difference between each of the sites
in terms of the biotic index. The discharge channel did have a signiﬁcantly greater (more
pollution tolerance) biotic index score in January and April. Upstream and downstream
sites were virtually identical for all three sampling dates.

A total of 53 taxa were collected with the modiﬁed-Hess sampler at the riﬂle sites
(Appendix 1). Forty taxa (39 in June, 23 in November) were collected at the upstream

sites, twenty-two taxa (22 in June, 15 in November) were collected within the impact

15

zone, and thirty taxa (30 in June, 25 in November) were collected at the downstream
recovery sites. The riﬂles in the impact zone supported signiﬁcantly fewer total
individuals than upstream or recovery sites for both sampling dates (Figure 16).
Upstream and recovery sites did not differ in June, but in November the upstream sites
had signiﬁcantly greater number of mdrvrduals (Figure 16). The impact zone also
supported a signiﬁcantly fewer number of taxa than upstream or recovery sites for both
sampling dates (Figure 17), though the other two sites did not differ signiﬁcantly. The
same is true for the mean # EPT taxa, though it is interesting to note that the recovery
zone had a greater number of EPT taxa for both sampling periods, although this was not
signiﬁcant (Figure 18). The recovery sites had a greater number (though not signiﬁcant)
of mayﬂy (Figure 19) and caddisﬂy (Figure 20) individuals in June, though in November
the upstream sites had greater numbers of both (signiﬁcant only for caddisﬂies). The
impact zone had signiﬁcantly fewer mayﬂies than upstream or recovery sites for both
sampling dates (Figure 19). However, the impact zone was not signiﬁcantly different
than upstream or recovery sites in terms of caddisﬂy individuals in June, and was not
different than recovery sites in November (Figure 20).

In June, there was no signiﬁcant difference in diversity between the upstream and
recovery Sites, though the recovery sites did have a slightly higher diversity (Figure 21).
The impact zone was signiﬁcantly less diverse than the upstream or recovery sites. In
November, the recovery sites were signiﬁcantly more diverse than upstream or impact
zone sites, and the impact zone and upstream sites were not signiﬁcantly different. There
was no signiﬁcant diﬁerence in the biotic index between the sites for either sampling date

(Figure 22). Identiﬁcation of worms and chironomids to more meaningﬁrl taxonomic

16

levels would probably help elucidate trends in the biotic index. Since sample sizes were
small (n=5 for each “site”lsampling date), the fact that most of the tests showed
signiﬁcant diﬂemmes suggests that biological differences between the riﬂles are quite
robust. Overall, the data indicate that benthic invertebrate abundance and diversity had
recovered to at least upstream levels by 3 kilometers downstream of the discharge
channel.

Figure 23 shows the trend in the relative proportion of functional feeding groups
at the three sites. Scrapers comprised only 10% ofthe taxa downstream ofthe industrial
discharge, compared to 28% upstream and 27% at Kilgore Road. Shredders comprised
26% of the taxa downstream of the effluent, compared to 15% upstream and only 5% at
Kilgore Road. The number of collector-gatherer taxa gradually increased in a
downstream direction, and there were slightly more predators upstream (Figure 23).

Macroinvertebrate abundance and diversity data collected with the ceramic tiles
are generally consistent with that from the Hess samples. There were few statistically
signiﬁcant trends for the measured metrics, largely because of small sample sizes and
high degree of variability in macroinvertebrate colonization. Ceramic tiles in the impact
zone had a larger number of individuals than other sites in July and September, yet the
Hess samples indicated a statistically signiﬁcant reduction in numbers of organisms ﬁ'om
this same area. The contrasting results are due to the fact that midges (Chironomidae)
and Oligochaetes were ennumerated for the tile samples but not for the Hess samples; the
impact zone had a comparatively higher percentage of these organisms than the upstream
or recovery sites (Figure 26). The impact zone consistently supported fewer number of

macroinvertebrate taxa (Figure 25) and EPT Taxa (Figure 27) than upstream or recovery

17

sites. The recovery sites supported the greatest number of taxa for 3 of the 4 sampling
dates, and had a greater number of EPT Taxa than upstream sites in July and August.
These eﬁ‘ects were not statistically signiﬁcant. The upstream sites usually supported the
greatest numbers of rmyﬂy individuals (Figure 28) and caddisﬂy individuals (Figure 29),
despite the fact that the recovery sites had a greater number oftaxa. Although few ofthe
observations were statistically signiﬁcant, the consistent trends in colonization at the

three “sites” indicate a classic example of impact and subsequent downstream recovery.

DISCUSSION

1. HABITAT QUALITY

Over the last 150 years, the entire Portage Creek watershed has been subjected to
intensive agricultural and urban development. I obtained aerial photographs for the
Portage Creek watershed dating back to 1935, and they showed that much of the
watershed was under intensive agriculture at some point in the last century, but the
amount of land under intensive agriculture has decreased in the last two to three decades.
Much of the land that was cleared for agricultural purposes has reverted back to early-
successional forest, and this natural regeneration has obscured the level to which the
watershed was disturbed in the late 1800’s and early 1900’s. Despite the positive
changes that have taken place within the terrestrial landscape, the habitat quality of
Portage Creek remains poor. The upper reaches of Portage Creek are almost entirely
channelized, with homogeneous depth and ﬂow, 100% sand and silt substrates, and low

structmal diversity.

18

lermacia & Upjohn Co. pumps 5-8 million gallons of heated water into Portage
Creek every day, increasing the total discharge by 50-150%. This has immediate effects
on habitat quality because the added volume of water gives the stream enough energy to
keep ﬁne sediment in suspension and transport it downstream. This increases depth and
ﬂow heterogeneity, and exposes the gravel and cobble substrates that comprised the
original streambed. The shift in substrate particle size has positive effects on stream
invertebrates (Cummins and Lauﬂ‘ 1969). The increase in habitat heterogeneity also
allows for a greater diversity of aquatic macrophytes, which are an important source of
cover and food for aquatic animals (Iversen et al. 1985, Carpenter and Lodge 1986,
Humphries 1996).

An alternative explanation for this shiﬂ in habitat quality is that the industry built
its discharge channel at a natural break in the stream continuum, where it made a
transition from a low gradient reach to a higher gradient reach. However, stream gradient
below the discharge channel is not diﬁerent ﬁ'om upstream. Stronger evidence was
attained using shells of freshwater bivalves in the Family Unionidae. Hundreds of broken
and heavily eroded bivalve shells were found in the riﬁles downstream of the industrial
emuent, yet none were found in the silt/sand substratum upstream. A ﬂash ﬂood in June
of 1997 nearly doubled the total discharge of Portage Creek for a brief period, and
ﬂushed the sediment out ﬁom a 15 meter section of the creek about 100 meters upstream
of the industrial eﬁluent. The ﬂash ﬂood exposed a gravel/cobble streambed by
deepening the channel by 0.2 to 0.6 meters. Unionid shells were numerous in this short
section, representing all of the species that had been found downstream. Within two

weeks, this short section of stream was ﬁlled in with sand, and the shells were once again

19

buried. This chance event is of particular interest for this study. It suggests that many
sections of Portage Creek that are now 100% sand and silt substrate once had a substrate
type suitable for ﬁ'eshwater mussels, and presumably other riﬂe fauna. Over time,
agricultural and urban development in the watershed caused excessive sedimentation of
the creek. This buried the existing streambed, including the assemblage of freshwater
mussels that inhabited the stream. The industry discharged enough water into the creek
to allow it to regain some of its initial habitat quality. It took a point-source pollutant to
reveal what a history of non-point source pollution had buried. Unfortunately, no live

unionid mussels were found in Portage Creek aﬁer two years of study.

11. BENTHIC MACROINVERTEBRATES ASSESSMENT

The results from the Hess samples and ceramic tile samples indicate that the
industrial eﬂuent has a negative impact on the quality of Portage Creek. The riﬂes
immediately downstream of the eﬂuent (“the impact zone”) consistently supported fewer
taxa, a lower community diversity, and have fewer numbers of pollution-intolerant taxa
such as mayﬂies (Ephemeroptera) and caddisﬂies (Trichoptera) than riﬂes upstream or
much farther downstream Despite the fact that the industrial eﬂuent elevates total
stream temperature by as much as 8-12 °C in the winter, temperature is unlikely to be the
cause for the reduction in diversity we see immediately downstream of the industrial
eﬂuent. Although the stream temperature at the recovery sites was slightly cooler than
sites immediately downstream of the discharge channel, it remained 6-10°C above
upstream (natural) temperatures. Despite only a minor recovery of stream temperature at

the recovery sites, there was a very substantial recovery of the macroinvertebrate

20

community. The recovery zone consistently supported a greater number of pollution-
intolerant EPT taxa than the upstream site, indicating that conditions were as good or
better than upstream conditions. This suggests that the factor that limits the
macroinvertebrate community in the impact zone must be localized. This is not meant to
imply that temperature has no effect on the benthic invertebrates. Some taxa were found
at Kilgore Road that were not found at the upstream sites, including the mayﬂies
Tricorythodes sp., Ephemerella lata, Ephemera simulam', the caddisﬂy Psychomyia
ﬂavida, and the snail Goniobasis livescens. Perhaps thermal constancy and warmer
temperatures is the factor that allows these taxa to exist at these sites. The introduced
bivalve, Corbiculaﬂuminea, is also found at the Kilgore Road riﬂes, and its presence is
almost certainly attributed to the warmer temperatures provided by the eﬂuent. This
clam is widely distributed throughout the upper midwest in streams that are either
warmed by industrial eﬂuents (such as nuclear power plants), or whose temperature does
not drop below this species lower lethal limit of 4-5 °C (Mattice and Dye 1978, Graney et
al. 1980). Although the Kilgore Road communities are comprised of Slightly different
taxa, the overall community composition reﬂects favorable water quality and habitat
quality. So although temperature may have an effect on the invertebrates, it is unlikely
that the effect of temperature alone is negative.

Iron deposition is more likely to have altered the macroinvertebrate community
structure in the riﬂes within the impact zone. Ferric hydroxide is evident within the
discharge channel and for about 1000 meters downstream, where the sediments and
aquatic macrophytes are covered with a ﬁne layer of rusty-colored material. Ferric

hydroxide deposition has been shown to have both direct and indirect effects on stream

21

biota. It inhibits the colonization and growth of diatoms and green algae, which
comprises an important primary food source for aquatic food webs (Cummins and Klug
1979, Sode 1983, Sheldon and Skelly 1990, Wellnitz and Sheldon 1994). In a separate
study in Portage Creek, Kauﬁnan (unpublist data) found that tiles placed in the
discharge channel had Signiﬁcantly fewer diatoms, and lower chlorophyll-a than tiles
placed upstream or further downstream. In a small mountain stream in Vermont, Sheldon
and Skelly (1990) found that over a distance of17 meters, the epilithic community of
diatoms and green algae was almost entirely replaced by an iron-depositing bacterium
Leptothrix ochracea. The percent cover of L. ochracea increased ﬁ'om 0.1 % to 99.8 %
over that short distance, and this was due to a 24-fold increase in iron concentration and a
20~fold increase in manganese concentration. Within ~300 meters, the percentage of L
ochracea dropped to less than 10%, and the diatoms and ﬁlamentous algae regained
dominance. At the same study site, Wellnitz et al. (1994) found that macroinvertebrate
diversity and abundance was greatly reduced within the L. ochracea bloom, though
further downstream the diversity and abundance approached that of upstream. These
patterns were also evident in other streams in northern Vermont that also had blooms of
ironodepositing bacteria (Wellnitz et al. 1994). Many other studies have documented
similar drops in the diversity of the macroinvertebrate and ﬁsh communities due to iron
deposition, and the subsequent downstream recovery (Koryak et al.1972, Letterman and
Mitsch 1978, Rasmussen and Lindegaard 1988).

It is possible that ferric hydroxide and temperature act in synergy immediately
downstream of the discharge channel. The combined effects of poor food quality and

high temperature (increased metabolism) could have a large effect on some animals.

22

Clogging of gills with ferric hydroxide coupled with high temperature (increased
respiratory demands) could also have a large effect on some animals, especially those
with external gills (such as heptageniid mayﬂies).

Results from the Hester-Dendy artiﬁcial substrates indicate that the industrial
eﬂuent has no effect on the biological integrity of Portage Creek. The discharge channel
itself supports a sparse community of pollution-tolerant invertebrates, but whatever was
inhibiting the development of a healthy macroinvertebrate community within the
discharge channel was apparently not operating in the creek itself Important riﬂe
insects such as hydropsychid caddisﬂies and heptageniid mayﬂies were more abundant
downstream of the discharge channel than upstream. This indicates that the industrial
eﬂuent may actually improve conditions for some important pollution-intolerant taxa.
Given the difference in habitat conditions between upstream and downstream sites, it is
not surprising that we should ﬁnd a more diverse macroinvertebrate community
downstream of the industrial eﬂuent. All else being equal, an artiﬁcial substrate placed
in a good microhabitat (gravel or cobble) should be colonized by a greater diversity of
macroinvertebrates than one placed in a poor microhabitat (sand or silt) because there is a
greater local source pool of potential colonizers (Osman 1982). If anything, it was
surprising that we didn’t see an even greater number and diversity of macroinvertebrates
downstream of the industrial eﬂuent compared to upstream.

However, in addition to improving habitat quality (as measured by depth, ﬂow
velocity, and substrate particle size), the industrial eﬂuent also reduces water quality due
to thermal pollution and ferric hydroxide precipitation. The Letter ultimately degrades

habitat quality by blanketing the stream substratum. This reduction in water quality is

23

thought to account for the discrepancy between invertebrate community that could
potentially colonize the favorable habitat downstream of the eﬂuent versus the
invertebrate community that actually inhabits such sites.

This result is partially dependent on the sampling method chosen. If one were
interested in expressing the shift in habitat quality in terms ofwhat it means for the
macroinvertebrate community, then it would not be valid to use a sampling method that
attempts to overcome habitat diﬂ‘erences between sites (such as artiﬁcial substrates). A
more active sampling method (such as a Hess sampler or a Surber sampler) would be
more effective at showing differences in macroinvertebrate communities due to habitat
quality. If an active sampling method had been used, a much larger positive effect of the
industrial eﬂuent on the stream biota would have been demonstrated. This is because
the sand habitat is inhabited by only a few taxa in Portage Creek: nematodes (Nematoda),
ﬂatworms (Turbellaria), worms (Oligochaeta), water mites (Hydracarina), clams
(Sphaeriidae), midges (Chironomidae) a few beetles (Coleoptera), and dragonﬂy and
darnselﬂy larvae (Odonata). Many more taxa are found in riﬂes downstream of the

discharge channel.

SUMMARY

The variable of interest in water resource management and biological assessment
has always been water quality; this is especially true for assessing the impacts of point-
source pollutants on aquatic systems (Karr 1991). Other water resource variables, such
as habitat quality and ﬂow regime, are often relegated to the status of confounding

variables. Standard procedures emphasize the importance of sampling identical habitats

24

upstream and downstream of a source of pollution, so that differences in
macroinvertebrate communities can be attributed to water quality. Artiﬁcial substrates
are tremendously popular in aquatic research, and much of their appeal is that they
provide a uniform colonization area so that the biological community can be assessed
independent of variation in the natural substratum (Rosenberg and Resh 1982, 1993).

What if a point-source of pollution alters habitat quality? In this case, habitat
quality should not be considered merely a confounding ﬁrctor that serves only to obscure
water quality differences between Sites. Instead, it should become as important as water
quality in the overall assessment of the pollutant’s impact on the stream (Rankin 1995).
Equal consideration of water quality and habitat quality causes the complexity of the
assessment to increase, but also provides a more comprehensive and realistic assessment
of the overall impact of the pollutant on biological integrity. The success of restoration
or remediation programs is dependent on an accurate assessment of the causes of
observed effects (Karr 1991, NRC 1992, Davis and Simon 1995).

In this study, if only the results from the Hess samples were presented, then it
would be clear that the industry was having a negative impact on the quality of Portage
Creek. This assumes that the control sites are representative of their respective reaches.
We know this is not the case, because the two upstream riﬂe Sites were the only sites
upstream of the discharge channel that had a riﬂe habitat suitable for the establishment
of a macroinvertebrate commlmity that we typically consider indicative of a healthy,
unpolluted stream. Using the same sampling method, riﬂes immdiately‘ downstream of
the discharge channel would compare favorably to ~ 95% of the locations upstream of the

discharge channel.

25

If only the results from the artiﬁcial substrates were presented, then we would
probably conclude that the industry was having no effect on the biological integrity of
Portage Creek, or even a slightly positive effect. This is due to the fact that the eﬂuent
increases total stream discharge and improves in-stream habitat quality. The habitat
analysis and the two sampling approaches provide comprehensive insight into the overall
effect of the industrial eﬂuent on Portage Creek: by improving habitat it has a slightly
positive effect on biological integrity at one scale, but by reducing water quality it has a
negative effect on biological integrity at another scale.

In this situation, the water resource manager laces a bit of a dilemma: is it better
to have good water quality at the expense of habitat quality, or good habitat quality at the
expense of water quality? Shutting down the eﬂuent would cause a rapid degradation of
habitat quality, and lower the productivity and diversity of Portage Creek. We would be
sacriﬁcing habitat quath for water quality. Alternatively, allowing the industry to
continue business as usual would maintain current biological integrity. We would be
sacriﬁcing localized reductions in water quality for habitat quality. Perhaps it is unwise
to set a precedent of trading water quality for habitat quality. However, restoring
biological integrity to aquatic systems that have been virtually destroyed by decades or
centuries of human activities is no easy task. It requires an understanding of which
factors most strongly limit biological integrity within each system, and an evaluation of
the costs and beneﬁts of diﬁ‘erent restoration or remediation strategies. Taking the
quickest and easiest approach to try to improve biological integrity, such as targeting
point sources of pollution rather than addressing non-point sources, may yield no net

improvement in the health of our aquatic systems (Karr 1991). This study demonstrates

26

that targeting a point-source of pollution may actually be detrimental to overall biological

integrity.

LITERATURE CITED

Allan, J .D. 1995. Stream ecology: Structure and Function of Running Waters. Chapman
& Hall. 388 pages.

Allan, J.D., and AS. Flecker. 1993. Biodiversity conservation in running waters:
identifying the major factors that threaten the destruction of riverine Species and
ecosystems. Bioscience 43:32-43.

Cairns, J. Jr., and J.R. Pratt. 1989. The scientiﬁc basis of bioassays. Hydrobiologia
188/ 189:5-20.

Cairns, J .Jr. 1976. Heated waste-water eﬂ‘ects on aquatic ecosystems. Pages 32-38 In:
Thermal Ecology 11 (G.W. Esch and R.W. McFarlane, editors) ERDA Symposium
Series (CONF-750425), pp. 302-307.

Carpenter, SR, and D.M. Lodge. 1986. Effects of submersed macrophytes on
ecosystem processes. Aquatic Botany 26:341-370

Cover, E.C., and RC. Harrel. 1978. Sequences of colonization, diversity, biomass, and
productivity of macroinvertebrates on artiﬁcial substrates in a ﬁ'eshwater canal.
Hydrobiologia 59:81-95.

Cummins, KW., and OH. Lauﬂ‘. 1969. The inﬂuence of substrate particle size on the
microdistribution of stream macrobenthos. Hydrobiologia 34:145-181.

Cummins, K.W., and M.J. Klug. 1979. Feeding ecology of stream invertebrates.
Annual Review of Ecology and Systematics. 10:147-172.

Davis, W.S., and T.P. Simon. 1995. Biological Assessment and Criteria: Tools for
Water Resource Planning and Decision Making. Lewis Publishers.

Gagnier, D.L., and RC. Bailey. 1994. Balancing loss of information and gains in
eﬂiciency in characterizing stream sediment samples. Journal of the North American
Benthological Society 13:170-180.

Graney, R.L., D.S. Cherry, J.H. Rodgers Jr., and J.Cairns Jr. 1980. The inﬂuence of
thermal discharges and substrate composition on the population structure and
distribution of the Asiatic clam, Corbiculaﬂuminea, in the New River, Virginia. The
Nautilus 94:130-135.

27

Hester, FE, and IS. Dendy. 1962. A multiple-plate sampler for aquatic
macroinvertebrates. Transactions of the American Fisheries Society 91:420-421.

Hilsenhoﬁ‘, W.H. 1987. An improved biotic index of organic stream pollution. The
Great Lakes Entomologist 20:31-39.

Humphries, P. 1996. Aquatic macrophytes, macroinvertebrate associations and water
levels in a lowland Tasmanian river. Hydrobiologia 321 :219-233.

Iversen, T.M., J. Thorup, T. Hansen, J. LodaL and J. Olsen. 1985. Quantitative estimates
and community structure of invertebrates in a macrophyte rich stream. Archiv ﬁir
Hydrobiologie 102:291-301.

Karr, J.R. 1981. Assessment of biotic integrity using ﬁsh communities. Fisheries 6:21-
27.

Karr, J .R. 1990. Biological integrity and the goal of environmental legislation: lessons
for conservation biology. Conservation Biology 4:244-250.

Karr, J .R. 1991. Biological integrity: a long-neglected aspect of water resource
management. Ecological Applications 1:66-84.

Karr, J .R., and 1.]. Schlosser. 1978. Water resources and the land-water interface.
Science 201 :229-234.

Kart, J.R., L.A. Toth, and DR. Dudley. 1985. Fish communities of midwestem rivers: a
history of degradation. Bioscience 35:90-95.

Koryak, M., M.A. Shapiro, and J.L. Sykora. 1972. Riﬂe zoobenthos in streams
receiving acid mine drainage. Water Research 6:1239-1247.

Lamberti, G.A., and MB. Berg. 1995. Invertebrates and other benthic features as
indicators of environmental change in Juday Creek, Indiana. Natural Areas Journal
15:249-258.

Lamberti, G.A., and V.H. Resh. 1979. Substrate relationships, spatial distribution
patterns, and sampling variability in a stream caddisﬂy population. Environmental
Entomology 8:561-567.

Lenat, DR. 1993. A biotic index for the southeastern United States: derivation and list
of tolerance values, with criteria for assigning water-quality ratings. Journal of the
North American Benthological Society 12:279-290.

Letterman, RD, and W.J. Mitsch. 1978. Impact of acid mine drainage on a mountain
stream in Pennsylvania. Environmental Pollution 17:53-73.

Maltby, L., and P. Calow. 1989. The application of bioassays in the resolution of
environmental problems: past, present and ﬁrture. Hydrobiologia 188/ 189:65-76.

28

Mason, W.T., and PP. Yevich. 1967. The use of phloxine B and rose bengal stains to
facilitate sorting benthic samples. Transactions of the American Microscopical
Society 86:221-223.

Mathers, GK, and TR Martin. 1969. A new multiplate sampler for collecting
macroinvertebrates of the stream biocies. Transactions of the Illinois Academy of
Science 62:331-333.

Mattice, 1.8., and LL. Dye. 1976. Thermal tolerance of the adult Asiatic clam. Pages
130-135 in: Thermal Ecology II (G.W. Esch and R.W. McFarlane, editors) ERDA
Symposium Series (CONF-750425), pp. 302-307.

Meier, P.G., D.L. Penrose, and L. Polak. 1979. The rate of colonization by macro-
invertebrates on artiﬁcial substrate samplers. Freshwater Biology 9:381-392.

Merritt, R.W., and K.W. Cummins. 1996. Trophic relations of macroinvertebrates.
Pages 453-474 in: Methods in Stream Ecology (F .R Hauer and G.A. Lamberti, eds.)
Academic Press.

Merritt, R.W., K.W. Cummins, and T.M. Burton. 1984. The role of aquatic insects in
the processing and cycling of nutrients. Pages 134-163 In: The Ecology of Aquatic
Insects (V .H. Resh and D.M. Rosenberg, eds.) Praeger Scientiﬁc, New York 625 p.

Michigan Department of Natural Resource (MDNR). 1991. Qualitative biological and
habitat survey protocols for wadeable streams and rivers. GLEAS Procedure No. 51.
Michigan DNR, Surface Water Quality Division, Great Lakes Environmental
Assessment Section.

Niemi, G.J., P. DeVore, N. Detenbeck, D. Taylor, A. Lima, J. Pastor, J.D. Yount, and
R]. Naiman. 1990. Overview of case studies on recovery of aquatic systems ﬁom
disturbance. Environmental Management 14:571-587.

NRC (National Research Council). 1992. Restoration of Aquatic Ecosystems: Science,
Technology, and Public Policy. National Academy Press, Washington, DC.

Osman, R.W. 1982. Artiﬁcial substrates as ecological islands. Pages 71-130 in:
Artiﬁcal Substrates (J. Cairns Jr., editor). Ann Arbor Science.

Paul, R.W.Jr., E.F. Benﬁeld, and J. Cairns Jr. 1978. Effects of thermal discharge on leaf
decomposition in a river ecosystem International Journal of Applied and Theoretical
Linmology 20:1759-1766.

Plaﬂcin, J.L., M.T. Barbour, K.D. Porter, SK Gross, and RM. Hughes. 1989. Rapid
bioassessment protocols for use in streams and rivers. Benthic macroinvertebrates
and ﬁsh. EPA 440-4-89-001. Ofﬁce of Water Regulations and Standards, US.
Environmental Protection Agency, Washington, DC.

29

Platts, W.S., W.F. Megahan, and G.W. Minshall. 1983. Methods for evaluating stream,
riparian, and biotic conditions. General Technical Report No. INT-138, US.
Department of Agriculture, Forest Service, Intermountain Forest and Range
Experiment Station, Ogden, Utah.

Power, M.E., R.J. Stout, C.E. Cushing, P.P. Harper, F .R. Hauer, W.J. Matthews, P.B.
Moyle, B. Statzner, and LR. Wais De Badgen. 1988. Biotic and abiotic controls in
river and stream communities. Journal of the North American Benthological Society
7:456-479.

Prati, L., R. Pavanello, and F. Pesarin. 1971. Assessment of surface water quality by a
single index of pollution. Water Research 5:741-751.

Rankin, ET. 1995. Habitat indices in water resource quality assessments. Pages 181-
208 in: Biological Assessment and Criteria: Tools for Water Resource Planning and
Decision Making (W.S. Davis and T.P. Simon, editors). Lewis Publishers 415 pages.

Rasmussen, K., and C. Lindegaard. 1988. Eﬁects of iron compounds on
macroinvertebrate communities in a Danish lowland river system. Water Research
22:1101-1108.

Richards, C., G.E Host, and J.W. Arthur. 1993. Identiﬁcation of predominant
environmental factors structuring stream macroinvertebrate communities within a
large agricultural catchment. Freshwater Biology 29:285-294.

Rosenberg, D.M., and V.H. Resh. 1982. The use of artiﬁcial substrates in the study of
ﬁeshwater benthic macroinvertebrates. Pages 175-235 in: Artiﬁcial Substrates (J.
Cairns Jr., editor). Ann Arbor Science, Ann Arbor, Michigan.

Rosenberg, D.M., and V.H. Resh. 1993. Freshwater biomonitoring and benthic
macroinvertebrates. Chapman & Hall. 488 pages.

Roth, N.E., J.D. Allan, and D.L. Erickson. 1995. Landscape inﬂuences on stream biotic
integrity assessed at multiple spatial scales. Landscape Ecology incorrect.

Scrimgeour, G.J., and D. Wicklum. 1996. Aquatic ecosystem health and integrity:
problems and potential solutions. Journal of the North American Benthological
Society 15:254-261.

Sheldon, SP, and D.K. Skelly. 1990. Differential colonization and growth of algae and
ferromanganese-depositing bacteria in a mountain stream. Journal of Freshwater
Ecology 5:475-485.

Sode, A. 1983. Effect of ferric hydroxide on algae and oxygen consumption by sediment
in a Danish stream. Archives fur Hydrobiologie 1:134-162.

30

Voss, E.G. 1972. Michigan Flora: A guide to the identiﬁcation and occurrence of the
native and naturalized seed-plants of the state: Part II Dicots (Saururaceae-
Cornaceae). Cranbook Institute of Science Bulletin 59.

Wallace, J.B., and J.R. Webster. 1996. The role of macroinvertebrates in stream
ecosystem function. Annual Review of Entomology 41:115-139. '

Ward, J.V. 1976. Eﬁ‘ects of thermal constancy and seasonal temperature displacement
on community structure of stream macroinvertebrates. In: Thermal Ecology 11 (G.W.
Esch and R.W. McFarlane, editors), ERDA Symposium Series (CONF-750425), pp.
302-307.

Ward, J.V., and J.A. Stanford. 1982. Thermal responses in the evolutionary ecology of
aquatic insects. Annual Review of Entomology 27:97-117.

Wellnitz, T.A., and SP. Sheldon. 1995. The effects of iron and manganese on diatom
colonization in a Vermont stream. Freshwater Biology 34:465-470.

Wellnitz, T.A., K.A. Grief, and SP Sheldon. 1994. Response of macroinvertebrates to
blooms of iron-depositing bacteria. Hydrobiologia 281 :1-17.

Williams, DD, and MR. Smith. 1996. Colonization dynamics of river benthos in
response to local changes in bed characteristics. Freshwater Biology 36:237-248.

31

TABLE 1. Water quality and habitat quality variables, as deﬁned by Karr (1991).

“ F” '“‘ ' -'--- "1W" '1: ‘w' t-s'. mam-imam: 3"“ ,‘ 1W 3.. . area-av WWW“ tit/km EMS-i=7
r‘ ‘ . ,

LASS , ”1”?" ..MPO

Water Quality Temperature, dissolved oxygen, pH, turbidity, nutrients
(primarily nitrogen and phosphorus), organic and morganic
chemicals, heavy metals, other toxic substances

Habitat Quality Substrate type and distribution, water depth, current velocity,
habitat diversity (pools, riﬂes), in-stream cover (woody debris,
undercut banks), spawning and nursery areas, basin size and
shape

  

-.-v.:.. WWWMRmmmIE-t mwxmwmwwmmmmw .Hil wrun- V‘ﬁh

 

32

TABLE 2. Explanation of metrics and indices used to compare meroinvertebrate
communities collected by the three sampling methods (HD = Hester-Dendy artiﬁcial
substrates, Hess = Hess samples, Tile = ceramic tiles).

 

 

 

(#Indrvrduals) 4'

Mean # Taxa

Mean # EPT Taxa

Ephemeroptera Individuals

Trichoptera Individuals

% Midges and Worms

H’ Diversity

Biotic Index

Functional Feeding Groups

Hess

Hess

 

Log-transformed number of
macroinvertebrates m a sample or collection

of samples.
Average number of macroinvertebrate taxa

in a sample or collection of samples.

Average number of taxa in a sample or
collection of samples which belong to the
insect families Ephemeroptera, Plecoptera,
or Trichoptera.
Average number of individuals in a sample
or collection of samples that belong to the
Family Ephemeroptera.
Average number of individuals in a sample
or collection of samples that belong to the
Family Trichoptera.
Percentage of total individuals in a sample
belonging to the family Chironomidae or the
Class Oligochaeta.
Shannon-Weiner diversity. This diversity
measure is calculated as:

H’ = -2(Na/N) 10810 (NJN)
Where N, = nmnber of individuals of species
i, andN=totalnumberofindividualsina
sample. The higher the value of H’, the more
diverse our sample (Hayek and BuzaS 1997).
Combines two commonly used biotic indices
for eastern North America (Hilsenhoﬂ~ 1987,
Lenat 1993). Hilsenhoﬂ’s values were used
preferentially, but Lemat’s were used when
Hilsenhoﬂ' failed to provide a value
(particularly for non-insect taxa). Scores
range ﬁ‘om 0 to 10, with low scores
indicating little pollution tolerance.
Percentage of each of the major functional
feeding groups (Scrapers, F iltering-
Collectors, Gathering-Collectors, Shredders,
and Predators), based on presence/absence of
taxa (Merritt and Cummins 1995).

33

TABLE 3. List of aquatic macrophyte species found within the study area. Relative
proportions of these species are indicated for upstream, downstream, and within the
industrial eﬂuent (0 = Absent, 1 = < 20%, 2 = 20-50%, 3 = 50-80%, 4 = >80%).

 

 

Elodea canadensrs 1 0 2
Ludwigia palustris 0 4 1
Potamogeton crispus 1 0 1
Potamogetonﬁliformis 4 0 1
Potamogeton zosteriformis 1 0 1
Sparganium spp. 0 0 3
Zannichellia palustris 0 0 1
Unidentiﬁed Poaceae 0 l l

 

34

UPSTREAM RIFFLE \

UPSTREAM

 
  
  

 
   
  
  

SITE 1

DISCHARGE

RAILROAD CROSSING \
CHANNEL

DISCHARGE SITE 2

CHANNEL\ \

IMPACT
ZONE

 

 

RECOVERY
\

DOWNSTREAM Sm: 5

A. B.

FIGURE 1. Map of Portage Creek. A) Broader view of Portage Creek (Scale: 1 inch = "
1000 meters), Showing the three riﬂe “sites” used for Hess samples and ceramic tile
colonization. The dashed lines perpendicular to the stream channel are locations for
habitat and ﬂow analyses. B). Narrower view of Portage Creek (Scale: 1 inch = “200
meters), showing the ﬁve stream sections where Hester-Dendy artiﬁcial substrates were
deployed, including the discharge channel itself. The circle in Figure 1.B indicates the
location where mussel shells were found following excavation by the storm.

35

 

  

. I. ‘
,. . ~
._ ,
' ,. , .
- ' . . ..
. .v r
-. .. . . ‘
. - . .. w
~ ‘. 1' ‘ ' .
., . ,‘s .
r . . .‘
". - ~ . ﬂ ,
.. ~ . . .
u ' . ’ . _ ~
. , - x . - r
. ‘. i . 1

a? :

    
 

Figure 2. A) Pharmacia & Upjohn’s industrial outfall. B) Conﬂuence of
the discharge channel with Portage Creek.

36

 

Figure 3. A) Trout habitat restoration site; one upstream site for
Hess sampling and ceramic tile colonization. B) Short rifﬂe section
created by the railroad bridge construction; upstream site for Hess
sampling, collection site for mayﬂies, and test location for mayﬂy
growth experiment.

37

 

Figure 4. A) Rifﬂe located just upstream of the Kilgore Road
bridge; downstream site for Hess sampling and ceramic tile
colonization. B) Portage Creek in the lower reaches of its
watershed, about 5 miles downstream of the study area. Two
crumbling paper mills in this area have left the soils and water
contaminated.

38

(In) or (m/s)

 

 

 

FIGURE 5. Mean depth (m), and mean velocity (m/s) upstream and downstream of the
industrial eﬂuent (+ Standard Errors). Values are taken from microhabitat

measurements of the Hester-Dendy artiﬁcial substrates (See Figure l for sampling
locations).

Cross-Stream Distance
0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0

 

 

 

P

G

O
1

 

FIGURE 6. Two representative depth proﬁles ﬁ'om upstream and downstream of the
industrial eﬂuent. The X-axis is inverted to illustrate the proﬁle as a cross-section of the
stream at each transect. See Figure 1 for transect locations.

39

100.00 1

  
 

 

& 80.00 '
[3. ISIlt&Sand
.3 60.00 ' ElGravel&Cobble
L-
3 40.00 -
a
m
e\° 20.00

0.00 —

Upstream Downstream
Location

FIGURE 7. Percentage of substrate types upstream and downstream of the industrial
eﬂuent. Values are taken from microhabitat measurements of Hester-Dendy artiﬁcial

substrates. See Figure l for sampling locations.

I October-May
m June-September

% Cover

 

 

Upstream Downstream

Sampling Date

FIGURE 8. Percentage of instream-cover upstream and downstream of the industrial
eﬂuent during the growing season (June-September) and non-growing season (October-
May). Values are taken from microhabitat measurements of Hester-Dendy artiﬁcial
substrates. See Figure 1 for sampling locations. A list of aquatic macrophytes is
presented in Table 3.

40

'm‘ 3.000 I

:3 2.500 ‘

'5 ClDischargGmnel
3 1.500-130mm

.3

g
E

 

October 1996 January 1997 All-ll 1997
Sanqiing Date

FIGURE 9. Log (# Individuals) colonizing Hester—Dendy artiﬁcial substrates at each
sampling unit over the three colonization periods (+ Standard Errors). N = 11-14
(Upstream), 4-7 (Discharge Channel), 10-12 (Downstream). Overall ANOVA p-values:
October 0.0429, January 0.0001, April 0.0029. Sites marked with the same letter are not
signiﬁcantly different (Tukeys LSD Method, p > 0.05). See Figure 1 for sampling
locations.

18.00 “
16.00 E
14.00 -

I Upstream
[:1 Discharge Channel
5 Downstream

Mean # Taxa

 

 

October 1996 Janmry 1997 April 1997
Sampling Date

FIGURE 10. Mean # Taxa colonizing Hester-Dendy artiﬁcial substrates at each sampling
unit over the three colonization periods (+ Standard Errors). N = 11-14 (Upstream), 4-7
(Discharge Channel), 10-12 (Downstream). Overall ANOVA p-values: October 0.0019,
January 0.0001, April 0.0001. Sites marked with the same letter are not signiﬁcantly
different (Tukeys LSD Method, p > 0.05). See Figure 1 for sampling locations.

41

EPT Taxa

 

   

January 1997 Alril 1997
Sanﬂing Date

FIGURE 11. Mean # EPT Taxa colonizing Hester-Dendy artiﬁcial substrates at each
sampling unit over the three colonization periods (+ Standard Errors). N = 11-14
(Upstream), 4-7 (Discharge Channel), 10-12 (Downstream). Overall ANOVA p-values:
October 0.1070, January 0.0011, April 0.0002. Sites marked with the same letter are not
signiﬁcantly different (Tukeys LSD Method, p > 0.05). See Figure 1 for sampling
locations.

9.00 -
8.00 -
7.00 .
6.00 -
5.00 -

   

IUpstrean
UDischarg Gum]
EDoumtream

y Individual:

 

FIGURE 12. Mean # Ephemeroptera individuals colonizing Hester-Dendy artiﬁcial
substrates at each sampling unit over the three colonization periods (+ Standard Errors).
N = 11-14 (Upstream), 4-7 (Discharge Channel), 10-12 (Downstream). Overall ANOVA
p-values: October 0.1791, January 0.0372, April 0.0333. Sites marked with the same
letter are not signiﬁcantly different (Tukeys LSD Method, p > 0.05). See Figure 1 for

sampling locations.

42

10.“) 1

8.001

    

7.00 ~
6.00 ~ lUpsuean

- DDisclnr-g Oml
4.00 T EIbwistIeam

l

.1.

Caddist'ly Individual
5:.
8

I“

HHEHHIIHIIH

  

~HlllllllllllllIlHlllIlllllHHHHllIIHIH

lIllllllllllllllllllllllllilllllllllll

III

II!

   

 

SanﬁngDate

FIGURE 13. Mean # Trichoptera individuals colonizing Hester-Dendy artiﬁcial substrates
at each sampling unit over the three colonization periods (+ Standard Errors). N = 11-14
(Upstream), 4-7 (Discharge Channel), 10-12 (Downstream). Overall ANOVA p-values:
October 0.4311, January 0.1487, April 0.0429. Sites marked with the same letter are not
signiﬁcantly diﬂ‘erent (1‘ ukeys LSD Method, p > 0.05). See Figure 1 for sampling
locations.

1.4M —
1.200 i
1.000 r
0.800 “
0.6m *
0.4M *
0.200 r
0.000 -

  

IUpstr-eun
UDisclnry Om!
IDovmstream

11' Diversity

   

lEl|lIHHHIIHIIHHUIIIHIHHHIHII

October 1996 January 1997 Alli] 1997
Stunning Date

 

FIGURE 14. H’ diversity of invertebrates colonizing Hester-Dendy artiﬁcial substrates at
each sampling unit over the three colonization periods (+ Standard Errors). Oligochaetes
and chironomids were excluded when calculating H’. N = 11-14 (Upstream), 4-7
(Discharge Channel), 10-12 (Downstream). Overall ANOVA p-values: October 0.0001,
January 0.0001, April 0.0001 . Sites marked with the same letter are not signiﬁcantly
diﬁ’erent (Tukeys LSD Method, p > 0.05). See Figure 1 for sampling locations.

43

   
   

  

12.“) r b
g 10.” 7
E 8.“ _ IUrstrean
.2 EIDisclupOnml
:5: 6'00 _ 1; EDmnstream
” 4.00 «

 

IllllllHWHIWHHIIIHHII!
U?!llllIIIIIIHHIHIIHWH

lF'lilHl||Hl|lHHHl

October 1996 January 1997 Apil 1997
Sam Date

FIGURE 15. Biotic index of invertebrates colonizing Hester-Dendy artiﬁcial substrates at
each sampling unit over the three colonization periods (+ Standard Errors). Oligochaetes
and chironomids were excluded when calculating the biotic index. N = 11-14
(Upstream), 4-7 (Discharge Channel), 10-12 (Downstream). Overall ANOVA p-values:
October 0.3632, January 0.0001, April 0.0054. Sites marked with the same letter are not
signiﬁcantly different (Tukeys LSD Method, p > 0.05). See Figure 1 for sampling
locations.

I Upstream
C1 Impact Zone
Recovery

'8

Log (# Individuals)

    

November 1997
Sampling Dates

 

FIGURE 16. Log (# Individuals) collected with the modiﬁed-Hess sampler at riﬂe sites
upstream and downstream of the industrial eﬂuent (+ Standard Errors), exclusive of the
chironomids and Oligochaetes. N = 5 per site. Overall ANOVA p-values: June 0.0079,
November 0.0010. Sites marked with the same letter are not signiﬁcantly diﬁerent
(Tukeys LSD Method, p > 0.05). See Figure 1 for sampling locations.

30.00 '1
25.00

20.00
I Upstream

15-00 [J lrnpact Zone

10.00 ERecow-ty

Mean # Taxa

5.00

 

 

0.00 L

 

Noveniier 1997
Sanmling Dates

FIGURE 17. Mean # Taxa collected with the modiﬁed-Hess sampler at riﬂe Sites
upstream and downstream of the industrial eﬂuent (+ Standard Errors) N = 5. Overall
AN OVA p—values: June 0.0006, November 0.0029. Sites marked with the same letter are
not signiﬁcantly diﬂerent (Tukeys LSD Method, p > 0.05). See Figure l for sampling
locations.

45

12.00 *

IUpstrean

1:] Impact Zone
IRecowry

# EPT Taxa

 

 

June 1997 Novenlier 1997
Sampling Dates

FIGURE 18. Mean # EPT Taxa collected with the modiﬁed-Hess sampler at riﬂe sites
upstream and downstream of the industrial eﬂuent (+ Standard Errors). N = 5. Overall
ANOVA p-values: June 0.0027, November 0.0004. Sites marked with the same letter are
not signiﬁcantly diﬁ‘erent (Tukeys LSD Method, p > 0.05). See Figure 1 for sampling
locations.

 

 

June 1997 Novenirer 1997
Sampling Dates

FIGURE 19. Mean # Ephemeroptera individuals collected with the modiﬁed-Hess

sampler at riﬂe sites upstream and downstream of the industrial eﬂuent (+ Standard
Errors). N = 5. Overall ANOVA p-values: June 0.0077, November 0.0093. Sites marked
with the same letter are not signiﬁcantly different (T ukeys LSD Method, p > 0.05). See
Figure 1 for sampling locations.

46

350.0 *
.Q
§ 300.0 *
:2
:5 250.0 - . lUpstream
I:
2, 200-0 - a Dlmpathone
a
a 150.0 ‘ ERecovery
E
U

 

 

June 1997 November 1997
Sampling Dates

FIGURE 20. Mean # Trichoptera individuals collected with the modiﬁed-Hess sampler at
riﬂe sites upstream and downstream of the industrial eﬂuent (+ Standard Errors). N = 5.
Overall AN OVA p-values: June 0.2907, November 0.374. Sites marked with the same
letter are not signiﬁcantly diﬂ'erent (Tukeys LSD Method, p > 0.05). See Figure 1 for

sampling locations.

1.200

1.000

0.800 b
I Upstream

0.600 U Impact Zone

11' Diversity

0.400 , E Recovery

   

0.200 .

0.000 1"
June 1997 November 1997
Sampling Dates

FIGURE 21. H’ Diversity of samples collected with the modiﬁed-Hess sampler at riﬂe
sites upstream and downstream of the industrial eﬂuent (+ Standard Errors), exclusive of
chironomidae and oligochaeta. N = 5. Overall AN OVA p-values: June 0.0332,
November 0.0012. Sites marked with the same letter are not signiﬁcantly different
(Tukeys LSD Method, p > 0.05). See Figure 1 for sampling locations.

47

7.00 ‘
6.00 l
5.00 §
4.00
3.00

2.00 4

I Upstream

C] Impact Zone
5 Recowry

Biotic Index

 

0.00 L
June 1997 Noveniier 1997

Sampling Dates

FIGURE 22. Biotic index calculated for samples collected with the modiﬁed-Hess
sampler at riﬂe Sites upstream and downstream of the industrial eﬂuent (+ Standard
Errors), exclusive of oligochaeta and chironomidae. N = 5. Overall ANOVA p-values:
June 0.0539, November 0.1494. Sites marked with the same letter are not signiﬁcantly
diﬂ‘erent (Tukeys LSD Method, p > 0.05). See Figure 1 for sampling locations.

 

50.0]
g 40.0 /.
'5 30.0 .‘ --O--Shredder
é- ‘ —O—-Scraper
5 20-0 +Filteru-Collector
e\° 10.0 +Gatherer—Colleaor
--0---Predator
0.0 -; # -

 

Upstream Impact Zone Recovery

Riﬂe Location

FIGURE 23. Functional feeding group composition of samples collected with the
modiﬁed Hess sampler at riﬂe sites upstream and downstream of the industrial eﬂuent.

Figures are based on presence/absence of taxa. See Figure 1 for sampling locations.

48

 

 

 

a 2.50 ~
:5
E 2.00 _ +Upstream
E -O--lrrpathone
an: 1.50“ "AH-Recovery
E
,q LII)“
0.50 l u I a
Jul-97 Aug-97 Sept-97 Nov-97
Month

FIGURE 24. Log (# Individuals) colonizing ceramic tiles at the three riﬂe sites over the
four colonization periods (:1: Standard Errors). N = 4—5. Overall AN OVA p-values: July
0.0717, August 0.0229, September 0.0332, November 0.1765. See Figure 1 for sampling
locations.

14.00 -
1200 ~
10.00 -
8.00 ~
6.00 .
4.00 «
2.00 —
0.00 I . . .
Jul-97 Aug-97 Sept-97 Nov-97

—O—Upstream
-D—-Inpathone
"*"R600le.

Mean # Taxa

 

 

 

Month

FIGURE 25. Mean # Taxa colonizing ceramic tiles at the three riﬂe sites over the four
colonization periods (:1: Standard Errors). N = 45. Overall ANOVA p-values: July
0.1060, August 0.1392, September 0.2238, November 0.0618. See Figure 1 for sampling
locations.

49

100.0

+Upstream
- D- - Inpact Zone

"Ir-Recovery

% Midges and Worms

 

 

 

FIGURE 26. % Midges and Worms (Oligochaeta) ) colonizing ceramic tiles at the three
riﬂe sites over the four colonization periods (:1: Standard Errors). N = 4—5. Overall
ANOVA p-values: July 0.0004, August 0.0009, September 0.0001, November 0.1502.
See Figure 1 for sampling locations.

8.1!) 1
7.11) r
6.1!) J
5.1!) .
4.1!) ~
3.1!) -
2.1!) 4
Lil) _
0.1!)

# EPT Taxa

 

 

 

FIGURE 27. Mean # EPT Taxa colonizing ceramic tiles at the three riﬂe Sites over the
four colonization periods (i Standard Errors). N = 4—5. Overall ANOVA p-values: July
0.0688, August 0.2242, September 0.1445, November 0.0127. See Figure 1 for sampling
locations.

50

20.0 -

 

 

.3

a

El

.2 16.0

'6

i 12.0 - +upmem
8 - O- - Irrpaet Zorn
a. 8.0 -

g Ir - Recovery
5 4.0 —

.E

G.

“1 0.0 .

 

FIGURE 28. Mean # Ephemeroptera colonizing ceramic tiles at the three riﬂe sites over

the four colonization periods (:1: Standard Errors). N = 4—5. Overall ANOVA p-values:
July 0.0223, August 0.0066, September 0.0838, November 0.0657. Figure l for sampling
locations.

70.0 1
60.0
50.0
40.0
30.0
20.0
10.0

0.0

l

J

—O— Upstream

- D- - Impact Zone

I

l

- - £- - Recovery

1

l

Trichoptera Individuals

 

 

 

Jul-97 Aug-97 Sep-97 Nov-97
Month

FIGURE 29. Mean # Trichoptera colonizing ceramic tiles at the three riﬂe sites over the
four colonization periods (:1: Standard Errors). N = 4—5. Overall ANOVA p-values: July
0.2862, August 0.0581, September 0.2743, November 0.1666. Figure 1 for sampling
locations.

51

CHAPTER TWO
GROWTH AND SURVIVAL OF ST ENACRON HVT ERPUNCTATUM
(EPHEMEROPTERA: HEPTAGENIIDAE) EXPOSED
TO AN INDUSTRIAL EFFLUENT

INTRODUCTION

A critical component of water resource management is the prediction or detection
of adverse effects of stressors on aquatic ecosystems (Rosenberg and Resh 1993, Davis
and Simon 1995). Historically, much of the focus of water resource assessment and
management programs was on establishing water quality standards for physico-chemical
variables, and controlling point-sources of pollution (Karr 1991, Yoder 1995).
Considerable emphasis was placed on the development of acute and chronic laboratory
bioassays for a suite of pollutants (Malins 1989, Maltby and Calow 1989). The ultimate
goal of these laboratory bioassays was to predict the impacts of pollutants on entire
ecosystems. Despite strict regulation of point-source pollution by state and federal
regulatory agencies in the past few decades, there is still potential for these to
compromise the biological integrity of aquatic ecosystems. This is due to the fact that
laboratory bioassays will rarely provide an accurate and realistic prediction or assessment
of the eﬂ‘ects of a point-source of pollution on ecosystem quality (Cairns 1983, Cairns
and Pratt 1989, Maltby and Calow 1989). This is because they often fail to address
potential synergistic interactions of the toxicant with physical, chemical, or biological
modifying factors in the environment (Sprague 1995). Bioassays conducted within
natural systems (in situ) have become popular in the last 15 years (Winger et al. 1984,
Munawar and Munawar 1987, Chappie and Burton 1997). Test conditions are not

necessarily meant to emulate the natural system, but are meant to enclose the organisms

52

withinadeﬁnedspacewheretheycanbeexposed to some ofthedynamic environmental
variables that would often go unaccounted for in laboratory studies. In situ bioassays are
particularly well suited for determining the impact of point-source pollutants on
representative aquatic Species (Muliss et al. 1996), and ideally community structure
(Cairns and Pratt 1989).

An example of a point-source pollutant that may have negative impacts on aquatic
ecosystems yet is largely unregulated is cooling water eﬂuents. We studied the effect of
an industrial cooling-water eﬂuent on benthic macroinvertebrate community structure.
The eﬂuent is thermally constant, ranging ﬁom 17 to 25 °C throughout the year. In the
winter months the eﬂuent raises the stream temperature 8 to 12°C above ambient. The
eﬂuent contains high levels of ferrous iron, which is oxidized to ferric iron and then
hydrolyzed to ferric hydroxide, which forms a yellow-brown precipitate (‘yellowboy”)
that settles out onthe substratum. It isknownto inhibitthe growthofbenthicalgae
(especially diatoms and green algae) (Sheldon and Skelly 1990, Wellnitz and Sheldon
1995), depress macroinvertebrate diversity by interfering with feeding and respiration
(Koryak et al. 1972, Rasmussen and Lindegaard 1988, Wellnitz et al. 1994), and has
consequences for higher trophic levels such as ﬁsh (Letterman and Mitsch 1978).

Neither the thermal nature of the eﬂuent nor the levels of iron exceed water
quality standards. However, there could be an additive effect of temperature and ferric
hydroxide in the eﬂuent, or there may be some other factor that acts in synergy with
these factors to inﬂuence the biota. Mayﬂies (Ephemeroptera) were conspicuously scarce
or absent in riﬂes immediately downstream of the industrial eﬂuent (Chapter One),

though were abundant at sites ﬁrrther upstream and much farther downstream. Their

53

scarcity caused particular concern about the effect that the industrial eﬂuent was having
on the biological integrity of Portage Creek. We conducted in situ bioassays to try to
assess the speciﬁc eﬁ‘ect of the eﬂuent on the mayﬂy Stenacron interpunctatum (Family
Heptageniidae). Our speciﬁc objectives were:
(1) Compare mayﬂy growth and mortality withinand downstream ofthe
industrial eﬂuent versus and upstream control (Experiment 1),
(2) Try to separate the effects of temperature and food quality on rmyﬂy growth

and mortality (Experiment 2).

METHODS

STUDY SITE: Portage Creek is a third-order tn'butary of the Kalamazoo River in
Kalamazoo County, Michigan (Figure 1). Its 52-square kilometer drainage area is a
glacial outwash plain, with up to 350 feet of sand and gravel overlying a shale bedrock
valley. It originates from seeps and wetlands, and travels north for about 15 kilometers to
itsconﬂuencewiththe Kalamazoo Riverinthe cityofKalamazoo. Landuse inthe
watershed in primarily agricultural, urban, and industrial. The industrial eﬂuent that is
the focus of this study is located in the city of Portage (Figure 3A, 3B). The eﬂuent is
non-contact cooling water utilized by a large pharmaceutical company. Though not
accessible to the general public, people can rent canoes upstream of the industrial eﬂuent
and canoe past it. There was concern over the yellow-brown precipitate that covered the
substratum within and downstream of the discharge channel, and a more general concern

that such a conspicuous source of pollution was detracting ﬁ'om the aesthetic beauty of

54

the creek. This prompted the Michigan Department of Environmental Quality to initiate

an investigation into the eﬂ'ects of this eﬂuent on Portage Creek.

CHOICE or TEST ORGANISM: Results from comparative ﬁeld sampling (Chapter One)
showthat mayﬂies are scarce or absent within and immediately downstream (”1000
meters) of the industrial eﬂuent. The scarcity of the mayﬂy Stenacron interpunctatum is
particularly noticeable because it is abundant at riﬂes upstream and well downstream
(>1500 meters) of the industrial eﬂuent. Late-instars are readily distinguished from
similar mayﬂies in the genus Stenonema by their pointed gills and gray/black coloration
pattern. Behavioral studies indicate that it is an oppOrtunistic collector-gatherer
(McShaﬁ‘rey and McCaﬁ‘erty 1986). It obtains most of its nutrition by brushing loosely
adhered material ﬁom the surfaces of stones. S. interpunctatum was chosen because of
its abundance in Portage Creek, its apparent sensitivity to the physico-chemical
environment immediately downstream of the industrial eﬂuent (based on its

scarcity/absence), its nutritional requirements, and ease of handling and identiﬁcation.

EXPERIMENT ONE: The ﬁrst experiment was designed to determine if S. intetpunctatum
was capable of surviving and growing within and downstream of the discharge channel,
compared to an upstream control. The experimental design is illustrated in Figure 3.
Twelve growth chambers were partially ﬁlled with gravel (1-3 cm diameter) ﬁ'om the
mayﬂy collection site, along with eight 4 cm2 ceramic tiles previously colonized with
periphyton in Augusta Creek (Kalamazoo County, Michigan). Organisms were collected

at the Livery Canoe Launch located about 1500 meters upstream of the discharge

55

channel, where a short riﬂe section supports abundant mayﬂies (Figure 1). Ten mayﬂies
were placed into each of the growth chambers; twenty individuals were set aside to
determine initial size and weight.

Four growth chambers were placed into each of three acclimation chambers (50-
gallon buckets); two were transported to the downstream test sites, and the third remained
at the collection (control) site (Figure 2). The acclimation chambers were placed at their
respective test Sites until the water temperature equilibrated with the stream temperature,
which took about 3-4 hours. Two growth chambers were then placed into each of two
cages suspended in the stream, for a total of 4 growth chambers (40 individual mayﬂies)
per site. The growth chambers were leﬁ in the stream for six days (May 16 — May 22
1997). Daily measurements of temperature (min/max thermometer) and ﬂow velocity
(Marsh-McBirney ﬂowmeter) were taken. At the conclusion of the experiment, all

mayﬂies were killed, measured (total length, head capsule width), and weighed (dry

weight).

EXPERIMENT Two: The second experiment was designed to examine the growth and
survival of S. interpunctatum provided with different food. types at different Site
locations. This was a 2-way factorial, with two levels of the location treatment
(upstream, discharge channel), and two levels of the food treatment (good quality food,
poor quality food). The experimental design is illustrated in Figure 3. The good quality
food consisted of gravel (1 to 3 cm diameter) and 2.5 cm2 ceramic tiles that had been
placed inariﬂeupstreamofthe industrialeﬂuent, andwascoatedwithabioﬁlmhigh

in diatoms, ﬁlamentous green algae, and FPOM. The poor quality food consisted of

56

gravel (1 to 3 cm. Diameter) and 2.5 cm2 ceramic tiles that had been placed in the
discharge channel, and was coated with a thick gelatinous matrix of ferric hydroxide, a
blue-green bacterium, and perhaps other inorganic substances.

Each of the 12 growth chambers used in Experiment One were divided into two
chambers, andeachhalfwas partially ﬁlled witheither good quality food orpoorquality
food. Due to an unusually warm spring, the ﬁrst cohort of S. interpunctatum had already
emerged from upstream sections of Portage Creek. Thus, test animals were collected at a
riﬂe section about 3 kilometers downstream of the industrial eﬂuent (Kilgore Road)
(Figure 1). These individuals probably represented the second cohort of the summer,
since this site is strongly inﬂuenced by the thermal nature of the industrial eﬂuent, and
growth/development of mayﬂies is accelerated at these sites. Eight animals were placed
into each halfof the grth chamber (food treatment), for a total of 16 animals per
chamber. No measures were taken to acclimate the animals to the temperature at their
respective test sites because the temperature diﬁerences were small (‘24 °C). Two
growth chambers were placed into each of three cages suspended in the stream, for a total
of 6 growth chambers, and 6 replicates per food treatment. The growth chambers were
left in the stream for ﬁve days (May 18 — May 23 1998). Daily measurements of

temperature (min/max thermometer) were taken. At the conclusion of the experiment, all

rmyﬂies were killed, and weighed (dry weight).
RESPONSE VARIABLES: In the ﬁrst experiment, total length (mm), head capsule width

(mm), and dry weight (grams) were determined for each of the test animals. All three

were very well correlated with each other, so in the second experiment only dry weights

57

(grams) were determined. For uniformity, only dry weights were used in the statistical
analysis to determine a grth response. Grth was deﬁned as:

Growth = Final Dry Weight — Initial Dry Weight
For survival determination, all missing animals were presumed dead. Individuals that
had emerged within the growth chambers were considered survivors, but were not
included in the grth analysis.

For Experiment 1, I used a one-factor ANOVA to test the eﬁ‘ect ofsite on mayﬂy
growth and survival For Experiment 2, I used a two-factor ANOVA to test the effect of
site, food quality, and the interaction on mayﬂy growth and survival. I used the SAS©
statistical software program for the analyses. A test was considered signiﬁcant of the p-
value 5 0.05. Further statistical analysis, with a more accurate portrayal of
randomization, blocking, and treatment effects were only considered if the overall

AN OVA was signiﬁcant.

RESULTS
EXPERIMENT ONE: The temperature and ﬂow data are presented in Table 1. Water
temperatures were signiﬁcantly warmer in the discharge channel and the downstream test
site, as shown by degree-day accumulations. Therefore, growth was corrected for
degree- day accumulations (Mean Growth / ° Day). Summary data for the
growth/survival response is presented in Table 2, and presented graphically in Figures 4-
6.

There were no statistically signiﬁcant results ﬁom Experiment 1 (where

signiﬁcance is judged as a p—value less than 0.10). S. interpunctatum exhibited the

58

greatest absolute growth in the discharge channel (0.00219 grams), and the least growth
at the control site (0.00113 grams) (Figure 4), though the overall ANOVA for mean
growth was not signiﬁcant (p-value 0.5918). S. interpunctatum exhibited the greatest
growth at the downstream site when corrected for degree-day accumulation, and grew the
least at the upstream control site (Figure 5), though the overall ANOVA for mean growth
corrected for temperature is not signiﬁcant (p-value 0.8677). There was no trend in

survival between the three sites, and the overall ANOVA p-value was 0.6019 (Figure 6).

EXPERIMENT Two: The temperature data is presented in Table 1. Flow velocity was not
measured in Experiment 2, but was similar between the two sites. Summary data for the
growth/survival response is presented in Table 3. Figure 7 shows the interaction of food
and location on mayﬂy growth. Results from the two-factor AN OVA indicate few
statistically signiﬁcant results ﬁom Experiment 2 (Table 3) (where signiﬁcance is judged
as a p-value less than 0.10). Food quality appears to have the most obvious effect on
growth (p-value 0.1332), and growth normalized for degree days (p-valuc 0.0863).
Location, and the location x food interaction appear to have little effect on growth. There

was no difference in survival between the different treatments.

DISCUSSION

Stenacron interpunctatum grew faster within and downstream of the discharge
channel than upstream, and exhibited no survival response to the eﬂuent. The higher
temperatures are thought to have caused the slightly positive growth response.

Temperatureisperhapsthemostcriticalfactorinthelifecycleofmostaquaticinsects

59

(V annote and Sweeney 1980, Ward and Stanford 1982, Sweeney 1984). Many studies
have shown that warmer temperatures generally accelerate the life cycles of aquatic
insects by decreasing egg hatching time (Newell and Minshall 1978, Hurnpesch and
Elliot 1980), and accelerating larval grth and emergence (Hurnpesch 1981, Sweeney
and Vannote 1984, 1986). Development continues to increase with temperature until a
thermal threshold is reached (Sweeney 1984), and though this is species dependent,
aquatic insects can generally tolerate temperatures of up to 30-35 °C. The maximum
recorded temperature during the growth/survival experiments was 25 °C, which is likely
still within the optimal growth range for Stenacron interpunctatum. Flowers and
Hilsenhoff (1978) report that this species is often abundant in slower currents of small
eutrOphic streams of southern Wisconsin, and has an unusually high tolerance for silty
enviromnents. McCaﬂ'erty and Huﬂ‘ (1978) descn'be a complex life cycle for this species
inlndiana,withemergenceperiodsinthelate spring, summer, andearlyfall. Theselife
history studies suggest that S. interpunctatum has a fairly high tolerance for warmer
temperatures.

Food quantity and quality are also critical factors in the life cycle of aquatic
invertebrates (Cummins and Klug 1979, Anderson and Ormmins 1979, Sweeney 1984).
Food quantity and quality have been shown to aﬂ‘ect larval development time, size at
maturity, and fecundity of aquatic insects (Colbo and Porter 1979, Collins 1980, Webb
and Menitt 1987). In this study, S. interpunctatum grew poorly when provided with a
low quality food source (a gelatinous ferric hydroxide/bacterial bioﬁlm that developed
within the discharge channel), suggesting that food quality rather than temperature or

water quality are inhibiting some pollution-intolerant taxa from inhabiting riﬂes within

60

and immediately downstream of the eﬂuent. Other studies have demonstrated a general
reduction in macroinvertebrate community composition in areas with high dissolved iron
and ferric hydroxide (Koryak et al. 1972, Greenﬁeld and Ireland 1978, Letterman and
Mitsch 1978, Rasmussen and Lindegaard 1988, Wellnitz et al. 1994), and have often
attnhuted this to a reduction in the quantity and quality of food, destabilization of
substrate by ﬂocculent iron, or direct toxic eﬂects of Fe ions and/or the closely associated
iron-depositing sheathed bacteria Leptothrix spp.. Wellnitz et aL (1994) looked at
substrate choice, weight gain, and survival of selected aquatic insects on ferric I
hydroxide/Leptothrix ochracea substrates. Two of the three heptageniid myﬂies
(Epeorus, Heptagenia) preferred substrates ﬁ'ee of the iron, and Stenonema showed no
preference. All three genera showed a greater mortality within the bloom of iron-
depositing bacteria, and though all three ingested the iron bacteria, only Heptagenia
gained weight after 10 days. They found that Leptothrix ochracea encrusted the tracheal
gills of these mayﬂies, and appeared to hamper gill motion.

The quality and quantity of food resources is temperature-dependent. For
instance, temperature can inﬂuence the quantity of periphytic algae (McIntire and
Phinney 1965), or the microbiota that colonize detrital material and enhance its
nutritional quality (Webster and Benﬁeld 1986). So determining the relative importance
of temperature and food resources on aquatic invertebrate population and cormnunity
ecology is problematic because temperature affects both the quantity and quality of food,
and the ingestion and assimilation of the food by consumers (Sweeney 1984). In this
study there are several possible combined eﬁ‘ects of temperature and ferric hydroxide

precipitation on macroinvertebrates. Elevated temperature will increase respiratory

6l

dermnd, and combined with gill clogging or abrasion by particulate iron could have a
more detrimental effect than either one alone. Elevated temperature will increase
metabolic demand, and combined with poor food quality due to particulate iron in
suspension or on the stream substrate could also have a greater effect than either one
alone. In addition, temperature and iron could act in synergy with other environmental
variables. Thermal stress can be additive with toxic stress, especially when it is
accompanied by low oxygen (Sprague 1995). Thermal stress accompanied by changes in
pH or dissolved oxygen can also have additive effects. Oxygen and pH can also act in
synergy with the toxic stress of dissolved iron, suspended iron (gill abrasion), and iron-
coated substrates (poor nutrition, gill abrasion). The physiological, biochemical, or
behavioral traits of a species will determine its relative sensitivity to a particular stressor
(Sparague 1995). Life stage or size of the organism is a particularly important in
determining sensitivity. In this experiment, late-instar individuals were used, and these
larger, older individuals will generally be less sensitive to the adverse effects of
temperature or food quality than early-instar individuals.

These experiments were not designed to address all of the possible synergistic
effects of iron and temperature with other environmental variables. The experiments
were primarily designed to try to understand the why mayﬂies, especially Stenacron
interpunctatum, were scarce or absent in suitable habitats immediately downstream of the
industrial efﬂuent. The results are not conclusive. If the only effect of temperature
were to accelerate larval development, then S. interpunctatum might be expected to be
found downstream of the efﬂuent, but perhaps display an asynchronous emergence

pattern. However, benthic samples were taken during all times of the year, and mayﬂies

62

were never more than scarce. One might expect early instars to be more sensitive to the
warmer thermal regime, but this is unlikely given the multivoltine life cycle of S.
interpunctatum, and the results of previous studies on similar species. Based on
Experiment Two, ferric hydroxide precipitation and deposition perhaps better explains
the scarcity of mayﬂies in the discharge channel and riﬂes immediately downstream.
This result is supported by the studies that have examined macroinvertebrate community
structure in streams impacted by iron compounds.

Clearly, water quality standards are not protecting aquatic life in Portage Creek.
Comparative ﬁeld sampling of the benthic macroinvertebrate communities show that the
riﬂes immediately downstream of the eﬂuent consistently support fewer number of
taxa, and a larger proportion of pollution-tolerant taxa. Yet the industry consistently
meets eﬂuent quality standards for all physico-chemical criteria. Further studies in this
system with a broader range of taxa, a broader range of life stages, and exposure to a
broader range of environmental conditions (seasonal effects) might help to better
understand the scarcity of mayﬂies in riﬂes downstream of the industrial eﬂuent. This
would greatly enhance our understanding of the overall effect of the industrial eﬂuent on
the biological integrity of Portage Creek, and allow water resource managers to hone

their restoration or remediation eﬂ‘orts.

63

LITERATURE CITED

Anderson, NH, and K.W. Cummins. 1979. Inﬂuences of diet on the life histories of
aquatic insects. Journal of the Fisheries Research Board of Canada 36:335-342.

Cairns, J. Jr., and J.R. Pratt. 1989. The scientiﬁc basis of bioassays. Hydrobiologia
188/ 189:5-20.

Cairns, J .Jr. 1983. Are single species toxicity tests alone adequate for estimating
environmental hazard? Hydrobiologia 100:47-57.

Chappie, D.J., and A. Burton Jr. 1997. Optimization of in situ bioassays with Hyalella
azteca and Chironomus tentans. Environmental Toxicology and Chemistry 16:559-
564.

Colbo, MH., and G.N. Porter. 1979. Eﬂ'ects of the food supply on the life history of
Simuliidae (Diptera). Canadian Journal of Zoology 57:301-306.

Collins, NC. 1980. Developmental responses to food limitation as indicators of
environmental conditions for Ephydra cinerea Jones (Diptera). Ecology 61:650-661.

Cummins, K.W., and M.J. Klug. 1979. Feeding ecology of stream invertebrates.
Annual Review of Ecology and Systematics. 10:147-172.

Davis, W.S., and T.P. Simon. 1995. Biological Assessment and Criteria: Tools for
Water Resource Planning and Decision Making. Lewis Publishers.

Flowers, R.W., and W.L. Hilsenhoﬂ'. 1978. Life cycles and habitats of Wisconsin
Heptageniidae (Ephemeroptera). Hydrobiologia 60: 159-171.

Greenﬁeld, J.P., and M.P. Ireland. 1978. A survey of the macrofauna of a coal-waste
polluted Lancashire ﬂuvial system. Environmental Pollution 16: 105-122.

Humpesch, U.H. 1981. Eﬂ‘ect of temperature on larval growth of Ecafyonurus dispar
(Ephemeroptera: Heptageniidae) from two English lakes. Freshwater Biology
1 1:441-457.

Humpesch, U.H., and J.M. Elliot. 1980. Effect of temperature on the hatching time of
eggs of three Rhithrogena spp. (Ephemeroptera) from Austrian streams and an
English stream and river. Journal of Animal Ecology 49:643-661.

Karr, J.R. 1991. Biological integrity: a long-neglected aspect of water resource
management. Ecological Applications 1:66-84.

Koryak, M., MA. Shapiro, and J.L. Sykora. 1972. Riﬂe zoobenthos in streams
receiving acid mine drainage. Water Research 6: 1239-1247.

Letterman, RD, and W.J. Mitsch. 1978. Impact of acid mine drainage on a mountain
stream in Pennsylvania. Environmental Pollution 17:53-73.

Malins, DC. 1989. The use of environmental assays for impact assessment.
Hydrobiologia 1 88/1 89:87-91 .

Maltby, L., and P. Calow. 1989. The application of bioassays in the resolution of
environmental problems: past, present and future. Hydrobiologia 188/189z65-76.

McCaﬂ‘erty, W.P., and BL. Huff Jr. 1978. The life cycle of the mayﬂy Stenacron
interpunctatum (Ephemeroptera: Heptageniidae). Great Lakes Entomologist 11:209-

216.

McIntire, J .W., arxl H.K. Phinney. 1965. Laboratory studies of periphyton production
and cormnunity metabolism in lotic environments. Ecological Monographs 35:23 7-
258.

McShaﬂiey, D., and WP. McCaﬂerty. 1986. Feeding behavior ofStenacron
interpunctatum (Ephemeroptera: Heptageniidae). Journal of the North American
Benthological Society 5:200-210.

Muliss, R.M., D.M. Revitt, and R.B.E. Schutes. 1996. A statistical approach for the
assessment of the toxic inﬂuences on Gannarus pulex (Amphipoda) and Asellus
aquaticus (Isopoda) exposed to urban aquatic discharges. Water Research 5:123?-
1243.

Munawar, M., and LP. Munawar. 1987. Phytoplankton bioassays for evaluating toxicity
of in situ sediment contaminants. Hydrobiologia 149:87-105.

Newell, KL, and G.W. Minshall. 1978. Effect of temperature on the hatching time of
T ricorythodes minutus (Ephemeroptera: Tricorythidae). Journal of the Kansas
Entomological Society 51:504-506.

Rasmussen, K., and C. Lindegaard. 1988. Effects of iron compounds on
macroinvertebrate communities in a Danish lowland river system. Water Research

22:1101-1108.

Rosenberg, D.M., and V.H. Resh. 1993. Freshwater biomonitoring and benthic
macroinvertebrates. Clmpman & Hall. 488 pages.

Sheldon, SR, and D.K. Skelly. 1990. Diﬁerential colonization and grth of algae and
ferromanganese-depositing bacteria in a mountain stream. Journal of Freshwater
Ecology 5:475—485.

Sprague, J.B. 1995. Factors that modify toxicity. Pages 1012-1051 In: Fundamentals of
Aquatic Toxicology (G.M. Rand, editor). Taylor & Francis.

65

Sweeney, B.W. 1984. Factors inﬂuencing life history patterns of aquatic insects. Pages
56-100 In: The Ecology of Aquatic Insects (V .H. Resh and D.M. Rosenberg, editors).

Praeger Publishers.

Sweeney, B.W., and KL. Vannote. 1984. Inﬂuence of food quality and temperature on
the life history characteristics of the parthenogenetic mayﬂy, CIoeon triangulifer.
Freshwater Biology 14:621-630.

Sweeney, B.W., and KL. Vannote. 1986. Growth and production of a stream stoneﬂy:
inﬂuences of diet and temperature. Ecology 67:1396-1410.

Vannote, KL, and B.W. Sweeney. 1980. Geographic analysis of thermal equilibria: a
conceptual model for evaluating the effect of natural and modiﬁed thermal regimes
on aquatic insect communities. The American Naturalist 115:667-695.

Ward, J.V., and J.A. Stanford. 1982. Thermal responses in the evolutionary ecology of
aquatic insects. Annual Review of Entomology 27:97-117.

Webb, K.M., and R.W. Merritt. 1987. The inﬂuence of diet on the growth of Stenonema
vicarium (Walker) (Ephemeroptera: Heptageniidae). Hydrobiologia 153:253-259.

Webster, J .R, and BF. Benﬁeld. 1986. Vascular plant breakdown in freshwater
ecosystems. Annual Review of Ecology and Systematics 17:567-594.

Wellnitz, T.A., and SP. Sheldon. 1995. The effects of iron and manganese on diatom
colonization in a Vermont stream. Freshwater Biology 34:465-470.

Wellnitz, T.A., K.A. Grief, and SP Sheldon. 1994. Response of macroinvertebrates to
blooms of iron-depositing bacteria. Hydrobiologia 281:1-17.

Winger, P.V., M.J. Imlay, W.E. McMillan, T.W. Martin, J Takekawa, and W.W.
Johnson. 1984. Field and laboratory evaluation of the inﬂuence of copper—diquat on
apple snails in southern Florida. Environmental Toxicology and Chemistry 3:409-
424.

Yoder, CO. 1995. Policy issues and management applications of biological criteria.
Pages 327-343 In: Biological Assessment and Criteria: Tools for Water Resource
Planning and Decision Making (W.S. Davis and T.P. Simon, editors). Lewis
Publishers.

66

TABLE 1. Temperature and ﬂow data for in situ bioassay with the mayﬂy Stenacron
interpunctatum. Experiment 1: May 16 - May 22 (6 days), 1997, Experiment 2: May 18-
May 23 (5 days), 1998. Degree Days = 2 daily mean temperatures > 0° for duration of
experiment. Velocity not measured in Experiment 2.

 
 

 

 

EXPERIMENT 1

Flow Velocity (m/s) 0.22 0.22 0.25
Mean Temperature (° C) 12.2 21.3 16.1
Min/Max(°C) 10/15 14/24 13/19
Degree Days (° D) 73.2 127.8 96.6
EXPERIMENT 2

Mean Temperature (° C) 16.5 19.3

Min/Max(° C) (12/20) (15/24)

Degree Days (° D) 82.5 96.5

 

TABLE 2. Summary data for the in situ bioassay with the mayﬂy Stenacron
interpunctatum: Experiment 1. Dates: May l6-May 22 (6 days), 1997. Mean Growth: A
Dry Weight. The p-values for the overall AN OVA are indicated; tests are considered
signiﬁcant if the p-value < 0.05. See Figures 4-6 for graphical summary, and Table 1 for
temperature/ﬂow data.

 

   

 

 

 

7 ._ STATISTIC ‘ - 6
MEAN GROWTH (BY SITE) 0.5918
Upstream 1.13E-03 8.32E-04
Discharge Channel 2.19 E - 03 1.29 E - 03
Downstream 2.10 E - 03 2.42 E - 03
MEAN GROWTH/DEGREE DAY (BY SITE) 0.8677
Upstream 1.55E-05 1.14E-05
Discharge Channel 1.71 E - 05 1.00 E - 05
Downstream 2.18E-05 2.50E-05
MEAN SURVIVAL (BY SITE) 0.6019
Upstream 62.50 20.62
Discharge Channel 65.00 5.77
Downstream 53.33 15.28

 

67

TABLE 3. Summary data for the in situ bioassay with the mayﬂy Stenacron

interpunctatum: Experiment 2. Dates: May l8-May 23 1998 (5 days). The p-Vdues for

the overall ANOVA are indicated; tests are considered signiﬁcant if the p-value < 0.05.
GROWTH (LOCATION)

 

 

Upstream 3.62 E — 05 9.67 E — 04
Discharge 6.99 E — 04 9.87 E — 04
GROWTH (FOOD)
Upstream Food 6.92 E — 04 9.19 E — 04
Discharge Food 4.35 E - 05 1.04 E — 03
GROWTH (LOCATION x FOOD)
Upstream x Upstream Food 2.91 E — 04 9.20 E - 04
Upstream x Discharge Food -2.18 E - 04 1.03 E — 03
Discharge x Upstream Food 1.09 E — 03 7.91 E — 04
Discharge x Discharge Food 3.05 E — 04 1.07 E — 03
PERCENT SURVIVAL (LOCATION)
Upstream 62.5 10.7
Discharge 58.3 16.3
PERCENT SURVIVAL (FOOD)
Upstream Food 62.5 13.1
Discharge Food 58.3 14.4
DEPENDENT VARIABLE SOURCE P-VALUE
GROWTH Location 0.1978
Food . 0.1332
Location x Food 0.9284
SURVIVAL Location 0.8296
Food 0.8096
Location x Food 0.8385

 

68

UPSTREAM RIFFLE: ceramic tiles
conditioned here prior to Experi-
ment 2

   
 
  
    

RAILROAD BRIDGE RIFPLE: UPSTREAM

mayﬂy collection site and
upstream test site for Experi-
ment 1.

UPSTREAM TEST SITE: Experiment 2

DISCHARGE CHANNEL

DOWNSTREAM TEST SITE: Experiment 1

DOWNSTREAM RIFFLE:
mayﬂy collection Site for

ExPenmem 2' DOWNSTREAM

FIGURE 1. Map of Portage Creek. ﬁ'om Portage Police Station (Upstream Riﬂe) to
Milham Park (Downstream Riﬂe), showing locations used during the course of the mayﬂy
growth experiments. Scale: 1 inch = " 1000 meters. See text for details.

69

 

 

Figure 2. Elements of the mayﬂy grth experiment. A) Ceramic tiles placed in the
discharge channel. B) Ceramic tiles after ~1 month in the discharge channel,
displaying a heavy accumulation of ferric hydroxide and bacteria. C) A pair of growth
chambers used in Experiment One; one is open to show the arrangement of food tiles.
D) A growth chamber used in Experiment Two, showing the divider used to separate
the chamber into two compartments. E) A pair of growth chambers suspended within
the cage and immersed m the stream. F) All growth chambers (6 chambers, 3 cages)
at the industrial eﬂuent outfall during Experiment Two.

70

 

 

E 1
FOOD TYPE: High SEESMENT I l l l
L : ' -
StfeCaAITIIONS Upstream, Discharge Channel, Down I l I I
[anﬁllizgigzﬁvzﬁncgeafbers (2 cages) per srte, 10 I I I I

 

 

 

 

 

 

 

 

and Discharge (ferric hydroxide and bacteria).
LOCATIONS: Upstream, Discharge Channel
REPLICATIONS: 6 chambers (3 cages) per site, 2
compartments (food types) per chamber, 8 may-
ﬂies per compartment.

EXPERIMENT 2
FOOD TYPES: Upstream (diatoms and green algae), E
El

 

 

 

 

 

 

 

 

FIGURE 3. Experimental design for the mayﬂy grth experiments. See Figure 1 for a
map of test locations, and Figure 2 for phtographs of ceramic tiles, growth chambers,
cages, and the experiment in progress.

71

0.0040 -
0.0035 -
0.0030 -
0.0025 «
0.0020 <
0.0015 7
0.0010 —
0.0005 j
0.0000

Mean Growth (grams)

 

 

FIGURE 4. Mean Growth (grams) of caged Stenacron interpunctatum for the in situ
growth/mortality experiment conducted May 16 — May 22, 1997 (+ Standard Errors). N
(cages) = 4 (upstream), 4 (discharge), 3 (downstream). Overall ANOVA not signiﬁcant
(p—value 0.5918). See Table 2 for summary data, and Figure l for site locations.

0.00005 ‘

0.00004 “

Mean Growth/Degree Day

 

 

 

Upstream Discharge Downstream

Location

FIGURE 5. Mean growth (g) of caged Stenacron interpunctatwn, corrected for total
degree-day accumulation at each location (+ Standard Deviation). In situ
growth/mortality experiment conducted May 16 - May 22, 1997. N (cages) = 4
(upstream), 4 (discharge), 3 (downstream). Overall ANOVA not signiﬁcant (p-value
0.8677). Degree Days = 2 daily mean temperatures > 0° for 6-day experiment. See
Table 2 for summary data, and Figure 1 for site locations.

72

% Survival

 

 

 

 

Discharge Downstream
Location

FIGURE 6. Percent survival of caged Stenacron interpunctatum for the in situ
growth/mortality experiment conducted May 16-May 22 1997 (+ Standard Deviation). N
(cages) = 4 (upstream), 4 (discharge), 3 (downstream). Overall ANOVA not signiﬁcant
(p-value 0.6019). See Table 2 for summary data, and Figure 1 for site locations.

Mean Growth (grams)

 

 

FIGURE 7. Mean growth (grams) of caged Stenacron interpunctatum exposed to different
test locations and food sources (+ Standard Errors). N (chambers) = 6 for each
combination. Dates: May 18 — May 23 1998. Two-ANOVA not signiﬁcant (p-values:
Location 0.1978, Food 0.1332, Location x Food 0.9284). See Table 2 for summary data,
and Figure l for site locations.

73

APPENDIX ONE

Qualitative list of macroinvertebrate taxa collected in Portage Creek by various sampling
methods over the course of the study (June 1996 — June 1998). 1: HD Upstream, 2: HD
Discharge, 3: HD Downstream, 4: Hess Upstream, 5: Hess Impact Zone, 6: Hess
Downstream Recovery, 7: Tile Upstream, 8: Tile Impact Zone, 9: Tile Downstream
Recovery, 10: Qualitative (entire stream). * Denotes iron deposition.

   

 

 

PORI F ERA O
CNIDARIA
Hydra Sp. 0 o -
ECTOPROCI‘A
Cristatella mucedo -
TURBELLARIA
Dugesia tigrina 0 0 o 0 o o o -
NEMATODA - . . . . .
OLIGOCHAETA
Unident. Oligochaeta - o o o o o o o o
Bothrioneurum sp.
Rhyacodrilus sp.
Tubifex sp. -
Potamothrix sp.
Tubiﬁcidae w/o caps 0
Ophidonais sp.
Nais sp.
Dero sp.
Pristina sp.
Sparganophilus tamesis - -
Lumbriculus variegatus
Aelosoma sp. 0
HIRU DIN EA
Erpobdellidae
Helobdella stagnalis 0 o 0
Helobdella triserialis 0 0 o
Batrachobdella phalera
Placobdella papillifera -
PELECYPODA
Pisidium sp. 0 0 o o o -
Sphaerium striatinum 0 0
Alasmidonta calceolus
Strophitus undulatus
Elliptio dilatata
Lasmigona complanata
Actinonais ellipsiformes
Pyganodon grandis

74

APPENDIX ONE (CONTINUED)

 

 

Corbiculaﬂuminea
GASTROPODA
F errissia rivularis
Helisoma sp.
Planorbella 5p.
Armiger crista
Valvata tricarinata
Valvata bicarinata
Valvata sincera
F ossaria sp.
Goniobasis Iivescens
Physa sp.
Amnicola limosa
Viviparus georgianus
Campeloma decisum
HYDRACARINA
AMPHIPODA
Gammarus pseudolimnaeus
Hyalella azteca
ISOPODA
Caecidotea intermedius
DECAPODA
Orconectes propinquus
COLLEMBOLA
EPHEMEROPTERA
Stenonema sp.
Stenonema terminatum
Stenonema exiguum
Stenonemaﬁlscum
Stenacron interpunctatum
Baetis Ievitans
Baetis brunneicolor
Baetis sp.
Pseudocloeon 5p.
Ephemera simulans
Ephemeralla lata
Caenis sp.
Leptophlebia sp.
Paraleptophlebia sp.
Tricorythodes sp.
ODONATA

75

APPENDIX ONE (CONTINUED)

x5. I'- W .. - - . I; -- . . _ - .- - rv —.- .. r-- - ,.. . , ._, . - _. _‘ ~ . .. . , . . _. N
‘ - f, r-g"",‘f‘”"‘vw 1:42;... , 7 ._ I 'r ' ‘ _§r:i2§r '3 .‘.'":-.«‘.'!'.,-.f‘..‘,‘v.' 5'5". 4:. _"- 313.31; ' ‘1'} 1,. 'IJ'. E-i-‘W‘m-A f' 5y??- Iezrw 195715”- _.,.‘ w ‘ .. ’33": W v..- -I ,- 's' v.) ‘ .
, . 7‘ .' . I: ' I ‘ '. . ; ' ‘ ~. . 1 ,w , .. . * . I ‘ ., I < ,- .

“‘ ‘ ‘ ‘ 1 - . ‘-r‘-..'... , a" - . I_'_ '1, . . :.“~_-_- ‘ ., .* ~ u... < .A 1., mm »

 

Calopteryx maculatum o o o o
Aeshna sp. 0 o
Gomphus sp. .
Enallagma sp. .
Argia sp. . . ,
PLECOPTERA
T aeniopteryx sp. 0 . . .
HEMIPTERA
Plea sp.
Ranatraﬁzsca
Aquarius sp.
Gerris sp.
Belostomaﬂumineum
Notonecta sp.
MEGALOPTERA
Sialis sp. 0 -
Nigronia serricornis
TRICHOPTERA
Hydropsyche sp. 0 o -
Cheumatopsyche sp. 0 -
Helicopsyche borealis
Oecetis sp. -
Nectapsyche diarina - -
Lype diversa - -
Brachycentrus numerosus
Psychomyiaﬂavida
Pycnopsyche guttifer -
Pycnopsyche luculenta
Neophyllax sp.
Glossosoma nigrior . .
Hydroptila sp. 0 o o o . o
Molanna sp.
Phryganea sp.
COLEOPTERA
Macronychus glabratus 0
Stenelmis sp.
Optioservusfastiditus -
Dubiraphia sp. 0 -
Ancyronyx variegata - o
Ectopria nervosa
Peltodytes sp. 0 o
Laccophilus sp.
A gabus sp.

 

76

APPENDIX ONE (CONTINUED)

 

Tropisternus sp.
Sperchopsis sp.
Hydroporus sp.
Gyrinus sp.

DIPTERA
Tipula sp.
Eriocera sp.
Antocha sp.
Chrysops sp.
Hemerodromia sp.
Simulium vittatum
Simulium tuberosum
Anopheles sp.
Ceratopogonidae
Chironomidae

   

 

77

"Illlllllll'lllllllls